In [0]:

    

import tensorflow as tf
import os
import itertools

from experimental.attentive_uncertainty.contextual_bandits.pretrain_gnp import train  # local file import


    

In [0]:

    

savedir = '/tmp/tfbind8/models/'
context_dim = 2
num_actions = 5
num_target = 50
num_context = 512
data_hparams = tf.contrib.training.HParams(context_dim=context_dim,
                                           num_actions=num_actions,
                                           num_target=num_target,
                                           num_context=num_context)


    

In [0]:

    

model_type = 'anp'
x_y_encoder_net_sizes = [HIDDEN_SIZE]*3
global_latent_net_sizes = [HIDDEN_SIZE]*2
local_latent_net_sizes = None
x_encoder_net_sizes = None
HIDDEN_SIZE = 64
decoder_net_sizes = [HIDDEN_SIZE]*3 + [2*num_actions]
heteroskedastic_net_sizes = None
att_type = 'multihead'
att_heads = 8
data_uncertainty = False
beta = 1
temp = 1


    

In [0]:

    

betas = [0., 0.5, 1., 2., 5., 1000.]
temps = [1e-3, 0.5, 1., 2., 5., 1000.]


    

In [0]:

    

for beta, temp in itertools.product(betas, temps):
  print('beta', beta, 'temperature', temp, flush=True)
  model_hparams = tf.contrib.training.HParams(activation=tf.nn.relu,
                                            output_activation=tf.nn.relu,
                                            x_encoder_net_sizes=x_encoder_net_sizes,
                                            x_y_encoder_net_sizes=x_y_encoder_net_sizes,
                                            global_latent_net_sizes=global_latent_net_sizes,
                                            local_latent_net_sizes=local_latent_net_sizes,
                                            decoder_net_sizes=decoder_net_sizes, 
                                            heteroskedastic_net_sizes=heteroskedastic_net_sizes,
                                            att_type=att_type,
                                            att_heads=att_heads,
                                            model_type=model_type,
                                            data_uncertainty=data_uncertainty,
                                            beta=beta,
                                            temperature=temp)
  save_path = os.path.join(savedir, 'gnp_' + model_type + '_beta_' + str(beta) + '_temp_'+ str(temp) + '.ckpt')
  training_hparams = tf.contrib.training.HParams(lr=0.01,
                                                optimizer=tf.train.RMSPropOptimizer,
                                                num_iterations=10000,
                                                batch_size=10,
                                                num_context=num_context,
                                                num_target=num_target, 
                                                print_every=100,
                                                save_path=save_path,
                                                max_grad_norm=10.0,
                                                is_nll=True)

  train(data_hparams,
        model_hparams,
        training_hparams)


    

    




beta 0.0 temperature 0.001
it: 0, train recon loss: 6387034.0, local kl: 0.0 global kl: 0.0 valid reconstr loss: 44108340.0
Saving best model with reconstruction loss 44108340.0
it: 100, train recon loss: 3.799004554748535, local kl: 0.0 global kl: 0.0 valid reconstr loss: 4.032503604888916
Saving best model with reconstruction loss 4.0325036
it: 200, train recon loss: 3.2728283405303955, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.3462297916412354
Saving best model with reconstruction loss 3.3462298
it: 300, train recon loss: 2.59476900100708, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.627716302871704
Saving best model with reconstruction loss 2.6277163
it: 400, train recon loss: 2.2547805309295654, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.3351688385009766
Saving best model with reconstruction loss 2.3351688
it: 500, train recon loss: 2.110807418823242, local kl: 0.0 global kl: 0.0 valid reconstr loss: 5.039649963378906
it: 600, train recon loss: 7.900534152984619, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.866126537322998
Saving best model with reconstruction loss 1.8661265
it: 700, train recon loss: 1.6320852041244507, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.045217514038086
it: 800, train recon loss: 2.019601821899414, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.877802848815918
it: 900, train recon loss: 26.55901527404785, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.2718411684036255
Saving best model with reconstruction loss 1.2718412
it: 1000, train recon loss: 0.9949962496757507, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.6291958093643188
Saving best model with reconstruction loss 0.6291958
it: 1100, train recon loss: 0.42802175879478455, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.20938223600387573
Saving best model with reconstruction loss 0.20938224
it: 1200, train recon loss: 1.417797327041626, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.227940559387207
it: 1300, train recon loss: 0.9711161255836487, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.7778622508049011
it: 1400, train recon loss: 0.3565886914730072, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.0157487392425537
it: 1500, train recon loss: 0.27513134479522705, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.25777605175971985
it: 1600, train recon loss: 1.742950439453125, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.8368628025054932
it: 1700, train recon loss: -0.39728808403015137, local kl: 0.0 global kl: 0.0 valid reconstr loss: 121.77002716064453
it: 1800, train recon loss: 0.05531693249940872, local kl: 0.0 global kl: 0.0 valid reconstr loss: 265.8903503417969
it: 1900, train recon loss: -0.219674214720726, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.12111188471317291
Saving best model with reconstruction loss -0.121111885
it: 2000, train recon loss: 0.14142096042633057, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.15484212338924408
it: 2100, train recon loss: -0.17071925103664398, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.35871103405952454
it: 2200, train recon loss: -0.5744224190711975, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.012645683251321316
it: 2300, train recon loss: -0.24646292626857758, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.00905059278011322
it: 2400, train recon loss: -0.15465623140335083, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.4059184789657593
Saving best model with reconstruction loss -0.40591848
it: 2500, train recon loss: -0.285388708114624, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.17367659509181976
it: 2600, train recon loss: -0.8173168897628784, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.431240051984787
Saving best model with reconstruction loss -0.43124005
it: 2700, train recon loss: -0.37164416909217834, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.13876119256019592
it: 2800, train recon loss: 860.3157958984375, local kl: 0.0 global kl: 0.0 valid reconstr loss: 11.182978630065918
it: 2900, train recon loss: 0.12703049182891846, local kl: 0.0 global kl: 0.0 valid reconstr loss: 124.81571960449219
it: 3000, train recon loss: 0.06254253536462784, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.6858776211738586
Saving best model with reconstruction loss -0.6858776
it: 3100, train recon loss: 0.297012597322464, local kl: 0.0 global kl: 0.0 valid reconstr loss: 203.7034912109375
it: 3200, train recon loss: -0.563001811504364, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.2840899229049683
it: 3300, train recon loss: 5033.67236328125, local kl: 0.0 global kl: 0.0 valid reconstr loss: 756.8186645507812
it: 3400, train recon loss: 9.258255004882812, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.27062180638313293
it: 3500, train recon loss: -0.6421890258789062, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.6702020168304443
it: 3600, train recon loss: -0.7033683657646179, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.7187628149986267
Saving best model with reconstruction loss -0.7187628
it: 3700, train recon loss: -1.0136516094207764, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1050.50341796875
it: 3800, train recon loss: 6.177923202514648, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.3502994477748871
it: 3900, train recon loss: -0.8736940622329712, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.7131285667419434
it: 4000, train recon loss: -1.2038121223449707, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.6526997089385986
it: 4100, train recon loss: 168.49237060546875, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.22593696415424347
it: 4200, train recon loss: 1.7797157764434814, local kl: 0.0 global kl: 0.0 valid reconstr loss: 4.108527660369873
it: 4300, train recon loss: -0.9908382892608643, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.7397663593292236
Saving best model with reconstruction loss -0.73976636
it: 4400, train recon loss: 338.0594482421875, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.8244545459747314
Saving best model with reconstruction loss -0.82445455
it: 4500, train recon loss: 0.12761345505714417, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.2929624915122986
it: 4600, train recon loss: -1.1973421573638916, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.010319709777832
Saving best model with reconstruction loss -1.0103197
it: 4700, train recon loss: -0.8725327253341675, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.7693243622779846
it: 4800, train recon loss: -1.2184964418411255, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.9735482931137085
it: 4900, train recon loss: -1.3859398365020752, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.7947551012039185
it: 5000, train recon loss: -0.31567567586898804, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1441.57470703125
it: 5100, train recon loss: -1.0107579231262207, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.0151691436767578
Saving best model with reconstruction loss -1.0151691
it: 5200, train recon loss: -0.9016862511634827, local kl: 0.0 global kl: 0.0 valid reconstr loss: 404.5569152832031
it: 5300, train recon loss: 891.6041259765625, local kl: 0.0 global kl: 0.0 valid reconstr loss: 11.477273941040039
it: 5400, train recon loss: -0.7098957300186157, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.9862231612205505
it: 5500, train recon loss: -1.4005135297775269, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.0639445781707764
Saving best model with reconstruction loss -1.0639446
it: 5600, train recon loss: -1.3667165040969849, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.9920927286148071
it: 5700, train recon loss: 5038.34912109375, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.0451289415359497
it: 5800, train recon loss: -1.030645728111267, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.058269739151001
it: 5900, train recon loss: -0.4670732617378235, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.3686424195766449
it: 6000, train recon loss: -1.5054653882980347, local kl: 0.0 global kl: 0.0 valid reconstr loss: 10234.03515625
it: 6100, train recon loss: -1.1694194078445435, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.0679458379745483
Saving best model with reconstruction loss -1.0679458
it: 6200, train recon loss: -1.308934211730957, local kl: 0.0 global kl: 0.0 valid reconstr loss: 7277.1787109375
it: 6300, train recon loss: -1.2051606178283691, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.14448444545269012
it: 6400, train recon loss: -0.9100425839424133, local kl: 0.0 global kl: 0.0 valid reconstr loss: 14969.3994140625
it: 6500, train recon loss: -1.1947896480560303, local kl: 0.0 global kl: 0.0 valid reconstr loss: 14.910797119140625
it: 6600, train recon loss: -1.1090412139892578, local kl: 0.0 global kl: 0.0 valid reconstr loss: 647.41748046875
it: 6700, train recon loss: -1.1088249683380127, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.2106472253799438
Saving best model with reconstruction loss -1.2106472
it: 6800, train recon loss: -1.1315897703170776, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.0128449201583862
it: 6900, train recon loss: 0.021583225578069687, local kl: 0.0 global kl: 0.0 valid reconstr loss: 60.044334411621094
it: 7000, train recon loss: -1.1013495922088623, local kl: 0.0 global kl: 0.0 valid reconstr loss: 10257.212890625
it: 7100, train recon loss: -1.0417348146438599, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.2074103355407715
it: 7200, train recon loss: -1.4586032629013062, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.1917887926101685
it: 7300, train recon loss: -1.1810437440872192, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.0149248838424683
it: 7400, train recon loss: -0.5872988104820251, local kl: 0.0 global kl: 0.0 valid reconstr loss: 11595.81640625
it: 7500, train recon loss: -1.5147688388824463, local kl: 0.0 global kl: 0.0 valid reconstr loss: 780.5067749023438
it: 7600, train recon loss: -1.1604467630386353, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.2145413160324097
Saving best model with reconstruction loss -1.2145413
it: 7700, train recon loss: -1.5009857416152954, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.2082933187484741
it: 7800, train recon loss: 926.412109375, local kl: 0.0 global kl: 0.0 valid reconstr loss: 49607.94921875
it: 7900, train recon loss: -1.1263600587844849, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.9407891631126404
it: 8000, train recon loss: 7290.47802734375, local kl: 0.0 global kl: 0.0 valid reconstr loss: 285.7080383300781
it: 8100, train recon loss: -1.3873811960220337, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.9733722805976868
it: 8200, train recon loss: 36.63344955444336, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.4706871807575226
it: 8300, train recon loss: -1.0171222686767578, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.056503176689148
it: 8400, train recon loss: 2.9313387870788574, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.119999647140503
it: 8500, train recon loss: 3.1596827507019043, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.096611499786377
it: 8600, train recon loss: 3.233673095703125, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.0812954902648926
it: 8700, train recon loss: 3.00834584236145, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.061717987060547
it: 8800, train recon loss: 2.8156094551086426, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.039207696914673
it: 8900, train recon loss: 2.884107828140259, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.0095481872558594
it: 9000, train recon loss: 2.7969892024993896, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.015036106109619
it: 9100, train recon loss: 3.1389572620391846, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.0473732948303223
it: 9200, train recon loss: 2.9471073150634766, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.0133562088012695
it: 9300, train recon loss: 2.9639689922332764, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.951353073120117
it: 9400, train recon loss: 3.0359139442443848, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.9704716205596924
it: 9500, train recon loss: 3.0372302532196045, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.9581868648529053
it: 9600, train recon loss: 2.92311954498291, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.0168793201446533
it: 9700, train recon loss: 3.006552219390869, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.951402187347412
it: 9800, train recon loss: 3.0331201553344727, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.9898688793182373
it: 9900, train recon loss: 2.737353563308716, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.95811128616333
beta 0.0 temperature 0.5
it: 0, train recon loss: 719.5963745117188, local kl: 0.0 global kl: 0.0 valid reconstr loss: 124.08468627929688
Saving best model with reconstruction loss 124.08469
it: 100, train recon loss: 3.8541831970214844, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.837512969970703
Saving best model with reconstruction loss 3.837513
it: 200, train recon loss: 3.2499186992645264, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.4432015419006348
Saving best model with reconstruction loss 3.4432015
it: 300, train recon loss: 2.864241123199463, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.8654470443725586
it: 400, train recon loss: 2.2838375568389893, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.6760644912719727
Saving best model with reconstruction loss 2.6760645
it: 500, train recon loss: 1.7591315507888794, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.2883105278015137
Saving best model with reconstruction loss 2.2883105
it: 600, train recon loss: 3.0008926391601562, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.600368618965149
Saving best model with reconstruction loss 1.6003686
it: 700, train recon loss: 3.6982247829437256, local kl: 0.0 global kl: 0.0 valid reconstr loss: 4.601300239562988
it: 800, train recon loss: 0.8122152090072632, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.7493829131126404
Saving best model with reconstruction loss 0.7493829
it: 900, train recon loss: 2.5465242862701416, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.840136706829071
it: 1000, train recon loss: 0.4858965575695038, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.3463010787963867
Saving best model with reconstruction loss 0.34630108
it: 1100, train recon loss: 0.3838435709476471, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.3856421113014221
it: 1200, train recon loss: 0.7927756309509277, local kl: 0.0 global kl: 0.0 valid reconstr loss: 160.5375518798828
it: 1300, train recon loss: -0.24498091638088226, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.9619116187095642
it: 1400, train recon loss: -0.15826809406280518, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.7257960438728333
it: 1500, train recon loss: -0.4091218113899231, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.4403083026409149
Saving best model with reconstruction loss -0.4403083
it: 1600, train recon loss: -0.1928856372833252, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.16405509412288666
it: 1700, train recon loss: 4.259724140167236, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.4959880113601685
it: 1800, train recon loss: -0.08679676055908203, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.33297502994537354
it: 1900, train recon loss: -0.5475417971611023, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.018695944920182228
it: 2000, train recon loss: -0.4177434742450714, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.5119761824607849
Saving best model with reconstruction loss -0.5119762
it: 2100, train recon loss: 1.0374712944030762, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.3125524520874023
it: 2200, train recon loss: -0.5960872173309326, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.6362454891204834
Saving best model with reconstruction loss -0.6362455
it: 2300, train recon loss: 0.42264342308044434, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.4582953453063965
it: 2400, train recon loss: -0.1424863636493683, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.6070533990859985
it: 2500, train recon loss: 305.9012756347656, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.5177480578422546
it: 2600, train recon loss: 0.8550962209701538, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1048.82177734375
it: 2700, train recon loss: -0.5367326736450195, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.37382978200912476
it: 2800, train recon loss: -0.9215459823608398, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.5531755685806274
it: 2900, train recon loss: -0.9794694781303406, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.4684799313545227
it: 3000, train recon loss: -0.7184836864471436, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.8223695158958435
Saving best model with reconstruction loss -0.8223695
it: 3100, train recon loss: -1.1304049491882324, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.8713809847831726
Saving best model with reconstruction loss -0.871381
it: 3200, train recon loss: -0.6214811205863953, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.6349345445632935
it: 3300, train recon loss: -0.9176385402679443, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.8342528343200684
it: 3400, train recon loss: 0.7730697989463806, local kl: 0.0 global kl: 0.0 valid reconstr loss: 372.74163818359375
it: 3500, train recon loss: -1.1189758777618408, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.4748579263687134
it: 3600, train recon loss: 3.7023518085479736, local kl: 0.0 global kl: 0.0 valid reconstr loss: 208.8046875
it: 3700, train recon loss: 15.730340957641602, local kl: 0.0 global kl: 0.0 valid reconstr loss: 7.784844875335693
it: 3800, train recon loss: -0.8172935843467712, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.547559916973114
it: 3900, train recon loss: -0.97147536277771, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.8057301044464111
it: 4000, train recon loss: -1.2166486978530884, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.5073103308677673
it: 4100, train recon loss: -1.2513784170150757, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.2761847674846649
it: 4200, train recon loss: -0.8633514046669006, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.0267075300216675
Saving best model with reconstruction loss -1.0267075
it: 4300, train recon loss: 4.932765483856201, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.44093629717826843
it: 4400, train recon loss: -1.0742868185043335, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.3989531397819519
it: 4500, train recon loss: -0.9910043478012085, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.970230221748352
it: 4600, train recon loss: -1.3383570909500122, local kl: 0.0 global kl: 0.0 valid reconstr loss: 202.09878540039062
it: 4700, train recon loss: -1.163667917251587, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.8321536779403687
it: 4800, train recon loss: -1.3187061548233032, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.5589970350265503
it: 4900, train recon loss: -0.31945762038230896, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.2736340761184692
Saving best model with reconstruction loss -1.2736341
it: 5000, train recon loss: -1.2688287496566772, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.5094085335731506
it: 5100, train recon loss: 0.6465798020362854, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.0871856212615967
it: 5200, train recon loss: -0.9819766283035278, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.2353107929229736
it: 5300, train recon loss: -1.0904310941696167, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.1917473077774048
it: 5400, train recon loss: -1.1760882139205933, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.2310715913772583
it: 5500, train recon loss: -1.2970465421676636, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.115028738975525
it: 5600, train recon loss: -1.465874195098877, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.9587532877922058
it: 5700, train recon loss: -1.0882294178009033, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.1562139987945557
it: 5800, train recon loss: -1.1125497817993164, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.2174711227416992
it: 5900, train recon loss: -1.409653663635254, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.0817444324493408
it: 6000, train recon loss: -1.4345139265060425, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.1651138067245483
it: 6100, train recon loss: -1.25657320022583, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.8998550772666931
it: 6200, train recon loss: -1.2517329454421997, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.2636208534240723
it: 6300, train recon loss: -1.0508261919021606, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.2484639883041382
it: 6400, train recon loss: -1.1816896200180054, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.07586309313774109
it: 6500, train recon loss: -1.2111479043960571, local kl: 0.0 global kl: 0.0 valid reconstr loss: 24.899131774902344
it: 6600, train recon loss: -1.2747923135757446, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.2344534397125244
it: 6700, train recon loss: -1.4031423330307007, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.0329488515853882
it: 6800, train recon loss: -1.3100130558013916, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.1311031579971313
it: 6900, train recon loss: -1.1282299757003784, local kl: 0.0 global kl: 0.0 valid reconstr loss: 4534.35693359375
it: 7000, train recon loss: -0.9202293753623962, local kl: 0.0 global kl: 0.0 valid reconstr loss: 6194.39599609375
it: 7100, train recon loss: -1.082519769668579, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.2602769136428833
it: 7200, train recon loss: -0.629696786403656, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.6906168460845947
it: 7300, train recon loss: -0.6997606158256531, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.9832119345664978
it: 7400, train recon loss: -1.0653138160705566, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.3831675946712494
it: 7500, train recon loss: -1.2364436388015747, local kl: 0.0 global kl: 0.0 valid reconstr loss: 26.52674102783203
it: 7600, train recon loss: 276.0284118652344, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.9856899976730347
it: 7700, train recon loss: -1.3633372783660889, local kl: 0.0 global kl: 0.0 valid reconstr loss: 78.94963836669922
it: 7800, train recon loss: 0.4072686433792114, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.0976804494857788
it: 7900, train recon loss: 3.92461895942688, local kl: 0.0 global kl: 0.0 valid reconstr loss: 5.800103664398193
it: 8000, train recon loss: -1.21321702003479, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.189724087715149
it: 8100, train recon loss: -1.0626633167266846, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.2878895998001099
Saving best model with reconstruction loss -1.2878896
it: 8200, train recon loss: 1.9266802072525024, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.6965173482894897
it: 8300, train recon loss: 4.290986061096191, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.6992335319519043
it: 8400, train recon loss: -0.8456237316131592, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.8532463908195496
it: 8500, train recon loss: -1.1483415365219116, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.1497600078582764
it: 8600, train recon loss: -0.8231987357139587, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.0385905504226685
it: 8700, train recon loss: -0.6374051570892334, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.9818601608276367
it: 8800, train recon loss: -1.31861412525177, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.1906483173370361
it: 8900, train recon loss: -1.4228217601776123, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.8284306526184082
it: 9000, train recon loss: -1.4883756637573242, local kl: 0.0 global kl: 0.0 valid reconstr loss: 43.22502517700195
it: 9100, train recon loss: 8415.6962890625, local kl: 0.0 global kl: 0.0 valid reconstr loss: 8022.17138671875
it: 9200, train recon loss: -0.8642818927764893, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1483.747802734375
it: 9300, train recon loss: -0.7117264866828918, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.2603098154067993
it: 9400, train recon loss: 0.582650363445282, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.103991985321045
it: 9500, train recon loss: -0.9930367469787598, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.2481657266616821
it: 9600, train recon loss: -0.2937847971916199, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.2297054529190063
it: 9700, train recon loss: -1.4090298414230347, local kl: 0.0 global kl: 0.0 valid reconstr loss: 4270.16357421875
it: 9800, train recon loss: -1.0588465929031372, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.2163177728652954
it: 9900, train recon loss: -1.3756991624832153, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.2482048273086548
beta 0.0 temperature 1.0
it: 0, train recon loss: 1239.9683837890625, local kl: 0.0 global kl: 0.0 valid reconstr loss: 299.8367614746094
Saving best model with reconstruction loss 299.83676
it: 100, train recon loss: 3.646623373031616, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.9198174476623535
Saving best model with reconstruction loss 3.9198174
it: 200, train recon loss: 4.037714958190918, local kl: 0.0 global kl: 0.0 valid reconstr loss: 4.128327369689941
it: 300, train recon loss: 6.227508068084717, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.8779218196868896
Saving best model with reconstruction loss 3.8779218
it: 400, train recon loss: 2.4268674850463867, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.6427299976348877
Saving best model with reconstruction loss 2.64273
it: 500, train recon loss: 2.519711494445801, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.3765478134155273
Saving best model with reconstruction loss 2.3765478
it: 600, train recon loss: 1.7510930299758911, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.4404149055480957
it: 700, train recon loss: 1.4227432012557983, local kl: 0.0 global kl: 0.0 valid reconstr loss: 4.2783308029174805
it: 800, train recon loss: 1.7609878778457642, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.5193018913269043
it: 900, train recon loss: 1.5089385509490967, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.3849260807037354
Saving best model with reconstruction loss 1.3849261
it: 1000, train recon loss: 1.3496545553207397, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.3999134302139282
it: 1100, train recon loss: 1.1543844938278198, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.9710596799850464
it: 1200, train recon loss: 3.6685526371002197, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.7682257890701294
it: 1300, train recon loss: 0.8271864056587219, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.9644854068756104
it: 1400, train recon loss: 850.7556762695312, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.8824923038482666
it: 1500, train recon loss: 1.0893620252609253, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.909169614315033
Saving best model with reconstruction loss 0.9091696
it: 1600, train recon loss: 1.2092324495315552, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.8912230134010315
Saving best model with reconstruction loss 0.891223
it: 1700, train recon loss: 1.0755995512008667, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.4177379608154297
it: 1800, train recon loss: 1.1182639598846436, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.9085147380828857
it: 1900, train recon loss: 1.1050139665603638, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.2603243589401245
it: 2000, train recon loss: 0.7683650851249695, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.2400908470153809
it: 2100, train recon loss: 0.7936732172966003, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.9868494868278503
it: 2200, train recon loss: 1.3919947147369385, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.3126318454742432
it: 2300, train recon loss: 4.040109634399414, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.8692145347595215
it: 2400, train recon loss: 2.1596298217773438, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.4155092239379883
it: 2500, train recon loss: 0.713392972946167, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.598286509513855
Saving best model with reconstruction loss 0.5982865
it: 2600, train recon loss: 0.5283148288726807, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.8087368011474609
it: 2700, train recon loss: 0.8952435255050659, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.6862415671348572
it: 2800, train recon loss: 0.29204586148262024, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.7395071387290955
it: 2900, train recon loss: 0.5943732857704163, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.7588067650794983
it: 3000, train recon loss: 0.9741893410682678, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.535632848739624
Saving best model with reconstruction loss 0.53563285
it: 3100, train recon loss: 1.310385823249817, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.5512298345565796
it: 3200, train recon loss: 1.6948795318603516, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.7081737518310547
it: 3300, train recon loss: 1.2343207597732544, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.47951745986938477
Saving best model with reconstruction loss 0.47951746
it: 3400, train recon loss: 1.1066985130310059, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.0251232385635376
it: 3500, train recon loss: 1.6788809299468994, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.7489948272705078
it: 3600, train recon loss: 1.743161916732788, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.8769818544387817
it: 3700, train recon loss: 1.3225882053375244, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.3376266956329346
it: 3800, train recon loss: 1.2704508304595947, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.9498618245124817
it: 3900, train recon loss: 0.32162895798683167, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.8074977993965149
it: 4000, train recon loss: 1.09632408618927, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.311348795890808
it: 4100, train recon loss: 0.754086971282959, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.7994309663772583
it: 4200, train recon loss: 0.7937208414077759, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.4920189678668976
it: 4300, train recon loss: 2.2547593116760254, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.6603124737739563
it: 4400, train recon loss: 0.6448424458503723, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.6490240097045898
it: 4500, train recon loss: 0.704608142375946, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.2502574920654297
it: 4600, train recon loss: 1.0861033201217651, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.9959543347358704
it: 4700, train recon loss: 0.7529430985450745, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.4598523676395416
Saving best model with reconstruction loss 0.45985237
it: 4800, train recon loss: 0.4001934826374054, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.4643203616142273
it: 4900, train recon loss: -0.07793654501438141, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.2635157108306885
it: 5000, train recon loss: 0.4019518196582794, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.590057373046875
it: 5100, train recon loss: 0.35284504294395447, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.554025411605835
it: 5200, train recon loss: -0.005007552914321423, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.37397876381874084
Saving best model with reconstruction loss 0.37397876
it: 5300, train recon loss: 0.8086912631988525, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.198067307472229
it: 5400, train recon loss: 0.7284178733825684, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.7120290398597717
it: 5500, train recon loss: 1.236629605293274, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.570954442024231
it: 5600, train recon loss: 0.7678843140602112, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.8244554400444031
it: 5700, train recon loss: 5.849344253540039, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.1924079656600952
it: 5800, train recon loss: 0.5482019782066345, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.6644598841667175
it: 5900, train recon loss: 0.5776606202125549, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.6209133267402649
it: 6000, train recon loss: 0.48395514488220215, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.6383625864982605
it: 6100, train recon loss: 0.834168016910553, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.4169115722179413
it: 6200, train recon loss: 0.48940542340278625, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.5254985094070435
it: 6300, train recon loss: 0.6446852684020996, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.6109945774078369
it: 6400, train recon loss: 0.559222400188446, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.7358071208000183
it: 6500, train recon loss: 1.0740373134613037, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.45771196484565735
it: 6600, train recon loss: 0.5034765601158142, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.9774506688117981
it: 6700, train recon loss: 0.38633090257644653, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.6982764005661011
it: 6800, train recon loss: 1.2848122119903564, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.3333609104156494
it: 6900, train recon loss: 1.2561919689178467, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.3370305299758911
it: 7000, train recon loss: 0.721408486366272, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.80422443151474
it: 7100, train recon loss: 0.37463077902793884, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.7209855318069458
it: 7200, train recon loss: 0.7524888515472412, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.6972939372062683
it: 7300, train recon loss: 0.5008729696273804, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.37905651330947876
it: 7400, train recon loss: 30.253686904907227, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.49566948413848877
it: 7500, train recon loss: 0.2893916666507721, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.5547820925712585
it: 7600, train recon loss: 0.510678768157959, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.49836912751197815
it: 7700, train recon loss: 0.17479251325130463, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.3576500713825226
Saving best model with reconstruction loss 0.35765007
it: 7800, train recon loss: 0.4606214463710785, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.4539027810096741
it: 7900, train recon loss: 0.5603080987930298, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.3645426630973816
it: 8000, train recon loss: 0.46071311831474304, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.41826409101486206
it: 8100, train recon loss: 0.48338109254837036, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.8252427577972412
it: 8200, train recon loss: 4.509214878082275, local kl: 0.0 global kl: 0.0 valid reconstr loss: 8.48759937286377
it: 8300, train recon loss: 15055.0068359375, local kl: 0.0 global kl: 0.0 valid reconstr loss: 129.37388610839844
it: 8400, train recon loss: 2.157925605773926, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.699094772338867
it: 8500, train recon loss: 2.8314311504364014, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.6420931816101074
it: 8600, train recon loss: 1.938557505607605, local kl: 0.0 global kl: 0.0 valid reconstr loss: 71898.5078125
it: 8700, train recon loss: 1.9326472282409668, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.139737367630005
it: 8800, train recon loss: 2.6932504177093506, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.6695553064346313
it: 8900, train recon loss: 1.8140074014663696, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.8450627326965332
it: 9000, train recon loss: 1.7724162340164185, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.6782246828079224
it: 9100, train recon loss: 2.4102931022644043, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.7438421249389648
it: 9200, train recon loss: 1.356424331665039, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.8312487602233887
it: 9300, train recon loss: 1.1890857219696045, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.7027928829193115
it: 9400, train recon loss: 1.4916552305221558, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.9002094268798828
it: 9500, train recon loss: 1.3404133319854736, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.5168359279632568
it: 9600, train recon loss: 403.4531555175781, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.8273191452026367
it: 9700, train recon loss: 186.36453247070312, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.8397215604782104
it: 9800, train recon loss: 1.0542353391647339, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.7870402336120605
it: 9900, train recon loss: 0.9008796215057373, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.217402458190918
beta 0.0 temperature 2.0
it: 0, train recon loss: 423.49017333984375, local kl: 0.0 global kl: 0.0 valid reconstr loss: 171.3223114013672
Saving best model with reconstruction loss 171.32231
it: 100, train recon loss: 3.7377498149871826, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.9349398612976074
Saving best model with reconstruction loss 3.9349399
it: 200, train recon loss: 4.322741985321045, local kl: 0.0 global kl: 0.0 valid reconstr loss: 4.531877517700195
it: 300, train recon loss: 3.2598536014556885, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.497943878173828
Saving best model with reconstruction loss 3.4979439
it: 400, train recon loss: 3.191739320755005, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.056008815765381
Saving best model with reconstruction loss 3.0560088
it: 500, train recon loss: 1.9464929103851318, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.1685540676116943
it: 600, train recon loss: 1.4107921123504639, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.110842704772949
Saving best model with reconstruction loss 2.1108427
it: 700, train recon loss: 1.0355596542358398, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.1353169679641724
Saving best model with reconstruction loss 1.135317
it: 800, train recon loss: 0.8838842511177063, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.7335807085037231
Saving best model with reconstruction loss 0.7335807
it: 900, train recon loss: 0.7560232877731323, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.7052778005599976
Saving best model with reconstruction loss 0.7052778
it: 1000, train recon loss: 0.4074609875679016, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.5452098846435547
Saving best model with reconstruction loss 0.5452099
it: 1100, train recon loss: 0.5040917992591858, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.5326642394065857
Saving best model with reconstruction loss 0.53266424
it: 1200, train recon loss: 0.2914314866065979, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.6140015125274658
it: 1300, train recon loss: 0.36530637741088867, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.1304289549589157
Saving best model with reconstruction loss 0.13042895
it: 1400, train recon loss: -0.3026677072048187, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.030453765764832497
Saving best model with reconstruction loss -0.030453766
it: 1500, train recon loss: -0.006817481946200132, local kl: 0.0 global kl: 0.0 valid reconstr loss: 5.537072658538818
it: 1600, train recon loss: 0.08115752041339874, local kl: 0.0 global kl: 0.0 valid reconstr loss: 140.61985778808594
it: 1700, train recon loss: -0.5809379816055298, local kl: 0.0 global kl: 0.0 valid reconstr loss: 456.9562072753906
it: 1800, train recon loss: -0.056027647107839584, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.2661046087741852
Saving best model with reconstruction loss -0.2661046
it: 1900, train recon loss: -0.5170518159866333, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.20731990039348602
it: 2000, train recon loss: -0.6121858954429626, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.38927796483039856
it: 2100, train recon loss: 4.213685035705566, local kl: 0.0 global kl: 0.0 valid reconstr loss: 4.166958332061768
it: 2200, train recon loss: 5.315570831298828, local kl: 0.0 global kl: 0.0 valid reconstr loss: 4.141904830932617
it: 2300, train recon loss: 7.940433979034424, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.5938143730163574
it: 2400, train recon loss: 3.3329153060913086, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.1672091484069824
it: 2500, train recon loss: 3.320523262023926, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.49544358253479
it: 2600, train recon loss: 3.2719883918762207, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.1460094451904297
it: 2700, train recon loss: 3.11743426322937, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.5319550037384033
it: 2800, train recon loss: 2.421694755554199, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.9407267570495605
it: 2900, train recon loss: 1.314133644104004, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.5307178497314453
it: 3000, train recon loss: 1.139766812324524, local kl: 0.0 global kl: 0.0 valid reconstr loss: 6.03854513168335
it: 3100, train recon loss: 0.7557908296585083, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.8057004809379578
it: 3200, train recon loss: 0.8549626469612122, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.5852779150009155
it: 3300, train recon loss: 0.7359768152236938, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.9780494570732117
it: 3400, train recon loss: 0.13384929299354553, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1258.9619140625
it: 3500, train recon loss: 0.63546222448349, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.7525849938392639
it: 3600, train recon loss: 0.14544540643692017, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.042686939239502
it: 3700, train recon loss: -0.09403744339942932, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.2015519142150879
it: 3800, train recon loss: 0.03774639591574669, local kl: 0.0 global kl: 0.0 valid reconstr loss: 34.87210464477539
it: 3900, train recon loss: -0.08790402114391327, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.07218988239765167
it: 4000, train recon loss: -0.6045308709144592, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.14241516590118408
it: 4100, train recon loss: -0.3991522490978241, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.8365988731384277
it: 4200, train recon loss: -0.02773445099592209, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.5483104586601257
Saving best model with reconstruction loss -0.54831046
it: 4300, train recon loss: -0.4039359986782074, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.03090069629251957
it: 4400, train recon loss: -0.14048132300376892, local kl: 0.0 global kl: 0.0 valid reconstr loss: 6.404007911682129
it: 4500, train recon loss: 0.023486649617552757, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.24751873314380646
it: 4600, train recon loss: 8.868361473083496, local kl: 0.0 global kl: 0.0 valid reconstr loss: 63.82201385498047
it: 4700, train recon loss: -0.5125324726104736, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.4619584381580353
it: 4800, train recon loss: -0.7065737843513489, local kl: 0.0 global kl: 0.0 valid reconstr loss: 15.0985689163208
it: 4900, train recon loss: -0.4235965609550476, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.15431128442287445
it: 5000, train recon loss: -0.668428361415863, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.588712215423584
it: 5100, train recon loss: -0.41660431027412415, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.2782893478870392
it: 5200, train recon loss: -0.16444870829582214, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.4270799160003662
it: 5300, train recon loss: -0.5669248700141907, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.29629752039909363
it: 5400, train recon loss: -0.5534210205078125, local kl: 0.0 global kl: 0.0 valid reconstr loss: 28.780858993530273
it: 5500, train recon loss: -0.70354163646698, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.08335348218679428
it: 5600, train recon loss: -1.005014181137085, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.61949622631073
Saving best model with reconstruction loss -0.6194962
it: 5700, train recon loss: -0.4907837510108948, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.47029614448547363
it: 5800, train recon loss: -0.6857075691223145, local kl: 0.0 global kl: 0.0 valid reconstr loss: 475.8572692871094
it: 5900, train recon loss: -0.7968982458114624, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.5837432742118835
it: 6000, train recon loss: -0.43628576397895813, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.5760196447372437
it: 6100, train recon loss: -0.908057689666748, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.015011215582489967
it: 6200, train recon loss: -0.6554549336433411, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2130.586181640625
it: 6300, train recon loss: -0.7362249493598938, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.8416266441345215
Saving best model with reconstruction loss -0.84162664
it: 6400, train recon loss: -0.7157974243164062, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.3547360301017761
it: 6500, train recon loss: -0.8331186771392822, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.3323133587837219
it: 6600, train recon loss: -0.4053398370742798, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.6876675486564636
it: 6700, train recon loss: -0.8558889031410217, local kl: 0.0 global kl: 0.0 valid reconstr loss: 188.00595092773438
it: 6800, train recon loss: -0.7446691393852234, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.5873979926109314
it: 6900, train recon loss: 0.6567082405090332, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.5252069234848022
it: 7000, train recon loss: -0.5158934593200684, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.38725799322128296
it: 7100, train recon loss: 0.21622666716575623, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.697354257106781
it: 7200, train recon loss: 4654.37451171875, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.7185963988304138
it: 7300, train recon loss: 12.573673248291016, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1017.7664184570312
it: 7400, train recon loss: 12.010482788085938, local kl: 0.0 global kl: 0.0 valid reconstr loss: 300.9341125488281
it: 7500, train recon loss: 3.878800392150879, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.859071731567383
it: 7600, train recon loss: 2.9739785194396973, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.065006732940674
it: 7700, train recon loss: 1.8574819564819336, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.1380741596221924
it: 7800, train recon loss: 1.5129443407058716, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.5449864864349365
it: 7900, train recon loss: 2.779031276702881, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.107843041419983
it: 8000, train recon loss: 0.6750022768974304, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.0272808074951172
it: 8100, train recon loss: 12.647054672241211, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.5873510837554932
it: 8200, train recon loss: 0.6177831888198853, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.5754653215408325
it: 8300, train recon loss: 0.4293174147605896, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.12400682270526886
it: 8400, train recon loss: 0.589385986328125, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.27177804708480835
it: 8500, train recon loss: -0.005691963247954845, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.6062710285186768
it: 8600, train recon loss: 6.98002815246582, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.1682378500699997
it: 8700, train recon loss: 3.178361415863037, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1349.5465087890625
it: 8800, train recon loss: -0.1787811666727066, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.11645621806383133
it: 8900, train recon loss: -0.41007205843925476, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.1808498650789261
it: 9000, train recon loss: -0.33017298579216003, local kl: 0.0 global kl: 0.0 valid reconstr loss: 566.7682495117188
it: 9100, train recon loss: -0.0695415809750557, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.5226340293884277
it: 9200, train recon loss: 48.688697814941406, local kl: 0.0 global kl: 0.0 valid reconstr loss: 4958.78515625
it: 9300, train recon loss: -0.3558804392814636, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.594444215297699
it: 9400, train recon loss: -0.41426026821136475, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.23589545488357544
it: 9500, train recon loss: -0.5658360123634338, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.4891943335533142
it: 9600, train recon loss: -0.17553958296775818, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.7915793657302856
it: 9700, train recon loss: -0.1629447042942047, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.0255056619644165
it: 9800, train recon loss: 807.4552612304688, local kl: 0.0 global kl: 0.0 valid reconstr loss: 402053344.0
it: 9900, train recon loss: -0.44494879245758057, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.49434879422187805
beta 0.0 temperature 5.0
it: 0, train recon loss: 335.3072814941406, local kl: 0.0 global kl: 0.0 valid reconstr loss: 396.21148681640625
Saving best model with reconstruction loss 396.2115
it: 100, train recon loss: 3.81740403175354, local kl: 0.0 global kl: 0.0 valid reconstr loss: 4.522944450378418
Saving best model with reconstruction loss 4.5229445
it: 200, train recon loss: 3.909613609313965, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.9250295162200928
Saving best model with reconstruction loss 3.9250295
it: 300, train recon loss: 3.780914545059204, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.9009039402008057
Saving best model with reconstruction loss 3.900904
it: 400, train recon loss: 3.0990889072418213, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.3866498470306396
Saving best model with reconstruction loss 3.3866498
it: 500, train recon loss: 2.552492618560791, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.8080101013183594
Saving best model with reconstruction loss 2.80801
it: 600, train recon loss: 1.750773549079895, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.9593995809555054
Saving best model with reconstruction loss 1.9593996
it: 700, train recon loss: 1.1247199773788452, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.323281168937683
Saving best model with reconstruction loss 1.3232812
it: 800, train recon loss: 1.2684303522109985, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.2415049076080322
Saving best model with reconstruction loss 1.2415049
it: 900, train recon loss: 1.2190006971359253, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.010211706161499
Saving best model with reconstruction loss 1.0102117
it: 1000, train recon loss: 1.4039108753204346, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.0134464502334595
it: 1100, train recon loss: 0.5834790468215942, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.7578284740447998
Saving best model with reconstruction loss 0.7578285
it: 1200, train recon loss: 0.5116479992866516, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.3687126636505127
it: 1300, train recon loss: 7.290945053100586, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.7110438346862793
Saving best model with reconstruction loss 0.71104383
it: 1400, train recon loss: 3.971310615539551, local kl: 0.0 global kl: 0.0 valid reconstr loss: 4.242537021636963
it: 1500, train recon loss: 3.470386505126953, local kl: 0.0 global kl: 0.0 valid reconstr loss: 5.870308876037598
it: 1600, train recon loss: 3.1632750034332275, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.3552069664001465
it: 1700, train recon loss: 2.561929225921631, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.84424889087677
it: 1800, train recon loss: 0.7115905284881592, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.9946154356002808
it: 1900, train recon loss: 0.6815507411956787, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.748683750629425
it: 2000, train recon loss: 0.7425452470779419, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.8873979449272156
it: 2100, train recon loss: 0.4011002480983734, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.714877188205719
it: 2200, train recon loss: 0.04575018212199211, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.6525130271911621
Saving best model with reconstruction loss 0.652513
it: 2300, train recon loss: 0.3281480669975281, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.28872233629226685
Saving best model with reconstruction loss 0.28872234
it: 2400, train recon loss: 1.6707651615142822, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.1117043495178223
it: 2500, train recon loss: 0.8065871596336365, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.6894826889038086
it: 2600, train recon loss: -0.23474159836769104, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.029995430260896683
Saving best model with reconstruction loss 0.02999543
it: 2700, train recon loss: 0.4495563805103302, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.7157716751098633
it: 2800, train recon loss: -0.4316357672214508, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.6588285565376282
it: 2900, train recon loss: -0.45816779136657715, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.1101481169462204
it: 3000, train recon loss: -0.3365594446659088, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.18500842154026031
Saving best model with reconstruction loss -0.18500842
it: 3100, train recon loss: -0.6678529977798462, local kl: 0.0 global kl: 0.0 valid reconstr loss: 501.74609375
it: 3200, train recon loss: -0.6412489414215088, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.27066874504089355
it: 3300, train recon loss: -0.6004784107208252, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.5384457111358643
Saving best model with reconstruction loss -0.5384457
it: 3400, train recon loss: -0.7232464551925659, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.26698634028434753
it: 3500, train recon loss: -0.6965947151184082, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.656711220741272
Saving best model with reconstruction loss -0.6567112
it: 3600, train recon loss: -0.6422535181045532, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.25503548979759216
it: 3700, train recon loss: -0.5572492480278015, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.6892637610435486
Saving best model with reconstruction loss -0.68926376
it: 3800, train recon loss: -0.766207218170166, local kl: 0.0 global kl: 0.0 valid reconstr loss: 83.65840148925781
it: 3900, train recon loss: -0.6289870738983154, local kl: 0.0 global kl: 0.0 valid reconstr loss: 15.352360725402832
it: 4000, train recon loss: -0.6343206763267517, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.8331045508384705
Saving best model with reconstruction loss -0.83310455
it: 4100, train recon loss: 22.349634170532227, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.7374066114425659
it: 4200, train recon loss: -0.3868575990200043, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.721003532409668
it: 4300, train recon loss: -0.2453071027994156, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.834837019443512
Saving best model with reconstruction loss -0.834837
it: 4400, train recon loss: -0.9945761561393738, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.5613254308700562
it: 4500, train recon loss: -0.8431412577629089, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.8632501363754272
Saving best model with reconstruction loss -0.86325014
it: 4600, train recon loss: 0.4189586341381073, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.0250945091247559
Saving best model with reconstruction loss -1.0250945
it: 4700, train recon loss: -0.8130814433097839, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.8210074305534363
it: 4800, train recon loss: -1.2204139232635498, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.6593442559242249
it: 4900, train recon loss: -1.1082592010498047, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.9423577189445496
it: 5000, train recon loss: -0.7884652018547058, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.8420218229293823
it: 5100, train recon loss: -0.6859878897666931, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.955750584602356
it: 5200, train recon loss: 3.7803733348846436, local kl: 0.0 global kl: 0.0 valid reconstr loss: 5.2681884765625
it: 5300, train recon loss: 2.8101108074188232, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.130138397216797
it: 5400, train recon loss: 3.1778407096862793, local kl: 0.0 global kl: 0.0 valid reconstr loss: 5.0356879234313965
it: 5500, train recon loss: 2.515721082687378, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.80122971534729
it: 5600, train recon loss: 0.7894365191459656, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.27562421560287476
it: 5700, train recon loss: 279.8890380859375, local kl: 0.0 global kl: 0.0 valid reconstr loss: 607.8740234375
it: 5800, train recon loss: -0.6556195616722107, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.6159858107566833
it: 5900, train recon loss: -1.056037187576294, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.6908010244369507
it: 6000, train recon loss: -0.41039425134658813, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.5563753843307495
it: 6100, train recon loss: -0.46263620257377625, local kl: 0.0 global kl: 0.0 valid reconstr loss: 231.78382873535156
it: 6200, train recon loss: 2.380772352218628, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.5424685478210449
it: 6300, train recon loss: -0.7092760801315308, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.531481146812439
it: 6400, train recon loss: 606.5640258789062, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.7135154008865356
it: 6500, train recon loss: -0.533838152885437, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.7865197658538818
it: 6600, train recon loss: -0.9200351238250732, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.7712985873222351
it: 6700, train recon loss: 12.285896301269531, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3412.0361328125
it: 6800, train recon loss: 47.78459548950195, local kl: 0.0 global kl: 0.0 valid reconstr loss: 8982684.0
it: 6900, train recon loss: -0.7416858077049255, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.8954954743385315
it: 7000, train recon loss: 143.9360809326172, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.7970762252807617
it: 7100, train recon loss: -0.8336532711982727, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.0798430442810059
Saving best model with reconstruction loss -1.079843
it: 7200, train recon loss: 2.6491079330444336, local kl: 0.0 global kl: 0.0 valid reconstr loss: 129.4964599609375
it: 7300, train recon loss: -0.1271759271621704, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.8137320876121521
it: 7400, train recon loss: 177.87440490722656, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.415382146835327
it: 7500, train recon loss: 0.08712512999773026, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.9287977814674377
it: 7600, train recon loss: -0.6246806979179382, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.6435983180999756
it: 7700, train recon loss: -1.0462772846221924, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.9883207082748413
it: 7800, train recon loss: -0.8892644047737122, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.07007312029600143
it: 7900, train recon loss: 0.4204608201980591, local kl: 0.0 global kl: 0.0 valid reconstr loss: 337.57958984375
it: 8000, train recon loss: 1.6073158979415894, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.0374951362609863
it: 8100, train recon loss: 2.0182361602783203, local kl: 0.0 global kl: 0.0 valid reconstr loss: 79.11174774169922
it: 8200, train recon loss: 0.5217728614807129, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.131646752357483
it: 8300, train recon loss: 2.894605875015259, local kl: 0.0 global kl: 0.0 valid reconstr loss: 5.728950500488281
it: 8400, train recon loss: -0.8248770833015442, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.32453301548957825
it: 8500, train recon loss: -0.45630744099617004, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.0234774351119995
it: 8600, train recon loss: 38742.89453125, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.5472376942634583
it: 8700, train recon loss: 19.500946044921875, local kl: 0.0 global kl: 0.0 valid reconstr loss: 33.16558074951172
it: 8800, train recon loss: -1.1439135074615479, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.8738557696342468
it: 8900, train recon loss: -0.9947298765182495, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.9669309854507446
it: 9000, train recon loss: -1.1527776718139648, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.0435672998428345
it: 9100, train recon loss: -0.8021901249885559, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.1141672134399414
Saving best model with reconstruction loss -1.1141672
it: 9200, train recon loss: -1.181738018989563, local kl: 0.0 global kl: 0.0 valid reconstr loss: 368.2458801269531
it: 9300, train recon loss: -0.9455931186676025, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.9783245325088501
it: 9400, train recon loss: -0.9908097982406616, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.349497079849243
it: 9500, train recon loss: -0.8689686059951782, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.9968957304954529
it: 9600, train recon loss: -0.9824861884117126, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.9128878116607666
it: 9700, train recon loss: -0.8099238276481628, local kl: 0.0 global kl: 0.0 valid reconstr loss: 253.07171630859375
it: 9800, train recon loss: -0.8503050208091736, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.9338252544403076
it: 9900, train recon loss: -1.1325860023498535, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.0415514707565308
beta 0.0 temperature 1000.0
it: 0, train recon loss: 1749.7825927734375, local kl: 0.0 global kl: 0.0 valid reconstr loss: 231.23291015625
Saving best model with reconstruction loss 231.23291
it: 100, train recon loss: 61098540.0, local kl: 0.0 global kl: 0.0 valid reconstr loss: 4.266154766082764
Saving best model with reconstruction loss 4.266155
it: 200, train recon loss: 4.084741115570068, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.9994633197784424
Saving best model with reconstruction loss 3.9994633
it: 300, train recon loss: 4.073908805847168, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.8007259368896484
Saving best model with reconstruction loss 3.800726
it: 400, train recon loss: 3.357111930847168, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.9326441287994385
Saving best model with reconstruction loss 2.9326441
it: 500, train recon loss: 2.001112937927246, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.607705593109131
it: 600, train recon loss: 2.7804086208343506, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.755932331085205
Saving best model with reconstruction loss 2.7559323
it: 700, train recon loss: 1.9648381471633911, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.03020977973938
Saving best model with reconstruction loss 2.0302098
it: 800, train recon loss: 1.3492079973220825, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.2884840965270996
Saving best model with reconstruction loss 1.2884841
it: 900, train recon loss: 1.3214620351791382, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.3014461994171143
it: 1000, train recon loss: 3.622807264328003, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.850867748260498
it: 1100, train recon loss: 3.4975240230560303, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.5830698013305664
it: 1200, train recon loss: 3.4330170154571533, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.5031306743621826
it: 1300, train recon loss: 3.234985589981079, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.517305612564087
it: 1400, train recon loss: 3.2971012592315674, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.203155755996704
it: 1500, train recon loss: 1.7265067100524902, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.9682891368865967
it: 1600, train recon loss: 1.5477300882339478, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.7890987396240234
it: 1700, train recon loss: 1.0734641551971436, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.439800500869751
it: 1800, train recon loss: 1.2008849382400513, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.0991661548614502
Saving best model with reconstruction loss 1.0991662
it: 1900, train recon loss: 1.3846379518508911, local kl: 0.0 global kl: 0.0 valid reconstr loss: 6.187711238861084
it: 2000, train recon loss: 1.2227506637573242, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.7972519397735596
Saving best model with reconstruction loss 0.79725194
it: 2100, train recon loss: 0.8492265343666077, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.8170004487037659
it: 2200, train recon loss: 0.05925031378865242, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.6862220168113708
Saving best model with reconstruction loss 0.686222
it: 2300, train recon loss: 10.679670333862305, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.377225875854492
it: 2400, train recon loss: 15.836116790771484, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.2996799945831299
Saving best model with reconstruction loss 0.29968
it: 2500, train recon loss: 0.5428287982940674, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.10211379081010818
Saving best model with reconstruction loss 0.10211379
it: 2600, train recon loss: -0.30837225914001465, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.73797607421875
it: 2700, train recon loss: 1.1650269031524658, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1862.1151123046875
it: 2800, train recon loss: 3.4774105548858643, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.17667581140995026
Saving best model with reconstruction loss -0.17667581
it: 2900, train recon loss: -0.19162331521511078, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.14240899682044983
it: 3000, train recon loss: 0.2804172933101654, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.2526029348373413
it: 3100, train recon loss: 0.49638545513153076, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.2505981922149658
Saving best model with reconstruction loss -0.2505982
it: 3200, train recon loss: 50.695472717285156, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.04645240679383278
it: 3300, train recon loss: -0.11969591677188873, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.1899387687444687
it: 3400, train recon loss: -0.3222772479057312, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.647223949432373
it: 3500, train recon loss: -0.4614163935184479, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.21153892576694489
it: 3600, train recon loss: -0.42009660601615906, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.28508898615837097
it: 3700, train recon loss: -0.5561041235923767, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.15966945886611938
it: 3800, train recon loss: -0.023131782189011574, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.530582845211029
Saving best model with reconstruction loss -0.53058285
it: 3900, train recon loss: -0.5400124192237854, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.3221554756164551
it: 4000, train recon loss: -0.9354110360145569, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.19843034446239471
it: 4100, train recon loss: -0.7961596250534058, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.03473089635372162
it: 4200, train recon loss: -0.1900913566350937, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.45374348759651184
it: 4300, train recon loss: -0.6901845335960388, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.618448793888092
Saving best model with reconstruction loss -0.6184488
it: 4400, train recon loss: -0.26738160848617554, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.42938685417175293
it: 4500, train recon loss: -0.7942544221878052, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.610234797000885
it: 4600, train recon loss: -0.8451592922210693, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.6291857361793518
Saving best model with reconstruction loss -0.62918574
it: 4700, train recon loss: -0.3616108000278473, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.7157253623008728
Saving best model with reconstruction loss -0.71572536
it: 4800, train recon loss: -0.9718771576881409, local kl: 0.0 global kl: 0.0 valid reconstr loss: 282.29248046875
it: 4900, train recon loss: -0.9053916335105896, local kl: 0.0 global kl: 0.0 valid reconstr loss: 699.6489868164062
it: 5000, train recon loss: -0.5639207363128662, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.6616871356964111
it: 5100, train recon loss: -0.7354364395141602, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.7204939126968384
Saving best model with reconstruction loss -0.7204939
it: 5200, train recon loss: 1321.9854736328125, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.4851935803890228
it: 5300, train recon loss: -0.7336492538452148, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.5032526254653931
it: 5400, train recon loss: -0.8041770458221436, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.8318588733673096
Saving best model with reconstruction loss -0.8318589
it: 5500, train recon loss: -0.7910515069961548, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.8880369663238525
Saving best model with reconstruction loss -0.88803697
it: 5600, train recon loss: 5.269011974334717, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.7021155953407288
it: 5700, train recon loss: -0.5900288224220276, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.3256085216999054
it: 5800, train recon loss: -1.020426630973816, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.8843429088592529
it: 5900, train recon loss: -0.7115097641944885, local kl: 0.0 global kl: 0.0 valid reconstr loss: 10.953478813171387
it: 6000, train recon loss: -0.8552223443984985, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.44333046674728394
it: 6100, train recon loss: -0.7524386048316956, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.15181586146354675
it: 6200, train recon loss: -1.1288903951644897, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.8201348185539246
it: 6300, train recon loss: -0.6440576314926147, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.9211224317550659
Saving best model with reconstruction loss -0.92112243
it: 6400, train recon loss: 1094.224853515625, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2185.2685546875
it: 6500, train recon loss: -0.47078439593315125, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.8794786334037781
it: 6600, train recon loss: 5.590939998626709, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.7927379012107849
it: 6700, train recon loss: -1.1611006259918213, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.587803840637207
it: 6800, train recon loss: -1.0395784378051758, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.9856030941009521
Saving best model with reconstruction loss -0.9856031
it: 6900, train recon loss: -0.8123779892921448, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.0276511907577515
Saving best model with reconstruction loss -1.0276512
it: 7000, train recon loss: -0.4459500312805176, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.9655831456184387
it: 7100, train recon loss: -1.0023561716079712, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.8387416005134583
it: 7200, train recon loss: 189.8543701171875, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.5679975748062134
it: 7300, train recon loss: -0.4034551680088043, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.1038756370544434
Saving best model with reconstruction loss -1.1038756
it: 7400, train recon loss: 54.454742431640625, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.0077651739120483
it: 7500, train recon loss: -1.2888554334640503, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.15037648379802704
it: 7600, train recon loss: -0.8788406848907471, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.0408309698104858
it: 7700, train recon loss: -1.26964271068573, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.09571375697851181
it: 7800, train recon loss: -0.9800695180892944, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.9263272285461426
it: 7900, train recon loss: -0.8724908232688904, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.0217632055282593
it: 8000, train recon loss: -1.1990242004394531, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.0408945083618164
it: 8100, train recon loss: 5.360708236694336, local kl: 0.0 global kl: 0.0 valid reconstr loss: 5.304825305938721
it: 8200, train recon loss: 4.352898120880127, local kl: 0.0 global kl: 0.0 valid reconstr loss: 48.23567199707031
it: 8300, train recon loss: 5.051218032836914, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.8147978782653809
it: 8400, train recon loss: -1.0265791416168213, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.7146909832954407
it: 8500, train recon loss: -0.8588811755180359, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.0150116682052612
it: 8600, train recon loss: -1.089416742324829, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.31165191531181335
it: 8700, train recon loss: 3333.4677734375, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.7595304846763611
it: 8800, train recon loss: -1.2997112274169922, local kl: 0.0 global kl: 0.0 valid reconstr loss: 4.037296295166016
it: 8900, train recon loss: 7.8243327140808105, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.0197865962982178
it: 9000, train recon loss: -1.2592554092407227, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.7166915535926819
it: 9100, train recon loss: 5114.51806640625, local kl: 0.0 global kl: 0.0 valid reconstr loss: 8830.703125
it: 9200, train recon loss: 4.88761043548584, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.9370226860046387
it: 9300, train recon loss: -0.4997693598270416, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.188031792640686
Saving best model with reconstruction loss -1.1880318
it: 9400, train recon loss: -1.0928412675857544, local kl: 0.0 global kl: 0.0 valid reconstr loss: 6886.697265625
it: 9500, train recon loss: 1116.3369140625, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.1827337741851807
it: 9600, train recon loss: 0.4133402109146118, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.5959275960922241
it: 9700, train recon loss: 5.208222389221191, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.4524197280406952
it: 9800, train recon loss: -0.9268180131912231, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.994110107421875
it: 9900, train recon loss: -1.3256937265396118, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.0256816148757935
beta 0.5 temperature 0.001
it: 0, train recon loss: 65022560.0, local kl: 0.0 global kl: 0.007208447903394699 valid reconstr loss: 8727100.0
Saving best model with reconstruction loss 8727100.0
it: 100, train recon loss: 3.7100563049316406, local kl: 0.0 global kl: 0.0008413156610913575 valid reconstr loss: 4.0715131759643555
Saving best model with reconstruction loss 4.071513
it: 200, train recon loss: 3.652820348739624, local kl: 0.0 global kl: 0.00022017236915417016 valid reconstr loss: 3.3246870040893555
Saving best model with reconstruction loss 3.324687
it: 300, train recon loss: 2.8462157249450684, local kl: 0.0 global kl: 4.284338137949817e-05 valid reconstr loss: 2.608734607696533
Saving best model with reconstruction loss 2.6087346
it: 400, train recon loss: 2.4481379985809326, local kl: 0.0 global kl: 0.00012740830425173044 valid reconstr loss: 2.1109390258789062
Saving best model with reconstruction loss 2.110939
it: 500, train recon loss: 1.6291704177856445, local kl: 0.0 global kl: 0.0001530434819869697 valid reconstr loss: 1.4357550144195557
Saving best model with reconstruction loss 1.435755
it: 600, train recon loss: 1.0917967557907104, local kl: 0.0 global kl: 0.0003153732977807522 valid reconstr loss: 0.9707289934158325
Saving best model with reconstruction loss 0.970729
it: 700, train recon loss: 0.9829326272010803, local kl: 0.0 global kl: 0.0005968910409137607 valid reconstr loss: 6.899240016937256
it: 800, train recon loss: 725.1961669921875, local kl: 0.0 global kl: 0.00027505357866175473 valid reconstr loss: 1.5816352367401123
it: 900, train recon loss: 4.299914836883545, local kl: 0.0 global kl: 6.600162305403501e-05 valid reconstr loss: 17.761512756347656
it: 1000, train recon loss: 3.151331901550293, local kl: 0.0 global kl: 1.0517058144504654e-12 valid reconstr loss: 3.3154969215393066
it: 1100, train recon loss: 3.0775396823883057, local kl: 0.0 global kl: 6.299030741452327e-12 valid reconstr loss: 2.878319501876831
it: 1200, train recon loss: 1.97402024269104, local kl: 0.0 global kl: 1.770250612764812e-13 valid reconstr loss: 4.139997482299805
it: 1300, train recon loss: 1.8153941631317139, local kl: 0.0 global kl: 2.4128392426409073e-12 valid reconstr loss: 132.34095764160156
it: 1400, train recon loss: 0.9293930530548096, local kl: 0.0 global kl: 1.3860163017298532e-12 valid reconstr loss: 2.229388475418091
it: 1500, train recon loss: 0.6237261295318604, local kl: 0.0 global kl: 2.5202062658991053e-12 valid reconstr loss: 0.8762269616127014
Saving best model with reconstruction loss 0.87622696
it: 1600, train recon loss: 0.60980623960495, local kl: 0.0 global kl: 1.2637668689308157e-12 valid reconstr loss: 0.3334692418575287
Saving best model with reconstruction loss 0.33346924
it: 1700, train recon loss: 10.653547286987305, local kl: 0.0 global kl: 1.6780368761182274e-11 valid reconstr loss: 0.431106299161911
it: 1800, train recon loss: 0.16578397154808044, local kl: 0.0 global kl: 9.690660004047641e-15 valid reconstr loss: 0.19104409217834473
Saving best model with reconstruction loss 0.19104409
it: 1900, train recon loss: 0.4796474277973175, local kl: 0.0 global kl: 4.0650260935137794e-12 valid reconstr loss: 0.1838192492723465
Saving best model with reconstruction loss 0.18381925
it: 2000, train recon loss: -0.2691743075847626, local kl: 0.0 global kl: 1.859182946484239e-12 valid reconstr loss: 0.24046094715595245
it: 2100, train recon loss: 0.06563717871904373, local kl: 0.0 global kl: 1.4266365866433262e-13 valid reconstr loss: 0.051936376839876175
Saving best model with reconstruction loss 0.051936377
it: 2200, train recon loss: -0.39470142126083374, local kl: 0.0 global kl: 4.534778802467443e-11 valid reconstr loss: 8.401134490966797
it: 2300, train recon loss: -0.18791843950748444, local kl: 0.0 global kl: 3.036613929058296e-12 valid reconstr loss: 0.8123766183853149
it: 2400, train recon loss: 18.410661697387695, local kl: 0.0 global kl: 3.03445636673505e-13 valid reconstr loss: 40.37916946411133
it: 2500, train recon loss: -0.35728368163108826, local kl: 0.0 global kl: 4.783346028297553e-13 valid reconstr loss: 163.54222106933594
it: 2600, train recon loss: -0.5934011340141296, local kl: 0.0 global kl: 1.4395359904106897e-12 valid reconstr loss: -0.37921300530433655
Saving best model with reconstruction loss -0.379213
it: 2700, train recon loss: -0.3181394636631012, local kl: 0.0 global kl: 5.447919892986874e-12 valid reconstr loss: -0.34005749225616455
it: 2800, train recon loss: -0.27702876925468445, local kl: 0.0 global kl: 2.208025429162319e-12 valid reconstr loss: -0.6608628034591675
Saving best model with reconstruction loss -0.6608628
it: 2900, train recon loss: -0.5216187834739685, local kl: 0.0 global kl: 2.441019608667716e-12 valid reconstr loss: -0.7514338493347168
Saving best model with reconstruction loss -0.75143385
it: 3000, train recon loss: -0.6125711798667908, local kl: 0.0 global kl: 9.156239134944233e-14 valid reconstr loss: -0.4777279198169708
it: 3100, train recon loss: 6.133727073669434, local kl: 0.0 global kl: 3.3255759257500017e-12 valid reconstr loss: 1.3724042177200317
it: 3200, train recon loss: -0.754298746585846, local kl: 0.0 global kl: 2.099723173110135e-12 valid reconstr loss: -0.3692876696586609
it: 3300, train recon loss: 8.809680938720703, local kl: 0.0 global kl: 1.9675927553919337e-13 valid reconstr loss: 0.8008824586868286
it: 3400, train recon loss: -0.8220838904380798, local kl: 0.0 global kl: 1.8651191702190317e-12 valid reconstr loss: 0.6170347332954407
it: 3500, train recon loss: -0.7780019044876099, local kl: 0.0 global kl: 2.2598058840861412e-11 valid reconstr loss: 0.21797089278697968
it: 3600, train recon loss: -0.9829217791557312, local kl: 0.0 global kl: 3.7573867628371005e-12 valid reconstr loss: 2.960824966430664
it: 3700, train recon loss: -1.181379795074463, local kl: 0.0 global kl: 2.8154284459347423e-12 valid reconstr loss: -0.7627430558204651
Saving best model with reconstruction loss -0.76274306
it: 3800, train recon loss: -0.8583449125289917, local kl: 0.0 global kl: 1.0245831960631335e-12 valid reconstr loss: -0.32975178956985474
it: 3900, train recon loss: -1.16036856174469, local kl: 0.0 global kl: 1.621272560647924e-14 valid reconstr loss: 0.40170222520828247
it: 4000, train recon loss: -0.9379642605781555, local kl: 0.0 global kl: 1.723968190425751e-14 valid reconstr loss: -0.9118205904960632
Saving best model with reconstruction loss -0.9118206
it: 4100, train recon loss: -0.7912328839302063, local kl: 0.0 global kl: 3.208113028702053e-12 valid reconstr loss: -0.8855897784233093
it: 4200, train recon loss: -0.7660507559776306, local kl: 0.0 global kl: 1.8412216196139752e-12 valid reconstr loss: -0.7847574353218079
it: 4300, train recon loss: -1.1644260883331299, local kl: 0.0 global kl: 2.8663460494016135e-12 valid reconstr loss: -0.8823724389076233
it: 4400, train recon loss: -0.8628166317939758, local kl: 0.0 global kl: 7.741585150711217e-12 valid reconstr loss: -0.5526741743087769
it: 4500, train recon loss: -0.8229710459709167, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.9934677481651306
Saving best model with reconstruction loss -0.99346775
it: 4600, train recon loss: -1.079738736152649, local kl: 0.0 global kl: 8.050504707313166e-14 valid reconstr loss: -0.9291249513626099
it: 4700, train recon loss: -1.0274271965026855, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.24870461225509644
it: 4800, train recon loss: 4094879.25, local kl: 0.0 global kl: 7.037125222819629e-12 valid reconstr loss: -0.8560770750045776
it: 4900, train recon loss: -1.2751610279083252, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.11688757687807083
it: 5000, train recon loss: -1.0458370447158813, local kl: 0.0 global kl: 6.417245207446243e-12 valid reconstr loss: -0.7814134359359741
it: 5100, train recon loss: -0.9617224931716919, local kl: 0.0 global kl: 7.113428769633945e-14 valid reconstr loss: 5.563917636871338
it: 5200, train recon loss: -1.0733327865600586, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.7165655493736267
it: 5300, train recon loss: -0.8794255256652832, local kl: 0.0 global kl: 2.9157801173518294e-11 valid reconstr loss: -0.9750270843505859
it: 5400, train recon loss: 0.22205308079719543, local kl: 0.0 global kl: 1.7852164191367592e-11 valid reconstr loss: -0.8534848690032959
it: 5500, train recon loss: -1.2065569162368774, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.008034586906433
Saving best model with reconstruction loss -1.0080346
it: 5600, train recon loss: -1.2117395401000977, local kl: 0.0 global kl: 3.752506985699178e-12 valid reconstr loss: -0.8079783320426941
it: 5700, train recon loss: -0.368334025144577, local kl: 0.0 global kl: 3.7679304121240875e-12 valid reconstr loss: -1.0201152563095093
Saving best model with reconstruction loss -1.0201153
it: 5800, train recon loss: -1.0757321119308472, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.0682153701782227
Saving best model with reconstruction loss -1.0682154
it: 5900, train recon loss: -1.1187373399734497, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.1443145275115967
Saving best model with reconstruction loss -1.1443145
it: 6000, train recon loss: -1.2979092597961426, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.906826376914978
it: 6100, train recon loss: 15.598413467407227, local kl: 0.0 global kl: 0.0 valid reconstr loss: 5.1666107177734375
it: 6200, train recon loss: -1.0980640649795532, local kl: 0.0 global kl: 0.0 valid reconstr loss: 242.62449645996094
it: 6300, train recon loss: -0.9519946575164795, local kl: 0.0 global kl: 4.4027281820291364e-12 valid reconstr loss: -0.9605701565742493
it: 6400, train recon loss: -1.2257535457611084, local kl: 0.0 global kl: 3.4559410454909623e-12 valid reconstr loss: -1.0005018711090088
it: 6500, train recon loss: -1.0971125364303589, local kl: 0.0 global kl: 4.8530623963927155e-14 valid reconstr loss: 4.605196475982666
it: 6600, train recon loss: -0.5996896028518677, local kl: 0.0 global kl: 4.784145302139109e-12 valid reconstr loss: -0.9669707417488098
it: 6700, train recon loss: 2937.996826171875, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.2585750222206116
it: 6800, train recon loss: -1.1828150749206543, local kl: 0.0 global kl: 2.1803586713886602e-11 valid reconstr loss: -0.8772443532943726
it: 6900, train recon loss: -0.9206933379173279, local kl: 0.0 global kl: 9.81704748417335e-15 valid reconstr loss: 136.3768310546875
it: 7000, train recon loss: -0.834149181842804, local kl: 0.0 global kl: 3.109800286206499e-13 valid reconstr loss: -0.8758039474487305
it: 7100, train recon loss: -0.6839855909347534, local kl: 0.0 global kl: 1.3727314098800625e-14 valid reconstr loss: 16.053930282592773
it: 7200, train recon loss: -1.0483171939849854, local kl: 0.0 global kl: 1.1407364636228934e-11 valid reconstr loss: -0.8687145709991455
it: 7300, train recon loss: 13.571636199951172, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.0321396589279175
it: 7400, train recon loss: -0.8291754722595215, local kl: 0.0 global kl: 9.424575703187443e-12 valid reconstr loss: -1.1524255275726318
Saving best model with reconstruction loss -1.1524255
it: 7500, train recon loss: -1.1776882410049438, local kl: 0.0 global kl: 3.974988740940155e-13 valid reconstr loss: 20.282791137695312
it: 7600, train recon loss: -1.1182591915130615, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.988357424736023
it: 7700, train recon loss: -0.9750130772590637, local kl: 0.0 global kl: 4.6480397655757066e-14 valid reconstr loss: -0.4735536575317383
it: 7800, train recon loss: -1.1634362936019897, local kl: 0.0 global kl: 0.0 valid reconstr loss: 69.36351776123047
it: 7900, train recon loss: -1.142237901687622, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.9848895072937012
it: 8000, train recon loss: -1.3986676931381226, local kl: 0.0 global kl: 3.92602617083071e-14 valid reconstr loss: 44.72579574584961
it: 8100, train recon loss: -1.3133667707443237, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.8027783036231995
it: 8200, train recon loss: -1.3679581880569458, local kl: 0.0 global kl: 3.66166905824139e-12 valid reconstr loss: -1.2017546892166138
Saving best model with reconstruction loss -1.2017547
it: 8300, train recon loss: -1.182869791984558, local kl: 0.0 global kl: 4.091171845743702e-14 valid reconstr loss: -0.7951279282569885
it: 8400, train recon loss: -0.9878972172737122, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.096455454826355
it: 8500, train recon loss: -1.1188571453094482, local kl: 0.0 global kl: 1.4816141071178138e-14 valid reconstr loss: -1.058915376663208
it: 8600, train recon loss: 6.504535675048828, local kl: 0.0 global kl: 0.0 valid reconstr loss: 14.441956520080566
it: 8700, train recon loss: -0.9529950022697449, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.1767038106918335
it: 8800, train recon loss: -1.1259461641311646, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.063381552696228
it: 8900, train recon loss: -1.2984380722045898, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.0372563600540161
it: 9000, train recon loss: -1.3827130794525146, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.0890671014785767
it: 9100, train recon loss: -0.9686031341552734, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.0529423952102661
it: 9200, train recon loss: -1.104433536529541, local kl: 0.0 global kl: 0.0 valid reconstr loss: 279.2958984375
it: 9300, train recon loss: -0.851521372795105, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.0380789041519165
it: 9400, train recon loss: -1.0932159423828125, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.6525730490684509
it: 9500, train recon loss: -0.8251216411590576, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.9599425196647644
it: 9600, train recon loss: -0.5989768505096436, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.3005915880203247
Saving best model with reconstruction loss -1.3005916
it: 9700, train recon loss: -1.0496872663497925, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.1127756834030151
it: 9800, train recon loss: -1.023743987083435, local kl: 0.0 global kl: 0.0 valid reconstr loss: 9.595807075500488
it: 9900, train recon loss: 4.271275997161865, local kl: 0.0 global kl: 0.0 valid reconstr loss: 23.623085021972656
beta 0.5 temperature 0.5
it: 0, train recon loss: 2224.67333984375, local kl: 0.0 global kl: 0.005422009155154228 valid reconstr loss: 582.6435546875
Saving best model with reconstruction loss 582.64355
it: 100, train recon loss: 3.471472978591919, local kl: 0.0 global kl: 0.0010869416873902082 valid reconstr loss: 3.5405566692352295
Saving best model with reconstruction loss 3.5405567
it: 200, train recon loss: 3.1964945793151855, local kl: 0.0 global kl: 0.0011802369263023138 valid reconstr loss: 3.069071054458618
Saving best model with reconstruction loss 3.069071
it: 300, train recon loss: 2.3908116817474365, local kl: 0.0 global kl: 0.00022323112352751195 valid reconstr loss: 2.5199217796325684
Saving best model with reconstruction loss 2.5199218
it: 400, train recon loss: 1.9926563501358032, local kl: 0.0 global kl: 0.00028063650825060904 valid reconstr loss: 2.9629008769989014
it: 500, train recon loss: 1.9780651330947876, local kl: 0.0 global kl: 0.00015606096712872386 valid reconstr loss: 1.9074785709381104
Saving best model with reconstruction loss 1.9074786
it: 600, train recon loss: 1.741161823272705, local kl: 0.0 global kl: 0.0012459170538932085 valid reconstr loss: 2.0692968368530273
it: 700, train recon loss: 4.958874225616455, local kl: 0.0 global kl: 3.353039937792346e-05 valid reconstr loss: 6.909205913543701
it: 800, train recon loss: 2.041393756866455, local kl: 0.0 global kl: 0.00011801987420767546 valid reconstr loss: 3.7814621925354004
it: 900, train recon loss: 0.7471837401390076, local kl: 0.0 global kl: 8.010215242393315e-05 valid reconstr loss: 0.5921878218650818
Saving best model with reconstruction loss 0.5921878
it: 1000, train recon loss: 0.6960228085517883, local kl: 0.0 global kl: 0.0006022566230967641 valid reconstr loss: 0.5022331476211548
Saving best model with reconstruction loss 0.50223315
it: 1100, train recon loss: 3.3473103046417236, local kl: 0.0 global kl: 0.000171552412211895 valid reconstr loss: 18.512434005737305
it: 1200, train recon loss: 0.14592918753623962, local kl: 0.0 global kl: 0.0002738407638389617 valid reconstr loss: 0.09659935534000397
Saving best model with reconstruction loss 0.096599355
it: 1300, train recon loss: 40924.76953125, local kl: 0.0 global kl: 7.86530872574076e-05 valid reconstr loss: 58.603309631347656
it: 1400, train recon loss: 3.769108295440674, local kl: 0.0 global kl: 0.00014452690083999187 valid reconstr loss: 3.875965118408203
it: 1500, train recon loss: 0.8969539999961853, local kl: 0.0 global kl: 0.00018635101150721312 valid reconstr loss: 1.1607375144958496
it: 1600, train recon loss: 0.010925937443971634, local kl: 0.0 global kl: 8.918616367736831e-05 valid reconstr loss: -0.21132685244083405
Saving best model with reconstruction loss -0.21132685
it: 1700, train recon loss: 0.197373628616333, local kl: 0.0 global kl: 0.0005247958470135927 valid reconstr loss: 0.3080838918685913
it: 1800, train recon loss: -0.13762053847312927, local kl: 0.0 global kl: 0.00046198387281037867 valid reconstr loss: 0.07054431736469269
it: 1900, train recon loss: 0.3681943714618683, local kl: 0.0 global kl: 7.416789594572037e-05 valid reconstr loss: 1.8172721862792969
it: 2000, train recon loss: 251.26516723632812, local kl: 0.0 global kl: 6.841337017249316e-05 valid reconstr loss: -0.011941762641072273
it: 2100, train recon loss: -0.13655178248882294, local kl: 0.0 global kl: 0.00011547184840310365 valid reconstr loss: 0.0034880666062235832
it: 2200, train recon loss: -0.4558835029602051, local kl: 0.0 global kl: 0.0008356195176020265 valid reconstr loss: -0.4375854730606079
Saving best model with reconstruction loss -0.43758547
it: 2300, train recon loss: -0.5260111689567566, local kl: 0.0 global kl: 9.657934424467385e-05 valid reconstr loss: 1.155539870262146
it: 2400, train recon loss: 0.19307588040828705, local kl: 0.0 global kl: 0.00011604843894019723 valid reconstr loss: -0.2721056640148163
it: 2500, train recon loss: 14.615463256835938, local kl: 0.0 global kl: 4.2407358705531806e-05 valid reconstr loss: 16727.015625
it: 2600, train recon loss: -0.7953683137893677, local kl: 0.0 global kl: 3.186420872225426e-05 valid reconstr loss: -0.540833592414856
Saving best model with reconstruction loss -0.5408336
it: 2700, train recon loss: -0.06179377809166908, local kl: 0.0 global kl: 0.0005122192669659853 valid reconstr loss: -0.09934622049331665
it: 2800, train recon loss: -0.8362953066825867, local kl: 0.0 global kl: 4.574724880512804e-05 valid reconstr loss: 9229.3876953125
it: 2900, train recon loss: -0.4941282570362091, local kl: 0.0 global kl: 0.00046494673006236553 valid reconstr loss: -0.641780436038971
Saving best model with reconstruction loss -0.64178044
it: 3000, train recon loss: -0.07255934923887253, local kl: 0.0 global kl: 4.7360175813082606e-05 valid reconstr loss: -0.7585498690605164
Saving best model with reconstruction loss -0.75854987
it: 3100, train recon loss: -0.3668455481529236, local kl: 0.0 global kl: 0.00037340610288083553 valid reconstr loss: -0.7010971307754517
it: 3200, train recon loss: -0.5445757508277893, local kl: 0.0 global kl: 0.00022729006013832986 valid reconstr loss: 1541.9967041015625
it: 3300, train recon loss: 1.83805251121521, local kl: 0.0 global kl: 9.410674829268828e-05 valid reconstr loss: -0.8111006021499634
Saving best model with reconstruction loss -0.8111006
it: 3400, train recon loss: -0.1251314878463745, local kl: 0.0 global kl: 4.4471169530879706e-05 valid reconstr loss: 3023.457763671875
it: 3500, train recon loss: -1.07277512550354, local kl: 0.0 global kl: 1.8677479602047242e-05 valid reconstr loss: -0.8129165172576904
Saving best model with reconstruction loss -0.8129165
it: 3600, train recon loss: -0.6863555312156677, local kl: 0.0 global kl: 4.896334576187655e-05 valid reconstr loss: 3.777026653289795
it: 3700, train recon loss: -1.2861906290054321, local kl: 0.0 global kl: 7.331396773224697e-05 valid reconstr loss: -0.7532785534858704
it: 3800, train recon loss: -0.4424067735671997, local kl: 0.0 global kl: 0.004550456535071135 valid reconstr loss: 5172.166015625
it: 3900, train recon loss: 1.2917144298553467, local kl: 0.0 global kl: 0.0001193004209198989 valid reconstr loss: 0.7906771302223206
it: 4000, train recon loss: -1.3383926153182983, local kl: 0.0 global kl: 1.4283732525655068e-05 valid reconstr loss: -0.5999814867973328
it: 4100, train recon loss: 86.33789825439453, local kl: 0.0 global kl: 0.00038938308716751635 valid reconstr loss: 950.2227172851562
it: 4200, train recon loss: -0.8896148204803467, local kl: 0.0 global kl: 0.00020616513211280107 valid reconstr loss: 42.52956008911133
it: 4300, train recon loss: -0.7475872039794922, local kl: 0.0 global kl: 0.00016682779823895544 valid reconstr loss: -0.7825934886932373
it: 4400, train recon loss: -1.0074199438095093, local kl: 0.0 global kl: 3.830057175946422e-05 valid reconstr loss: 12.571244239807129
it: 4500, train recon loss: -0.9969781041145325, local kl: 0.0 global kl: 0.00013838295126333833 valid reconstr loss: -1.103714108467102
Saving best model with reconstruction loss -1.1037141
it: 4600, train recon loss: -0.5557494163513184, local kl: 0.0 global kl: 7.254351658048108e-05 valid reconstr loss: -0.9024062156677246
it: 4700, train recon loss: -1.052284598350525, local kl: 0.0 global kl: 1.6797606804175302e-05 valid reconstr loss: -0.9092065095901489
it: 4800, train recon loss: -1.494708776473999, local kl: 0.0 global kl: 1.9361372324055992e-05 valid reconstr loss: -1.217478632926941
Saving best model with reconstruction loss -1.2174786
it: 4900, train recon loss: 118.85405731201172, local kl: 0.0 global kl: 3.827176624326967e-05 valid reconstr loss: -0.5020117163658142
it: 5000, train recon loss: -0.5556018948554993, local kl: 0.0 global kl: 0.00031974553712643683 valid reconstr loss: -0.8484842777252197
it: 5100, train recon loss: -1.0580748319625854, local kl: 0.0 global kl: 6.312159530352801e-05 valid reconstr loss: 2622.549560546875
it: 5200, train recon loss: -0.915024995803833, local kl: 0.0 global kl: 0.00032184558222070336 valid reconstr loss: -0.9169852137565613
it: 5300, train recon loss: -1.1095027923583984, local kl: 0.0 global kl: 6.399372068699449e-05 valid reconstr loss: -1.1612154245376587
it: 5400, train recon loss: -0.9799491167068481, local kl: 0.0 global kl: 0.00010446917440276593 valid reconstr loss: -1.1828036308288574
it: 5500, train recon loss: -1.2834904193878174, local kl: 0.0 global kl: 3.8365997170330957e-05 valid reconstr loss: -1.1532268524169922
it: 5600, train recon loss: -1.3256484270095825, local kl: 0.0 global kl: 0.000210475642234087 valid reconstr loss: 7765.44140625
it: 5700, train recon loss: -1.0602856874465942, local kl: 0.0 global kl: 4.7978297516237944e-05 valid reconstr loss: -1.1797077655792236
it: 5800, train recon loss: -1.2058842182159424, local kl: 0.0 global kl: 0.00013337790733203292 valid reconstr loss: -1.1947107315063477
it: 5900, train recon loss: -1.539300560951233, local kl: 0.0 global kl: 0.00010365719208493829 valid reconstr loss: -1.126944899559021
it: 6000, train recon loss: -0.3294139802455902, local kl: 0.0 global kl: 1.6439171304227784e-06 valid reconstr loss: 34.70487594604492
it: 6100, train recon loss: 3.280485153198242, local kl: 0.0 global kl: 6.492538523161784e-05 valid reconstr loss: 3.236551284790039
it: 6200, train recon loss: 3.063467264175415, local kl: 0.0 global kl: 1.4601947441406082e-05 valid reconstr loss: 3.1464035511016846
it: 6300, train recon loss: 3.6417343616485596, local kl: 0.0 global kl: 1.6072835506975025e-08 valid reconstr loss: 3.065359115600586
it: 6400, train recon loss: 3.1207425594329834, local kl: 0.0 global kl: 2.102861706987369e-08 valid reconstr loss: 3.085378646850586
it: 6500, train recon loss: 3.253887891769409, local kl: 0.0 global kl: 1.173871444493102e-09 valid reconstr loss: 3.0386524200439453
it: 6600, train recon loss: 2.714986562728882, local kl: 0.0 global kl: 2.6054323143398506e-09 valid reconstr loss: 2.9510419368743896
it: 6700, train recon loss: 3.1225404739379883, local kl: 0.0 global kl: 1.5069673287015917e-09 valid reconstr loss: 3.0254154205322266
it: 6800, train recon loss: 3.0052881240844727, local kl: 0.0 global kl: 1.026177920415705e-09 valid reconstr loss: 3.0738229751586914
it: 6900, train recon loss: 3.008406639099121, local kl: 0.0 global kl: 5.77027592374435e-10 valid reconstr loss: 3.058900833129883
it: 7000, train recon loss: 3.0780491828918457, local kl: 0.0 global kl: 5.716396800359291e-10 valid reconstr loss: 2.999257802963257
it: 7100, train recon loss: 3.137413263320923, local kl: 0.0 global kl: 1.985775294199854e-10 valid reconstr loss: 2.97904634475708
it: 7200, train recon loss: 2.9617552757263184, local kl: 0.0 global kl: 1.5251951090977656e-10 valid reconstr loss: 2.9923994541168213
it: 7300, train recon loss: 3.236116886138916, local kl: 0.0 global kl: 1.587036752015436e-10 valid reconstr loss: 2.9661004543304443
it: 7400, train recon loss: 3.2073545455932617, local kl: 0.0 global kl: 1.2758094580789248e-10 valid reconstr loss: 3.0076510906219482
it: 7500, train recon loss: 2.9972825050354004, local kl: 0.0 global kl: 9.470267625655282e-11 valid reconstr loss: 2.995589256286621
it: 7600, train recon loss: 3.0539798736572266, local kl: 0.0 global kl: 3.1914783316100426e-11 valid reconstr loss: 2.9601027965545654
it: 7700, train recon loss: 2.797006130218506, local kl: 0.0 global kl: 5.247264775465332e-11 valid reconstr loss: 2.963338851928711
it: 7800, train recon loss: 2.9148364067077637, local kl: 0.0 global kl: 1.681539976705615e-11 valid reconstr loss: 2.9664087295532227
it: 7900, train recon loss: 2.7877607345581055, local kl: 0.0 global kl: 9.120317348565443e-12 valid reconstr loss: 2.9476475715637207
it: 8000, train recon loss: 2.89556884765625, local kl: 0.0 global kl: 1.4115428444150258e-11 valid reconstr loss: 2.9579882621765137
it: 8100, train recon loss: 2.900176763534546, local kl: 0.0 global kl: 4.410154966910662e-12 valid reconstr loss: 2.9572646617889404
it: 8200, train recon loss: 2.9094724655151367, local kl: 0.0 global kl: 5.87765816920105e-12 valid reconstr loss: 2.9557950496673584
it: 8300, train recon loss: 2.94905686378479, local kl: 0.0 global kl: 8.45982311981075e-12 valid reconstr loss: 2.917334794998169
it: 8400, train recon loss: 2.7650599479675293, local kl: 0.0 global kl: 2.4966701882983555e-12 valid reconstr loss: 2.9645588397979736
it: 8500, train recon loss: 3.0383903980255127, local kl: 0.0 global kl: 5.369552892597884e-12 valid reconstr loss: 2.978610038757324
it: 8600, train recon loss: 3.224085807800293, local kl: 0.0 global kl: 2.8821574033638386e-12 valid reconstr loss: 2.948763370513916
it: 8700, train recon loss: 2.8645801544189453, local kl: 0.0 global kl: 1.4681490615245374e-12 valid reconstr loss: 2.926130771636963
it: 8800, train recon loss: 2.688215732574463, local kl: 0.0 global kl: 1.1648132545311052e-12 valid reconstr loss: 3.0136120319366455
it: 8900, train recon loss: 2.7431230545043945, local kl: 0.0 global kl: 1.28714042462591e-12 valid reconstr loss: 2.9794042110443115
it: 9000, train recon loss: 2.7170610427856445, local kl: 0.0 global kl: 9.919577095493515e-13 valid reconstr loss: 2.9403791427612305
it: 9100, train recon loss: 3.079477071762085, local kl: 0.0 global kl: 1.561686327131162e-12 valid reconstr loss: 2.9459774494171143
it: 9200, train recon loss: 2.880915880203247, local kl: 0.0 global kl: 6.764441437545621e-13 valid reconstr loss: 2.9485256671905518
it: 9300, train recon loss: 2.9148380756378174, local kl: 0.0 global kl: 6.703245272222935e-13 valid reconstr loss: 2.8962819576263428
it: 9400, train recon loss: 2.955155611038208, local kl: 0.0 global kl: 1.2481866668717645e-12 valid reconstr loss: 2.8803513050079346
it: 9500, train recon loss: 2.9963536262512207, local kl: 0.0 global kl: 7.492041926085435e-13 valid reconstr loss: 2.950434684753418
it: 9600, train recon loss: 2.8668012619018555, local kl: 0.0 global kl: 8.264410748630435e-13 valid reconstr loss: 2.916525363922119
it: 9700, train recon loss: 2.9407873153686523, local kl: 0.0 global kl: 6.160166777721687e-13 valid reconstr loss: 2.9472904205322266
it: 9800, train recon loss: 3.0258066654205322, local kl: 0.0 global kl: 1.128103036332484e-12 valid reconstr loss: 2.9541802406311035
it: 9900, train recon loss: 2.671281099319458, local kl: 0.0 global kl: 1.1479044711298902e-12 valid reconstr loss: 2.9453470706939697
beta 0.5 temperature 1.0
it: 0, train recon loss: 14200.1396484375, local kl: 0.0 global kl: 0.0032195746898651123 valid reconstr loss: 3708.628173828125
Saving best model with reconstruction loss 3708.6282
it: 100, train recon loss: 3.72330641746521, local kl: 0.0 global kl: 0.0011039914097636938 valid reconstr loss: 4.213127613067627
Saving best model with reconstruction loss 4.2131276
it: 200, train recon loss: 3.6106903553009033, local kl: 0.0 global kl: 6.639982893830165e-05 valid reconstr loss: 15841095.0
it: 300, train recon loss: 3.1168782711029053, local kl: 0.0 global kl: 0.0001267190818907693 valid reconstr loss: 4.978559494018555
it: 400, train recon loss: 2.4858057498931885, local kl: 0.0 global kl: 7.951963925734162e-05 valid reconstr loss: 2.385699987411499
Saving best model with reconstruction loss 2.3857
it: 500, train recon loss: 2.3855621814727783, local kl: 0.0 global kl: 1.5277493730536662e-05 valid reconstr loss: 2.1151225566864014
Saving best model with reconstruction loss 2.1151226
it: 600, train recon loss: 2.2259867191314697, local kl: 0.0 global kl: 5.020999378757551e-05 valid reconstr loss: 2.1627895832061768
it: 700, train recon loss: 1.216884732246399, local kl: 0.0 global kl: 0.0004923687083646655 valid reconstr loss: 5.831252574920654
it: 800, train recon loss: 1.4224786758422852, local kl: 0.0 global kl: 0.00014376504987012595 valid reconstr loss: 1.9288512468338013
Saving best model with reconstruction loss 1.9288512
it: 900, train recon loss: 1.9411334991455078, local kl: 0.0 global kl: 0.0001206729793921113 valid reconstr loss: 1.4042447805404663
Saving best model with reconstruction loss 1.4042448
it: 1000, train recon loss: 1.8456262350082397, local kl: 0.0 global kl: 3.96781470044516e-05 valid reconstr loss: 1.4628838300704956
it: 1100, train recon loss: 1.0535589456558228, local kl: 0.0 global kl: 2.5963017833419144e-05 valid reconstr loss: 0.9721038937568665
Saving best model with reconstruction loss 0.9721039
it: 1200, train recon loss: 0.9636987447738647, local kl: 0.0 global kl: 0.00020974311337340623 valid reconstr loss: 2.9852068424224854
it: 1300, train recon loss: 1.1830060482025146, local kl: 0.0 global kl: 0.0002002163528231904 valid reconstr loss: 577.5053100585938
it: 1400, train recon loss: 0.8865138292312622, local kl: 0.0 global kl: 0.0005283377831801772 valid reconstr loss: 0.8561415076255798
Saving best model with reconstruction loss 0.8561415
it: 1500, train recon loss: 1.2121292352676392, local kl: 0.0 global kl: 0.00016697755199857056 valid reconstr loss: 0.8888267874717712
it: 1600, train recon loss: 0.8351790904998779, local kl: 0.0 global kl: 1.8579705283627845e-05 valid reconstr loss: 1.2856252193450928
it: 1700, train recon loss: 1.13729989528656, local kl: 0.0 global kl: 4.7431778511963785e-05 valid reconstr loss: 0.8658641576766968
it: 1800, train recon loss: 0.9919235706329346, local kl: 0.0 global kl: 6.967315130168572e-05 valid reconstr loss: 0.8080043196678162
Saving best model with reconstruction loss 0.8080043
it: 1900, train recon loss: 0.5719459056854248, local kl: 0.0 global kl: 3.7740810512332246e-05 valid reconstr loss: 0.7093366384506226
Saving best model with reconstruction loss 0.70933664
it: 2000, train recon loss: 1.135062575340271, local kl: 0.0 global kl: 5.9708771004807204e-05 valid reconstr loss: 1.0533102750778198
it: 2100, train recon loss: 0.6199114322662354, local kl: 0.0 global kl: 5.810119091620436e-06 valid reconstr loss: 0.8579612374305725
it: 2200, train recon loss: 0.46819013357162476, local kl: 0.0 global kl: 7.240269042085856e-05 valid reconstr loss: 1.027410864830017
it: 2300, train recon loss: 1.6129652261734009, local kl: 0.0 global kl: 0.001364941825158894 valid reconstr loss: 155.35203552246094
it: 2400, train recon loss: 0.6996219158172607, local kl: 0.0 global kl: 4.060842911712825e-05 valid reconstr loss: 1.4470672607421875
it: 2500, train recon loss: 14.426251411437988, local kl: 0.0 global kl: 3.185380410286598e-05 valid reconstr loss: 0.6399458646774292
Saving best model with reconstruction loss 0.63994586
it: 2600, train recon loss: 0.6513916254043579, local kl: 0.0 global kl: 2.07153116207337e-05 valid reconstr loss: 0.37242498993873596
Saving best model with reconstruction loss 0.372425
it: 2700, train recon loss: 1.2404913902282715, local kl: 0.0 global kl: 0.00013004188076592982 valid reconstr loss: 0.6796057224273682
it: 2800, train recon loss: 0.8134826421737671, local kl: 0.0 global kl: 2.132416921085678e-05 valid reconstr loss: 0.5654997825622559
it: 2900, train recon loss: 0.4149126708507538, local kl: 0.0 global kl: 9.256077464669943e-05 valid reconstr loss: 0.4820074439048767
it: 3000, train recon loss: 0.4994181990623474, local kl: 0.0 global kl: 2.2498063117382117e-05 valid reconstr loss: 0.6201460957527161
it: 3100, train recon loss: -0.04091451317071915, local kl: 0.0 global kl: 1.8354943676968105e-05 valid reconstr loss: 0.2542131245136261
Saving best model with reconstruction loss 0.25421312
it: 3200, train recon loss: 0.8726164698600769, local kl: 0.0 global kl: 8.400610386161134e-05 valid reconstr loss: 0.3489379286766052
it: 3300, train recon loss: 0.5117580890655518, local kl: 0.0 global kl: 8.314885781146586e-05 valid reconstr loss: 0.37315189838409424
it: 3400, train recon loss: 0.5032445192337036, local kl: 0.0 global kl: 6.761444819858298e-05 valid reconstr loss: 0.44970062375068665
it: 3500, train recon loss: 3.2783071994781494, local kl: 0.0 global kl: 8.54784666444175e-05 valid reconstr loss: 2.6422629356384277
it: 3600, train recon loss: 3.222046136856079, local kl: 0.0 global kl: 0.0001503094390500337 valid reconstr loss: 2.6617870330810547
it: 3700, train recon loss: 1.0796457529067993, local kl: 0.0 global kl: 0.00017349128029309213 valid reconstr loss: 1.3139169216156006
it: 3800, train recon loss: 1.182563066482544, local kl: 0.0 global kl: 7.457270112354308e-05 valid reconstr loss: 0.6671333909034729
it: 3900, train recon loss: 0.1788439005613327, local kl: 0.0 global kl: 7.40402247174643e-05 valid reconstr loss: 0.5868748426437378
it: 4000, train recon loss: 0.2386709302663803, local kl: 0.0 global kl: 2.780259819701314e-05 valid reconstr loss: 0.3640163838863373
it: 4100, train recon loss: 6.870812892913818, local kl: 0.0 global kl: 0.0003284444974269718 valid reconstr loss: 2.656252145767212
it: 4200, train recon loss: 1.6031243801116943, local kl: 0.0 global kl: 4.340176747064106e-05 valid reconstr loss: 1.0329034328460693
it: 4300, train recon loss: 3.2694790363311768, local kl: 0.0 global kl: 6.949782255105674e-05 valid reconstr loss: 1.075859785079956
it: 4400, train recon loss: 0.7739877104759216, local kl: 0.0 global kl: 3.2381627534050494e-05 valid reconstr loss: 0.5070226192474365
it: 4500, train recon loss: 0.6004735827445984, local kl: 0.0 global kl: 0.0003634525346569717 valid reconstr loss: 0.9774641394615173
it: 4600, train recon loss: 0.3450879454612732, local kl: 0.0 global kl: 6.772591405024286e-06 valid reconstr loss: 0.1323801875114441
Saving best model with reconstruction loss 0.13238019
it: 4700, train recon loss: 0.4501029849052429, local kl: 0.0 global kl: 8.494002031511627e-06 valid reconstr loss: 0.36762845516204834
it: 4800, train recon loss: 0.1355149894952774, local kl: 0.0 global kl: 9.802552995097358e-06 valid reconstr loss: 0.24418866634368896
it: 4900, train recon loss: 0.1897725760936737, local kl: 0.0 global kl: 0.00018757773796096444 valid reconstr loss: 0.5503507852554321
it: 5000, train recon loss: 0.4369807839393616, local kl: 0.0 global kl: 5.53134195797611e-05 valid reconstr loss: 0.3459164500236511
it: 5100, train recon loss: 0.28701305389404297, local kl: 0.0 global kl: 0.0003644131065811962 valid reconstr loss: 0.5734808444976807
it: 5200, train recon loss: 0.2185474932193756, local kl: 0.0 global kl: 0.00018606730736792088 valid reconstr loss: 0.11856392025947571
Saving best model with reconstruction loss 0.11856392
it: 5300, train recon loss: 0.3237029016017914, local kl: 0.0 global kl: 4.753952816827223e-05 valid reconstr loss: 0.12597988545894623
it: 5400, train recon loss: 0.28543150424957275, local kl: 0.0 global kl: 0.00012780216638930142 valid reconstr loss: 0.11529389768838882
Saving best model with reconstruction loss 0.1152939
it: 5500, train recon loss: 0.15618175268173218, local kl: 0.0 global kl: 1.1441735296102706e-05 valid reconstr loss: 0.3515111804008484
it: 5600, train recon loss: 0.24980440735816956, local kl: 0.0 global kl: 3.7759000406367704e-06 valid reconstr loss: 0.17627227306365967
it: 5700, train recon loss: 0.15366384387016296, local kl: 0.0 global kl: 3.5643151932163164e-05 valid reconstr loss: 0.4653524160385132
it: 5800, train recon loss: 0.1296602487564087, local kl: 0.0 global kl: 7.263095176313072e-05 valid reconstr loss: 0.3429887592792511
it: 5900, train recon loss: 0.10177643597126007, local kl: 0.0 global kl: 0.00011243864719290286 valid reconstr loss: 0.0893988236784935
Saving best model with reconstruction loss 0.08939882
it: 6000, train recon loss: 0.25957563519477844, local kl: 0.0 global kl: 2.4684248273842968e-05 valid reconstr loss: 4.156861305236816
it: 6100, train recon loss: 3.9159610271453857, local kl: 0.0 global kl: 3.7935678847134113e-06 valid reconstr loss: 4.086606025695801
it: 6200, train recon loss: 1.7696439027786255, local kl: 0.0 global kl: 3.8078244415373774e-06 valid reconstr loss: 2.315690755844116
it: 6300, train recon loss: 1.4363088607788086, local kl: 0.0 global kl: 4.24828658651677e-06 valid reconstr loss: 0.958664059638977
it: 6400, train recon loss: 1.0610270500183105, local kl: 0.0 global kl: 1.4815200302109588e-05 valid reconstr loss: 1.2162569761276245
it: 6500, train recon loss: 1.3083943128585815, local kl: 0.0 global kl: 0.0007965298136696219 valid reconstr loss: 1.5512504577636719
it: 6600, train recon loss: 0.5497885346412659, local kl: 0.0 global kl: 1.377978492200782e-06 valid reconstr loss: 0.5859917402267456
it: 6700, train recon loss: 0.7344800233840942, local kl: 0.0 global kl: 8.295734005514532e-06 valid reconstr loss: 0.41547414660453796
it: 6800, train recon loss: 1.6508234739303589, local kl: 0.0 global kl: 0.0004651937633752823 valid reconstr loss: 0.41185253858566284
it: 6900, train recon loss: 0.8211336135864258, local kl: 0.0 global kl: 2.182222669944167e-05 valid reconstr loss: 0.88615882396698
it: 7000, train recon loss: 0.09223759174346924, local kl: 0.0 global kl: 2.300989581272006e-05 valid reconstr loss: 0.26828789710998535
it: 7100, train recon loss: 0.10771725326776505, local kl: 0.0 global kl: 6.633168231928721e-05 valid reconstr loss: 0.11948129534721375
it: 7200, train recon loss: 0.6417955160140991, local kl: 0.0 global kl: 7.072816515574232e-06 valid reconstr loss: 0.3841506838798523
it: 7300, train recon loss: 0.25024616718292236, local kl: 0.0 global kl: 1.7265414498979226e-05 valid reconstr loss: 4397.6591796875
it: 7400, train recon loss: 0.8080999851226807, local kl: 0.0 global kl: 3.79855337087065e-05 valid reconstr loss: 0.0772835984826088
Saving best model with reconstruction loss 0.0772836
it: 7500, train recon loss: -0.09495668113231659, local kl: 0.0 global kl: 1.3964806385047268e-05 valid reconstr loss: 2.735177993774414
it: 7600, train recon loss: 1144.4871826171875, local kl: 0.0 global kl: 1.1042089681723155e-05 valid reconstr loss: 44292.0859375
it: 7700, train recon loss: 0.16145947575569153, local kl: 0.0 global kl: 1.34105755478231e-06 valid reconstr loss: 0.12783026695251465
it: 7800, train recon loss: 0.22755298018455505, local kl: 0.0 global kl: 2.93799462269817e-06 valid reconstr loss: 0.0350470133125782
Saving best model with reconstruction loss 0.035047013
it: 7900, train recon loss: 0.4415300190448761, local kl: 0.0 global kl: 5.592299930867739e-05 valid reconstr loss: -0.05051601305603981
Saving best model with reconstruction loss -0.050516013
it: 8000, train recon loss: 0.03426094725728035, local kl: 0.0 global kl: 4.051606993016321e-06 valid reconstr loss: 0.17141024768352509
it: 8100, train recon loss: 0.22191818058490753, local kl: 0.0 global kl: 8.042064791879966e-07 valid reconstr loss: 0.027984652668237686
it: 8200, train recon loss: 3323.290771484375, local kl: 0.0 global kl: 2.4770182790234685e-05 valid reconstr loss: 0.14393064379692078
it: 8300, train recon loss: 2909.2138671875, local kl: 0.0 global kl: 1.5972549590514973e-05 valid reconstr loss: 0.17341086268424988
it: 8400, train recon loss: 0.3252517879009247, local kl: 0.0 global kl: 1.6274780136882327e-06 valid reconstr loss: 0.08082092553377151
it: 8500, train recon loss: 0.8288635015487671, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.6227471828460693
it: 8600, train recon loss: 2.5047192573547363, local kl: 0.0 global kl: 0.00014649401418864727 valid reconstr loss: 2.4039368629455566
it: 8700, train recon loss: 0.8645339608192444, local kl: 0.0 global kl: 0.00010651765478542075 valid reconstr loss: 2.008237838745117
it: 8800, train recon loss: 0.5892065167427063, local kl: 0.0 global kl: 2.1505862605408765e-05 valid reconstr loss: 1.2054160833358765
it: 8900, train recon loss: 0.6588519811630249, local kl: 0.0 global kl: 0.0008720211917534471 valid reconstr loss: 0.8662136197090149
it: 9000, train recon loss: 232.3631591796875, local kl: 0.0 global kl: 0.005122426897287369 valid reconstr loss: 3.7743496894836426
it: 9100, train recon loss: 2.401848554611206, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.3285160064697266
it: 9200, train recon loss: 0.8396701216697693, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.9160388708114624
it: 9300, train recon loss: 2.5092055797576904, local kl: 0.0 global kl: 1.7875663615996018e-05 valid reconstr loss: 2.3517210483551025
it: 9400, train recon loss: 1.4837629795074463, local kl: 0.0 global kl: 6.273166218306869e-05 valid reconstr loss: 1.9715250730514526
it: 9500, train recon loss: 1.3353632688522339, local kl: 0.0 global kl: 2.2799134967499413e-05 valid reconstr loss: 1.0986382961273193
it: 9600, train recon loss: 1.8017014265060425, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.8071630001068115
it: 9700, train recon loss: 0.3401489853858948, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.652762770652771
it: 9800, train recon loss: 24544.677734375, local kl: 0.0 global kl: 6.080116145312786e-05 valid reconstr loss: 4680.2431640625
it: 9900, train recon loss: 1.6301759481430054, local kl: 0.0 global kl: 0.0005218586884438992 valid reconstr loss: 1.890445590019226
beta 0.5 temperature 2.0
it: 0, train recon loss: 412.3524475097656, local kl: 0.0 global kl: 0.008423915132880211 valid reconstr loss: 181.1405792236328
Saving best model with reconstruction loss 181.14058
it: 100, train recon loss: 3.949845790863037, local kl: 0.0 global kl: 0.0011924790451303124 valid reconstr loss: 4.6171417236328125
Saving best model with reconstruction loss 4.6171417
it: 200, train recon loss: 3.9930179119110107, local kl: 0.0 global kl: 3.957950684707612e-05 valid reconstr loss: 4.169563293457031
Saving best model with reconstruction loss 4.1695633
it: 300, train recon loss: 3.7324609756469727, local kl: 0.0 global kl: 9.43603299674578e-06 valid reconstr loss: 3.555501699447632
Saving best model with reconstruction loss 3.5555017
it: 400, train recon loss: 3.726012706756592, local kl: 0.0 global kl: 0.00019241969857830554 valid reconstr loss: 3.6772615909576416
it: 500, train recon loss: 2.321162462234497, local kl: 0.0 global kl: 0.00021088848006911576 valid reconstr loss: 2.307624101638794
Saving best model with reconstruction loss 2.307624
it: 600, train recon loss: 2.080692768096924, local kl: 0.0 global kl: 9.2022219178034e-06 valid reconstr loss: 2.0532562732696533
Saving best model with reconstruction loss 2.0532563
it: 700, train recon loss: 1.0264325141906738, local kl: 0.0 global kl: 1.5807910536125291e-12 valid reconstr loss: 0.9742600917816162
Saving best model with reconstruction loss 0.9742601
it: 800, train recon loss: 1.1190720796585083, local kl: 0.0 global kl: 1.68365321684405e-13 valid reconstr loss: 2.6438541412353516
it: 900, train recon loss: 0.7563635110855103, local kl: 0.0 global kl: 5.118377943702512e-12 valid reconstr loss: 1.812031865119934
it: 1000, train recon loss: 0.5424271821975708, local kl: 0.0 global kl: 1.348261779998694e-13 valid reconstr loss: 0.6982294917106628
Saving best model with reconstruction loss 0.6982295
it: 1100, train recon loss: 0.31055566668510437, local kl: 0.0 global kl: 1.2050091827142673e-12 valid reconstr loss: 0.7972965240478516
it: 1200, train recon loss: 0.8554630279541016, local kl: 0.0 global kl: 5.4012350148013866e-14 valid reconstr loss: 0.6269694566726685
Saving best model with reconstruction loss 0.62696946
it: 1300, train recon loss: 0.3425009846687317, local kl: 0.0 global kl: 2.087702406783354e-12 valid reconstr loss: 0.24507512152194977
Saving best model with reconstruction loss 0.24507512
it: 1400, train recon loss: -0.33553314208984375, local kl: 0.0 global kl: 3.529752878872472e-12 valid reconstr loss: -0.0991063043475151
Saving best model with reconstruction loss -0.099106304
it: 1500, train recon loss: -0.1371595561504364, local kl: 0.0 global kl: 2.2908792918219234e-12 valid reconstr loss: 0.15575304627418518
it: 1600, train recon loss: 0.08267179131507874, local kl: 0.0 global kl: 2.0848323067923502e-12 valid reconstr loss: 154.02845764160156
it: 1700, train recon loss: -0.11590214818716049, local kl: 0.0 global kl: 4.4478249208723675e-12 valid reconstr loss: 0.5350891351699829
it: 1800, train recon loss: -0.14544238150119781, local kl: 0.0 global kl: 1.2926779872218797e-13 valid reconstr loss: 0.45816555619239807
it: 1900, train recon loss: -0.3109581768512726, local kl: 0.0 global kl: 2.1978252551235755e-13 valid reconstr loss: -0.19245250523090363
Saving best model with reconstruction loss -0.1924525
it: 2000, train recon loss: 2734.6572265625, local kl: 0.0 global kl: 6.555866960411549e-13 valid reconstr loss: -0.1729348748922348
it: 2100, train recon loss: -0.29702216386795044, local kl: 0.0 global kl: 1.7723635059585519e-12 valid reconstr loss: -0.19423384964466095
Saving best model with reconstruction loss -0.19423385
it: 2200, train recon loss: 0.3926234543323517, local kl: 0.0 global kl: 4.885637033824608e-12 valid reconstr loss: 0.05056176707148552
it: 2300, train recon loss: -0.01725107617676258, local kl: 0.0 global kl: 1.5575062940753615e-13 valid reconstr loss: 0.025847138836979866
it: 2400, train recon loss: 221.2646942138672, local kl: 0.0 global kl: 2.3525660586276587e-12 valid reconstr loss: 4.8735785484313965
it: 2500, train recon loss: 0.19816549122333527, local kl: 0.0 global kl: 3.0243949705743844e-13 valid reconstr loss: -0.403827965259552
Saving best model with reconstruction loss -0.40382797
it: 2600, train recon loss: -0.7687643766403198, local kl: 0.0 global kl: 3.1967263038057414e-12 valid reconstr loss: -0.03882840648293495
it: 2700, train recon loss: 0.10467584431171417, local kl: 0.0 global kl: 1.0643708137081376e-12 valid reconstr loss: -0.5154905915260315
Saving best model with reconstruction loss -0.5154906
it: 2800, train recon loss: -0.5511628985404968, local kl: 0.0 global kl: 2.528244974486782e-12 valid reconstr loss: 5.0968756675720215
it: 2900, train recon loss: -0.5302512645721436, local kl: 0.0 global kl: 7.0985023414849024e-12 valid reconstr loss: 0.23332376778125763
it: 3000, train recon loss: -0.586376428604126, local kl: 0.0 global kl: 6.038058941726376e-12 valid reconstr loss: -0.5724685192108154
Saving best model with reconstruction loss -0.5724685
it: 3100, train recon loss: -0.8795145153999329, local kl: 0.0 global kl: 1.9546864127306662e-14 valid reconstr loss: -0.5154089331626892
it: 3200, train recon loss: -0.5404680371284485, local kl: 0.0 global kl: 3.0309088572266774e-12 valid reconstr loss: -0.6253629922866821
Saving best model with reconstruction loss -0.625363
it: 3300, train recon loss: -0.7302204966545105, local kl: 0.0 global kl: 1.5579664294773643e-12 valid reconstr loss: 842.3748168945312
it: 3400, train recon loss: -0.8191116452217102, local kl: 0.0 global kl: 2.3268643956075863e-12 valid reconstr loss: 5.262365818023682
it: 3500, train recon loss: -1.0054678916931152, local kl: 0.0 global kl: 2.715605518233133e-13 valid reconstr loss: 106.92742156982422
it: 3600, train recon loss: -0.3010183572769165, local kl: 0.0 global kl: 3.528857761558868e-12 valid reconstr loss: -0.6840701699256897
Saving best model with reconstruction loss -0.68407017
it: 3700, train recon loss: -0.6194821000099182, local kl: 0.0 global kl: 4.939546167925801e-12 valid reconstr loss: -0.6235607862472534
it: 3800, train recon loss: -0.9723273515701294, local kl: 0.0 global kl: 2.0611509027324004e-11 valid reconstr loss: 3582.730712890625
it: 3900, train recon loss: -0.9739328026771545, local kl: 0.0 global kl: 2.9601605455037427e-12 valid reconstr loss: 92.55559539794922
it: 4000, train recon loss: -1.1599663496017456, local kl: 0.0 global kl: 1.7818645864364768e-14 valid reconstr loss: 0.12539803981781006
it: 4100, train recon loss: -1.1053965091705322, local kl: 0.0 global kl: 1.3777772325807014e-12 valid reconstr loss: -0.7285417914390564
Saving best model with reconstruction loss -0.7285418
it: 4200, train recon loss: -0.36682093143463135, local kl: 0.0 global kl: 2.5788399193871214e-14 valid reconstr loss: 148.05792236328125
it: 4300, train recon loss: 0.10648109018802643, local kl: 0.0 global kl: 2.0573820425084932e-15 valid reconstr loss: -0.13708558678627014
it: 4400, train recon loss: 10.200301170349121, local kl: 0.0 global kl: 1.2526148521208036e-11 valid reconstr loss: 2.477616310119629
it: 4500, train recon loss: -0.9179554581642151, local kl: 0.0 global kl: 8.233316806105506e-12 valid reconstr loss: 79.8861312866211
it: 4600, train recon loss: -0.830219030380249, local kl: 0.0 global kl: 1.9246548799145557e-12 valid reconstr loss: -1.0658831596374512
Saving best model with reconstruction loss -1.0658832
it: 4700, train recon loss: -0.8062520623207092, local kl: 0.0 global kl: 2.5548575841721766e-12 valid reconstr loss: -0.8558210134506226
it: 4800, train recon loss: -1.2409738302230835, local kl: 0.0 global kl: 1.8984813721090177e-14 valid reconstr loss: -1.0507899522781372
it: 4900, train recon loss: -0.7885135412216187, local kl: 0.0 global kl: 7.957870473696005e-14 valid reconstr loss: 118.39527893066406
it: 5000, train recon loss: -0.9793991446495056, local kl: 0.0 global kl: 3.0715707755035737e-13 valid reconstr loss: -0.8400052785873413
it: 5100, train recon loss: -1.0054367780685425, local kl: 0.0 global kl: 6.361751403449745e-13 valid reconstr loss: 3653.2392578125
it: 5200, train recon loss: -1.2473397254943848, local kl: 0.0 global kl: 1.4280243654241076e-14 valid reconstr loss: -1.1179068088531494
Saving best model with reconstruction loss -1.1179068
it: 5300, train recon loss: -1.1328814029693604, local kl: 0.0 global kl: 7.310992089504253e-15 valid reconstr loss: -0.981267511844635
it: 5400, train recon loss: -0.8997434377670288, local kl: 0.0 global kl: 1.9539925233402755e-14 valid reconstr loss: 867.414306640625
it: 5500, train recon loss: -0.7479808926582336, local kl: 0.0 global kl: 5.905637090464211e-12 valid reconstr loss: -1.0457298755645752
it: 5600, train recon loss: -1.0617610216140747, local kl: 0.0 global kl: 1.2578501608351278e-13 valid reconstr loss: -0.7982127666473389
it: 5700, train recon loss: 4.447848796844482, local kl: 0.0 global kl: 1.95291699478517e-13 valid reconstr loss: 1.843818187713623
it: 5800, train recon loss: -1.1355700492858887, local kl: 0.0 global kl: 2.415018185430995e-11 valid reconstr loss: -1.0147520303726196
it: 5900, train recon loss: -1.1465786695480347, local kl: 0.0 global kl: 3.889984688532078e-12 valid reconstr loss: -0.6561508178710938
it: 6000, train recon loss: 3406.529296875, local kl: 0.0 global kl: 2.973649104731635e-12 valid reconstr loss: 1783.8446044921875
it: 6100, train recon loss: -1.1566636562347412, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.1894551515579224
Saving best model with reconstruction loss -1.1894552
it: 6200, train recon loss: -0.8756221532821655, local kl: 0.0 global kl: 0.0 valid reconstr loss: 197.6702117919922
it: 6300, train recon loss: 87.6680679321289, local kl: 0.0 global kl: 6.696032617270475e-14 valid reconstr loss: -0.9373677968978882
it: 6400, train recon loss: 134.9055633544922, local kl: 0.0 global kl: 1.578251418443699e-14 valid reconstr loss: -1.0339581966400146
it: 6500, train recon loss: -0.5615220069885254, local kl: 0.0 global kl: 1.8274964874720467e-12 valid reconstr loss: 3.7137653827667236
it: 6600, train recon loss: -1.2219401597976685, local kl: 0.0 global kl: 3.5984034762320816e-12 valid reconstr loss: 425.0392761230469
it: 6700, train recon loss: -1.1546005010604858, local kl: 0.0 global kl: 2.934177694650236e-14 valid reconstr loss: 53.582008361816406
it: 6800, train recon loss: -1.2171415090560913, local kl: 0.0 global kl: 1.7546402698853658e-13 valid reconstr loss: -0.7950325608253479
it: 6900, train recon loss: -0.9622594118118286, local kl: 0.0 global kl: 1.9165224962591765e-14 valid reconstr loss: -1.032884120941162
it: 7000, train recon loss: -1.0932754278182983, local kl: 0.0 global kl: 3.1151744053249664e-14 valid reconstr loss: -1.0141468048095703
it: 7100, train recon loss: -1.134742259979248, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.9457034468650818
it: 7200, train recon loss: -1.2993525266647339, local kl: 0.0 global kl: 5.440092820663267e-15 valid reconstr loss: -1.1934438943862915
Saving best model with reconstruction loss -1.1934439
it: 7300, train recon loss: -0.9774543642997742, local kl: 0.0 global kl: 3.1336044870045043e-14 valid reconstr loss: 1574.779541015625
it: 7400, train recon loss: -0.8743175268173218, local kl: 0.0 global kl: 2.910367259689739e-14 valid reconstr loss: 9.0841703414917
it: 7500, train recon loss: -1.0024563074111938, local kl: 0.0 global kl: 0.0 valid reconstr loss: 6404.01904296875
it: 7600, train recon loss: -1.108189582824707, local kl: 0.0 global kl: 0.0 valid reconstr loss: 79.10566711425781
it: 7700, train recon loss: -1.1999107599258423, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.2251322269439697
Saving best model with reconstruction loss -1.2251322
it: 7800, train recon loss: 2.236250400543213, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.0903608798980713
it: 7900, train recon loss: 28.156688690185547, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.201594740152359
it: 8000, train recon loss: -1.2764779329299927, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.2453761100769043
Saving best model with reconstruction loss -1.2453761
it: 8100, train recon loss: -1.0425260066986084, local kl: 0.0 global kl: 4.777248041598625e-12 valid reconstr loss: -1.2581614255905151
Saving best model with reconstruction loss -1.2581614
it: 8200, train recon loss: -1.2159653902053833, local kl: 0.0 global kl: 2.578492974691926e-14 valid reconstr loss: -0.1906966269016266
it: 8300, train recon loss: -1.2685118913650513, local kl: 0.0 global kl: 2.3359135806200193e-14 valid reconstr loss: 250.47801208496094
it: 8400, train recon loss: 2.9625542163848877, local kl: 0.0 global kl: 3.8191672047105385e-13 valid reconstr loss: 4.717771053314209
it: 8500, train recon loss: 2.951129913330078, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.6489222049713135
it: 8600, train recon loss: 2.1213626861572266, local kl: 0.0 global kl: 0.0 valid reconstr loss: 4.157703399658203
it: 8700, train recon loss: 1.7702457904815674, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.9453513622283936
it: 8800, train recon loss: 0.927756130695343, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.7806169390678406
it: 8900, train recon loss: 0.05668770894408226, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.29875659942626953
it: 9000, train recon loss: 0.21618637442588806, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.5714175701141357
it: 9100, train recon loss: 0.8897076845169067, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.19277213513851166
it: 9200, train recon loss: 1.1050100326538086, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.24420040845870972
it: 9300, train recon loss: 0.04821569100022316, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.5101004838943481
it: 9400, train recon loss: 0.25028669834136963, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.4787631928920746
it: 9500, train recon loss: 0.21161991357803345, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.47944673895835876
it: 9600, train recon loss: 7355.0078125, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.8715214729309082
it: 9700, train recon loss: 8.51093864440918, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.5052705407142639
it: 9800, train recon loss: -0.26619645953178406, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.15868446230888367
it: 9900, train recon loss: -0.19098879396915436, local kl: 0.0 global kl: 4.377921636322668e-13 valid reconstr loss: 0.0559132881462574
beta 0.5 temperature 5.0
it: 0, train recon loss: 813.1907348632812, local kl: 0.0 global kl: 0.0047151073813438416 valid reconstr loss: 691.0077514648438
Saving best model with reconstruction loss 691.00775
it: 100, train recon loss: 3.78196382522583, local kl: 0.0 global kl: 0.0007984127732925117 valid reconstr loss: 3.9896554946899414
Saving best model with reconstruction loss 3.9896555
it: 200, train recon loss: 3.8431522846221924, local kl: 0.0 global kl: 2.793375460896641e-05 valid reconstr loss: 3.8755671977996826
Saving best model with reconstruction loss 3.8755672
it: 300, train recon loss: 27.68033790588379, local kl: 0.0 global kl: 0.0004668154870159924 valid reconstr loss: 3.795754909515381
Saving best model with reconstruction loss 3.795755
it: 400, train recon loss: 2.9846479892730713, local kl: 0.0 global kl: 0.00012607644021045417 valid reconstr loss: 3.740363359451294
Saving best model with reconstruction loss 3.7403634
it: 500, train recon loss: 2.1170363426208496, local kl: 0.0 global kl: 9.108721314987633e-06 valid reconstr loss: 2.434893846511841
Saving best model with reconstruction loss 2.4348938
it: 600, train recon loss: 2.0035130977630615, local kl: 0.0 global kl: 1.4675883619474916e-08 valid reconstr loss: 1.9925007820129395
Saving best model with reconstruction loss 1.9925008
it: 700, train recon loss: 1.7473039627075195, local kl: 0.0 global kl: 6.864040585918829e-14 valid reconstr loss: 1.856979250907898
Saving best model with reconstruction loss 1.8569793
it: 800, train recon loss: 1.8305613994598389, local kl: 0.0 global kl: 3.7143834015485044e-13 valid reconstr loss: 1.828953742980957
Saving best model with reconstruction loss 1.8289537
it: 900, train recon loss: 1.8320064544677734, local kl: 0.0 global kl: 2.874239908579046e-12 valid reconstr loss: 1.7666888236999512
Saving best model with reconstruction loss 1.7666888
it: 1000, train recon loss: 2.063601016998291, local kl: 0.0 global kl: 3.014567762082976e-13 valid reconstr loss: 1.7312476634979248
Saving best model with reconstruction loss 1.7312477
it: 1100, train recon loss: 1.4717613458633423, local kl: 0.0 global kl: 1.983205266675725e-12 valid reconstr loss: 1.8342382907867432
it: 1200, train recon loss: 1.5550371408462524, local kl: 0.0 global kl: 1.0403983230489189e-11 valid reconstr loss: 1.542518973350525
Saving best model with reconstruction loss 1.542519
it: 1300, train recon loss: 1.364650845527649, local kl: 0.0 global kl: 2.686267874807413e-12 valid reconstr loss: 1.5858569145202637
it: 1400, train recon loss: 1.1350739002227783, local kl: 0.0 global kl: 1.4895341904352932e-12 valid reconstr loss: 1.1556397676467896
Saving best model with reconstruction loss 1.1556398
it: 1500, train recon loss: 18.727676391601562, local kl: 0.0 global kl: 1.0898296154415732e-11 valid reconstr loss: 1.060415506362915
Saving best model with reconstruction loss 1.0604155
it: 1600, train recon loss: 0.9899040460586548, local kl: 0.0 global kl: 4.383854390610509e-12 valid reconstr loss: 0.7188313007354736
Saving best model with reconstruction loss 0.7188313
it: 1700, train recon loss: 0.41087231040000916, local kl: 0.0 global kl: 4.401382036611778e-12 valid reconstr loss: 0.6974402070045471
Saving best model with reconstruction loss 0.6974402
it: 1800, train recon loss: 2.588740825653076, local kl: 0.0 global kl: 1.651280501224761e-11 valid reconstr loss: 0.6610276103019714
Saving best model with reconstruction loss 0.6610276
it: 1900, train recon loss: 0.28173843026161194, local kl: 0.0 global kl: 5.111244760769296e-12 valid reconstr loss: 0.5431622266769409
Saving best model with reconstruction loss 0.5431622
it: 2000, train recon loss: 3.6212940216064453, local kl: 0.0 global kl: 1.0769080072137172e-11 valid reconstr loss: 3.5474183559417725
it: 2100, train recon loss: 0.7546382546424866, local kl: 0.0 global kl: 2.820466082908979e-12 valid reconstr loss: 3.0893237590789795
it: 2200, train recon loss: 0.747747540473938, local kl: 0.0 global kl: 5.678582604140558e-13 valid reconstr loss: 0.6890972852706909
it: 2300, train recon loss: 0.41801947355270386, local kl: 0.0 global kl: 2.4121152124301215e-12 valid reconstr loss: 0.832358717918396
it: 2400, train recon loss: 0.3313145339488983, local kl: 0.0 global kl: 2.5673518866398126e-11 valid reconstr loss: 0.6450008153915405
it: 2500, train recon loss: 0.9981069564819336, local kl: 0.0 global kl: 4.1632947089809136e-12 valid reconstr loss: 19.11334228515625
it: 2600, train recon loss: 0.46463868021965027, local kl: 0.0 global kl: 8.380358906823204e-12 valid reconstr loss: 0.9356163144111633
it: 2700, train recon loss: 0.48597848415374756, local kl: 0.0 global kl: 1.5584755708175635e-14 valid reconstr loss: 0.6737067699432373
it: 2800, train recon loss: 0.08137553930282593, local kl: 0.0 global kl: 3.8160100079842607e-13 valid reconstr loss: 0.44330573081970215
Saving best model with reconstruction loss 0.44330573
it: 2900, train recon loss: 0.6371340155601501, local kl: 0.0 global kl: 3.379692359306574e-13 valid reconstr loss: 1.3148956298828125
it: 3000, train recon loss: 0.3028160035610199, local kl: 0.0 global kl: 4.3473905031454763e-13 valid reconstr loss: -0.02090727724134922
Saving best model with reconstruction loss -0.020907277
it: 3100, train recon loss: -0.23886632919311523, local kl: 0.0 global kl: 4.325095837032222e-12 valid reconstr loss: 0.9880253672599792
it: 3200, train recon loss: 0.1683024913072586, local kl: 0.0 global kl: 2.1034424202426294e-12 valid reconstr loss: 0.5569297671318054
it: 3300, train recon loss: 20.817726135253906, local kl: 0.0 global kl: 5.164705468851949e-14 valid reconstr loss: 0.15405914187431335
it: 3400, train recon loss: -0.11748279631137848, local kl: 0.0 global kl: 3.896560157867768e-12 valid reconstr loss: 19.205808639526367
it: 3500, train recon loss: -0.01921268180012703, local kl: 0.0 global kl: 4.236888617725754e-12 valid reconstr loss: 0.10081984847784042
it: 3600, train recon loss: 0.052499301731586456, local kl: 0.0 global kl: 6.481454262186048e-12 valid reconstr loss: -0.09590696543455124
Saving best model with reconstruction loss -0.095906965
it: 3700, train recon loss: -0.42344897985458374, local kl: 0.0 global kl: 5.1861293037802625e-14 valid reconstr loss: -0.07190200686454773
it: 3800, train recon loss: -0.5322306156158447, local kl: 0.0 global kl: 5.1717429573909166e-14 valid reconstr loss: -0.29911670088768005
Saving best model with reconstruction loss -0.2991167
it: 3900, train recon loss: -0.31432485580444336, local kl: 0.0 global kl: 2.4411516644923248e-14 valid reconstr loss: -0.25436151027679443
it: 4000, train recon loss: -0.6720394492149353, local kl: 0.0 global kl: 2.9074312402066482e-12 valid reconstr loss: -0.15805400907993317
it: 4100, train recon loss: -0.08824479579925537, local kl: 0.0 global kl: 3.0362830305552535e-12 valid reconstr loss: -0.43996185064315796
Saving best model with reconstruction loss -0.43996185
it: 4200, train recon loss: -0.14832372963428497, local kl: 0.0 global kl: 4.022563532268819e-14 valid reconstr loss: -0.35152390599250793
it: 4300, train recon loss: -0.3501993417739868, local kl: 0.0 global kl: 3.211320098728265e-14 valid reconstr loss: -0.09749484807252884
it: 4400, train recon loss: -0.36312171816825867, local kl: 0.0 global kl: 1.4651214269578716e-13 valid reconstr loss: -0.13864870369434357
it: 4500, train recon loss: 0.0032236422412097454, local kl: 0.0 global kl: 1.043921893373323e-13 valid reconstr loss: 1102.060791015625
it: 4600, train recon loss: -0.537543773651123, local kl: 0.0 global kl: 5.630232391418133e-12 valid reconstr loss: -0.27389100193977356
it: 4700, train recon loss: -0.27620929479599, local kl: 0.0 global kl: 7.464488012073889e-13 valid reconstr loss: 70.579345703125
it: 4800, train recon loss: -0.8307538628578186, local kl: 0.0 global kl: 1.865174681370263e-13 valid reconstr loss: -0.3489942252635956
it: 4900, train recon loss: 0.09370148926973343, local kl: 0.0 global kl: 1.9829970998586077e-13 valid reconstr loss: 4.063752174377441
it: 5000, train recon loss: -0.4850018620491028, local kl: 0.0 global kl: 1.865174681370263e-14 valid reconstr loss: -0.15509167313575745
it: 5100, train recon loss: -0.6073165535926819, local kl: 0.0 global kl: 7.310840301200106e-14 valid reconstr loss: 0.37499022483825684
it: 5200, train recon loss: -0.9179151654243469, local kl: 0.0 global kl: 1.4650017310380292e-12 valid reconstr loss: -0.534130871295929
Saving best model with reconstruction loss -0.5341309
it: 5300, train recon loss: -0.5575591921806335, local kl: 0.0 global kl: 4.155121559323849e-12 valid reconstr loss: -0.5192291140556335
it: 5400, train recon loss: -0.10315079987049103, local kl: 0.0 global kl: 1.1272233146897292e-14 valid reconstr loss: -0.1022358313202858
it: 5500, train recon loss: -0.3904525637626648, local kl: 0.0 global kl: 4.2159630821159144e-12 valid reconstr loss: -0.3634099066257477
it: 5600, train recon loss: -0.47793495655059814, local kl: 0.0 global kl: 1.5569837086282234e-13 valid reconstr loss: -0.273995578289032
it: 5700, train recon loss: -0.40659424662590027, local kl: 0.0 global kl: 4.530143621339633e-14 valid reconstr loss: -0.618455171585083
Saving best model with reconstruction loss -0.6184552
it: 5800, train recon loss: -0.679033100605011, local kl: 0.0 global kl: 7.102651800039439e-14 valid reconstr loss: -0.41165271401405334
it: 5900, train recon loss: -0.6592832803726196, local kl: 0.0 global kl: 5.606438924221635e-12 valid reconstr loss: -0.35326603055000305
it: 6000, train recon loss: -0.8834467530250549, local kl: 0.0 global kl: 1.0575047781902214e-13 valid reconstr loss: -0.6551831960678101
Saving best model with reconstruction loss -0.6551832
it: 6100, train recon loss: 3.8390073776245117, local kl: 0.0 global kl: 2.914335439641036e-16 valid reconstr loss: 62.77560043334961
it: 6200, train recon loss: -0.7799902558326721, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.12525597214698792
it: 6300, train recon loss: 0.8671324253082275, local kl: 0.0 global kl: 2.0020858343579306e-13 valid reconstr loss: -0.7732298374176025
Saving best model with reconstruction loss -0.77322984
it: 6400, train recon loss: 0.10711795091629028, local kl: 0.0 global kl: 9.793316331496715e-13 valid reconstr loss: -0.3056202530860901
it: 6500, train recon loss: -0.5836174488067627, local kl: 0.0 global kl: 3.0416884289063972e-12 valid reconstr loss: 1.9302252531051636
it: 6600, train recon loss: -0.9278726577758789, local kl: 0.0 global kl: 6.168893521008023e-15 valid reconstr loss: 157.0760498046875
it: 6700, train recon loss: 1.122251272201538, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.9503602981567383
it: 6800, train recon loss: -0.5788876414299011, local kl: 0.0 global kl: 2.1586898935055387e-14 valid reconstr loss: -0.5567150712013245
it: 6900, train recon loss: -0.6514870524406433, local kl: 0.0 global kl: 2.7481489306424578e-14 valid reconstr loss: -0.6780979037284851
it: 7000, train recon loss: -0.6248819231987, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.7327276468276978
it: 7100, train recon loss: -0.8275203704833984, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.7857770323753357
Saving best model with reconstruction loss -0.78577703
it: 7200, train recon loss: 1.2109419107437134, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.9999915957450867
Saving best model with reconstruction loss -0.9999916
it: 7300, train recon loss: -0.7736329436302185, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.5931157469749451
it: 7400, train recon loss: -0.5957210659980774, local kl: 0.0 global kl: 0.0 valid reconstr loss: 859.5025024414062
it: 7500, train recon loss: -1.046921968460083, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.7956264019012451
it: 7600, train recon loss: -0.783295750617981, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.8819735050201416
it: 7700, train recon loss: -0.8858333230018616, local kl: 0.0 global kl: 0.0 valid reconstr loss: 10.789769172668457
it: 7800, train recon loss: -0.8060868382453918, local kl: 0.0 global kl: 0.0 valid reconstr loss: 875.128662109375
it: 7900, train recon loss: -0.17739029228687286, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.028229832649231
Saving best model with reconstruction loss -1.0282298
it: 8000, train recon loss: -0.9622960686683655, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.654760718345642
it: 8100, train recon loss: 76.20210266113281, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.1286134719848633
Saving best model with reconstruction loss -1.1286135
it: 8200, train recon loss: 3.0725667476654053, local kl: 0.0 global kl: 0.0 valid reconstr loss: 4.258289337158203
it: 8300, train recon loss: 3.094804048538208, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.05126690864563
it: 8400, train recon loss: 2.7698020935058594, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.753702402114868
it: 8500, train recon loss: 3.0772576332092285, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.5326426029205322
it: 8600, train recon loss: 3.240898370742798, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.086474895477295
it: 8700, train recon loss: 3.0402395725250244, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.044029474258423
it: 8800, train recon loss: 2.726325273513794, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.0085036754608154
it: 8900, train recon loss: 2.8533098697662354, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.0139424800872803
it: 9000, train recon loss: 2.7358384132385254, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.116132974624634
it: 9100, train recon loss: 3.1405434608459473, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.088545560836792
it: 9200, train recon loss: 2.9025111198425293, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.952918767929077
it: 9300, train recon loss: 2.93292498588562, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.0506722927093506
it: 9400, train recon loss: 3.040475606918335, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.0512678623199463
it: 9500, train recon loss: 3.0096611976623535, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.100572347640991
it: 9600, train recon loss: 2.9247210025787354, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.0281713008880615
it: 9700, train recon loss: 2.9709126949310303, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.0221617221832275
it: 9800, train recon loss: 3.0703864097595215, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.9755799770355225
it: 9900, train recon loss: 2.6995692253112793, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.9933745861053467
beta 0.5 temperature 1000.0
it: 0, train recon loss: 232.04954528808594, local kl: 0.0 global kl: 0.008364617824554443 valid reconstr loss: 66.7215347290039
Saving best model with reconstruction loss 66.721535
it: 100, train recon loss: 3.657794713973999, local kl: 0.0 global kl: 0.0008518256363458931 valid reconstr loss: 4.050600528717041
Saving best model with reconstruction loss 4.0506005
it: 200, train recon loss: 3.900050401687622, local kl: 0.0 global kl: 2.3475597117794678e-05 valid reconstr loss: 3.9583287239074707
Saving best model with reconstruction loss 3.9583287
it: 300, train recon loss: 3.6579883098602295, local kl: 0.0 global kl: 2.148553539882414e-05 valid reconstr loss: 30.533353805541992
it: 400, train recon loss: 2.9087672233581543, local kl: 0.0 global kl: 1.1038438287869212e-06 valid reconstr loss: 3.128913640975952
Saving best model with reconstruction loss 3.1289136
it: 500, train recon loss: 2.522622585296631, local kl: 0.0 global kl: 5.791135890831356e-07 valid reconstr loss: 2.748342275619507
Saving best model with reconstruction loss 2.7483423
it: 600, train recon loss: 1.7481588125228882, local kl: 0.0 global kl: 2.0095036745715333e-13 valid reconstr loss: 2.1332149505615234
Saving best model with reconstruction loss 2.133215
it: 700, train recon loss: 2.4132494926452637, local kl: 0.0 global kl: 3.5186298319445086e-12 valid reconstr loss: 2.049454927444458
Saving best model with reconstruction loss 2.049455
it: 800, train recon loss: 1.31778883934021, local kl: 0.0 global kl: 1.5203463488155933e-12 valid reconstr loss: 7.216444969177246
it: 900, train recon loss: 0.9892701506614685, local kl: 0.0 global kl: 4.113376306236205e-14 valid reconstr loss: 2.1914634704589844
it: 1000, train recon loss: 0.6971538662910461, local kl: 0.0 global kl: 3.73364186789793e-12 valid reconstr loss: 1.5368350744247437
Saving best model with reconstruction loss 1.5368351
it: 1100, train recon loss: 0.5623058080673218, local kl: 0.0 global kl: 1.4369061496211089e-12 valid reconstr loss: 0.9410572648048401
Saving best model with reconstruction loss 0.94105726
it: 1200, train recon loss: 0.8534618020057678, local kl: 0.0 global kl: 6.7491233955718766e-12 valid reconstr loss: 0.6449869871139526
Saving best model with reconstruction loss 0.644987
it: 1300, train recon loss: 0.5034409165382385, local kl: 0.0 global kl: 1.85808313180047e-12 valid reconstr loss: 0.25448206067085266
Saving best model with reconstruction loss 0.25448206
it: 1400, train recon loss: 104.53550720214844, local kl: 0.0 global kl: 1.0027648850163828e-11 valid reconstr loss: 33.19124221801758
it: 1500, train recon loss: 0.013046066276729107, local kl: 0.0 global kl: 2.5135353867722365e-12 valid reconstr loss: 0.18123412132263184
Saving best model with reconstruction loss 0.18123412
it: 1600, train recon loss: 0.20998547971248627, local kl: 0.0 global kl: 2.499889184548465e-12 valid reconstr loss: 0.07118567824363708
Saving best model with reconstruction loss 0.07118568
it: 1700, train recon loss: 0.7326025366783142, local kl: 0.0 global kl: 2.2714408479118653e-12 valid reconstr loss: 0.15606515109539032
it: 1800, train recon loss: 137.6766815185547, local kl: 0.0 global kl: 2.209296547789341e-12 valid reconstr loss: -0.004712799098342657
Saving best model with reconstruction loss -0.004712799
it: 1900, train recon loss: 0.04222075641155243, local kl: 0.0 global kl: 1.1719965276046906e-12 valid reconstr loss: 0.3836975693702698
it: 2000, train recon loss: 0.7628743052482605, local kl: 0.0 global kl: 6.032341293149557e-12 valid reconstr loss: 6.885401248931885
it: 2100, train recon loss: -0.3540705442428589, local kl: 0.0 global kl: 6.591352463836131e-12 valid reconstr loss: 0.7051268815994263
it: 2200, train recon loss: 26.715599060058594, local kl: 0.0 global kl: 1.5459994395783383e-12 valid reconstr loss: 0.41721203923225403
it: 2300, train recon loss: -0.20837895572185516, local kl: 0.0 global kl: 4.6827541844152165e-12 valid reconstr loss: 50.08363342285156
it: 2400, train recon loss: 1.9480931758880615, local kl: 0.0 global kl: 2.2416790645962692e-13 valid reconstr loss: 1470.7420654296875
it: 2500, train recon loss: 0.149928018450737, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.16886849701404572
Saving best model with reconstruction loss -0.1688685
it: 2600, train recon loss: -0.6061707139015198, local kl: 0.0 global kl: 7.327818907221229e-14 valid reconstr loss: 0.2750190496444702
it: 2700, train recon loss: -0.25596773624420166, local kl: 0.0 global kl: 8.529288386682765e-14 valid reconstr loss: 0.1490553319454193
it: 2800, train recon loss: -0.16011899709701538, local kl: 0.0 global kl: 5.992428775414282e-14 valid reconstr loss: 217.68197631835938
it: 2900, train recon loss: -0.40373826026916504, local kl: 0.0 global kl: 3.061675912796602e-12 valid reconstr loss: 0.014730127528309822
it: 3000, train recon loss: 16.72557830810547, local kl: 0.0 global kl: 3.2686076067989234e-11 valid reconstr loss: 44.7846565246582
it: 3100, train recon loss: -0.5042106509208679, local kl: 0.0 global kl: 0.0 valid reconstr loss: 53.29459762573242
it: 3200, train recon loss: 0.017140401527285576, local kl: 0.0 global kl: 2.68367359584909e-12 valid reconstr loss: 33.602481842041016
it: 3300, train recon loss: -0.33662790060043335, local kl: 0.0 global kl: 7.03571922944235e-12 valid reconstr loss: -0.20862731337547302
Saving best model with reconstruction loss -0.20862731
it: 3400, train recon loss: -0.4451284408569336, local kl: 0.0 global kl: 1.3187402558845207e-12 valid reconstr loss: 1403.29931640625
it: 3500, train recon loss: -0.34780746698379517, local kl: 0.0 global kl: 4.056846872324549e-12 valid reconstr loss: 168.51324462890625
it: 3600, train recon loss: 1.5310031175613403, local kl: 0.0 global kl: 1.1495451292253822e-11 valid reconstr loss: 1.5086734294891357
it: 3700, train recon loss: -0.7121217250823975, local kl: 0.0 global kl: 3.894913037927328e-12 valid reconstr loss: -0.43320125341415405
Saving best model with reconstruction loss -0.43320125
it: 3800, train recon loss: -0.38254860043525696, local kl: 0.0 global kl: 3.032296636007459e-14 valid reconstr loss: -0.44131359457969666
Saving best model with reconstruction loss -0.4413136
it: 3900, train recon loss: -0.3477414548397064, local kl: 0.0 global kl: 8.357897707256257e-15 valid reconstr loss: 78.57489013671875
it: 4000, train recon loss: -0.8695311546325684, local kl: 0.0 global kl: 2.0550618498593742e-14 valid reconstr loss: -0.5584427714347839
Saving best model with reconstruction loss -0.5584428
it: 4100, train recon loss: -0.5472204089164734, local kl: 0.0 global kl: 5.172485269960836e-12 valid reconstr loss: 0.7172085046768188
it: 4200, train recon loss: -0.4821023941040039, local kl: 0.0 global kl: 1.6229656507604773e-11 valid reconstr loss: 125.46577453613281
it: 4300, train recon loss: -0.26727116107940674, local kl: 0.0 global kl: 1.2007062011321068e-13 valid reconstr loss: -0.41557401418685913
it: 4400, train recon loss: 0.7969025373458862, local kl: 0.0 global kl: 6.46024574980919e-14 valid reconstr loss: 8.434649467468262
it: 4500, train recon loss: 0.2016497701406479, local kl: 0.0 global kl: 1.8750626051833308e-12 valid reconstr loss: -0.7653447389602661
Saving best model with reconstruction loss -0.76534474
it: 4600, train recon loss: 14.756792068481445, local kl: 0.0 global kl: 4.2945647038550305e-12 valid reconstr loss: 7.361722946166992
it: 4700, train recon loss: -0.6703091263771057, local kl: 0.0 global kl: 2.288960687657493e-12 valid reconstr loss: -0.6707092523574829
it: 4800, train recon loss: -1.0375282764434814, local kl: 0.0 global kl: 9.151013280472853e-13 valid reconstr loss: 0.4935493767261505
it: 4900, train recon loss: 1.6806412935256958, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.2115263044834137
it: 5000, train recon loss: 444.61236572265625, local kl: 0.0 global kl: 4.8396078811130394e-12 valid reconstr loss: 303.9066162109375
it: 5100, train recon loss: -0.6008149981498718, local kl: 0.0 global kl: 1.826053197540034e-12 valid reconstr loss: -0.7996123433113098
Saving best model with reconstruction loss -0.79961234
it: 5200, train recon loss: -0.7553783059120178, local kl: 0.0 global kl: 2.248201624865942e-15 valid reconstr loss: -0.5412806272506714
it: 5300, train recon loss: 3.1609368324279785, local kl: 0.0 global kl: 7.444132116285473e-14 valid reconstr loss: 3.8246641159057617
it: 5400, train recon loss: -0.8288528919219971, local kl: 0.0 global kl: 1.8267574952712806e-12 valid reconstr loss: -0.5765736103057861
it: 5500, train recon loss: -0.7960267663002014, local kl: 0.0 global kl: 2.381775332516156e-14 valid reconstr loss: -0.6485699415206909
it: 5600, train recon loss: 167.77638244628906, local kl: 0.0 global kl: 3.6061431618605866e-14 valid reconstr loss: 0.029938431456685066
it: 5700, train recon loss: 3.3867809772491455, local kl: 0.0 global kl: 2.906246857753425e-12 valid reconstr loss: 53.87858200073242
it: 5800, train recon loss: -0.6837198734283447, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1891.6568603515625
it: 5900, train recon loss: -1.353438138961792, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.285277783870697
it: 6000, train recon loss: -1.1022847890853882, local kl: 0.0 global kl: 0.0 valid reconstr loss: 352.8850402832031
it: 6100, train recon loss: -0.9229846000671387, local kl: 0.0 global kl: 4.4389839026770517e-14 valid reconstr loss: -0.5297971963882446
it: 6200, train recon loss: -0.619230329990387, local kl: 0.0 global kl: 2.1344037648418634e-13 valid reconstr loss: -0.988309919834137
Saving best model with reconstruction loss -0.9883099
it: 6300, train recon loss: -0.8809382319450378, local kl: 0.0 global kl: 3.240810397819871e-14 valid reconstr loss: -0.5907385945320129
it: 6400, train recon loss: 1582.246826171875, local kl: 0.0 global kl: 2.077366056951746e-13 valid reconstr loss: -0.7193630337715149
it: 6500, train recon loss: -0.8978066444396973, local kl: 0.0 global kl: 2.942091015256665e-14 valid reconstr loss: 5134.85498046875
it: 6600, train recon loss: -0.9900282025337219, local kl: 0.0 global kl: 9.368329029202371e-12 valid reconstr loss: -0.6917920112609863
it: 6700, train recon loss: 0.6640861630439758, local kl: 0.0 global kl: 9.885321727853835e-15 valid reconstr loss: 13.751041412353516
it: 6800, train recon loss: 72.21983337402344, local kl: 0.0 global kl: 1.495431556364224e-11 valid reconstr loss: 38.294960021972656
it: 6900, train recon loss: 1.0507590770721436, local kl: 0.0 global kl: 7.799316747991725e-15 valid reconstr loss: 2.583458662033081
it: 7000, train recon loss: 0.33385977149009705, local kl: 0.0 global kl: 3.6790462711912275e-14 valid reconstr loss: 0.0791829526424408
it: 7100, train recon loss: -0.43603515625, local kl: 0.0 global kl: 4.32093597013683e-14 valid reconstr loss: -0.7512413859367371
it: 7200, train recon loss: -0.9170498847961426, local kl: 0.0 global kl: 1.823888262642015e-14 valid reconstr loss: -1.010778546333313
Saving best model with reconstruction loss -1.0107785
it: 7300, train recon loss: 769.7457885742188, local kl: 0.0 global kl: 0.0 valid reconstr loss: 41.97228240966797
it: 7400, train recon loss: 39.25861740112305, local kl: 0.0 global kl: 6.672884467207041e-12 valid reconstr loss: 0.8189677596092224
it: 7500, train recon loss: -1.2060489654541016, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.0057305097579956
it: 7600, train recon loss: -0.7028245329856873, local kl: 0.0 global kl: 1.0215981706418464e-13 valid reconstr loss: -0.6094385981559753
it: 7700, train recon loss: -1.1237515211105347, local kl: 0.0 global kl: 7.036472099430924e-14 valid reconstr loss: -0.6454492211341858
it: 7800, train recon loss: -0.8558049201965332, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.1155794858932495
Saving best model with reconstruction loss -1.1155795
it: 7900, train recon loss: 255.35020446777344, local kl: 0.0 global kl: 5.451306073211981e-11 valid reconstr loss: 1.6089633703231812
it: 8000, train recon loss: -1.0323501825332642, local kl: 0.0 global kl: 1.5950852834747953e-13 valid reconstr loss: -1.2245444059371948
Saving best model with reconstruction loss -1.2245444
it: 8100, train recon loss: -1.094887137413025, local kl: 0.0 global kl: 2.300883658948316e-12 valid reconstr loss: -1.0987286567687988
it: 8200, train recon loss: -1.402226448059082, local kl: 0.0 global kl: 1.233822072288504e-14 valid reconstr loss: -1.015486240386963
it: 8300, train recon loss: 1.698136329650879, local kl: 0.0 global kl: 6.622480341889059e-14 valid reconstr loss: -1.0224900245666504
it: 8400, train recon loss: -1.0264545679092407, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.1846809387207031
it: 8500, train recon loss: -1.0164785385131836, local kl: 0.0 global kl: 8.215650382226158e-15 valid reconstr loss: 10241.6748046875
it: 8600, train recon loss: 2.9074385166168213, local kl: 0.0 global kl: 0.0 valid reconstr loss: 232.767333984375
it: 8700, train recon loss: -1.1798170804977417, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.9506881833076477
it: 8800, train recon loss: -0.4729211926460266, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.8865511417388916
it: 8900, train recon loss: -1.3585095405578613, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.22907114028930664
it: 9000, train recon loss: 31.78639793395996, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.7109886407852173
it: 9100, train recon loss: -0.46082884073257446, local kl: 0.0 global kl: 0.0 valid reconstr loss: 17.569013595581055
it: 9200, train recon loss: -1.3208410739898682, local kl: 0.0 global kl: 0.0 valid reconstr loss: 16.14309310913086
it: 9300, train recon loss: -0.8384004831314087, local kl: 0.0 global kl: 1.3134774572241131e-13 valid reconstr loss: -0.27527207136154175
it: 9400, train recon loss: -1.019873857498169, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.1194508075714111
it: 9500, train recon loss: 6.434832572937012, local kl: 0.0 global kl: 1.5030709316077306e-11 valid reconstr loss: 0.9037910103797913
it: 9600, train recon loss: 4.3156352043151855, local kl: 0.0 global kl: 1.848868280696081e-14 valid reconstr loss: 641430336.0
it: 9700, train recon loss: 132.61387634277344, local kl: 0.0 global kl: 7.470024598887903e-12 valid reconstr loss: 3.4001970291137695
it: 9800, train recon loss: 2.285372018814087, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.4308691024780273
it: 9900, train recon loss: 1.220193862915039, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.4953688383102417
beta 1.0 temperature 0.001
it: 0, train recon loss: 238201872.0, local kl: 0.0 global kl: 0.012519720010459423 valid reconstr loss: 3106289.25
Saving best model with reconstruction loss 3106289.2
it: 100, train recon loss: 3.475613832473755, local kl: 0.0 global kl: 0.0002088243782054633 valid reconstr loss: 4.944136142730713
Saving best model with reconstruction loss 4.944136
it: 200, train recon loss: 2.960510730743408, local kl: 0.0 global kl: 5.984329618513584e-05 valid reconstr loss: 2.84735369682312
Saving best model with reconstruction loss 2.8473537
it: 300, train recon loss: 2.273653268814087, local kl: 0.0 global kl: 6.50685396976769e-05 valid reconstr loss: 2.29744815826416
Saving best model with reconstruction loss 2.2974482
it: 400, train recon loss: 1.8508565425872803, local kl: 0.0 global kl: 2.724995465541724e-05 valid reconstr loss: 1.91544508934021
Saving best model with reconstruction loss 1.9154451
it: 500, train recon loss: 1.7514338493347168, local kl: 0.0 global kl: 0.00029656122205778956 valid reconstr loss: 33.54578399658203
it: 600, train recon loss: 1.5601433515548706, local kl: 0.0 global kl: 7.388999074464664e-05 valid reconstr loss: 2.0026068687438965
it: 700, train recon loss: 0.693009078502655, local kl: 0.0 global kl: 4.6527926315320656e-05 valid reconstr loss: 1.3681128025054932
Saving best model with reconstruction loss 1.3681128
it: 800, train recon loss: 0.7517452836036682, local kl: 0.0 global kl: 0.00011209090007469058 valid reconstr loss: 11.539917945861816
it: 900, train recon loss: 0.6592154502868652, local kl: 0.0 global kl: 2.76646860584151e-05 valid reconstr loss: 0.6238606572151184
Saving best model with reconstruction loss 0.62386066
it: 1000, train recon loss: 0.8688645362854004, local kl: 0.0 global kl: 6.892271630931646e-05 valid reconstr loss: 2.01011061668396
it: 1100, train recon loss: 0.24112409353256226, local kl: 0.0 global kl: 0.00021663866937160492 valid reconstr loss: 2.2146852016448975
it: 1200, train recon loss: 0.2812499403953552, local kl: 0.0 global kl: 0.00041971198515966535 valid reconstr loss: 1.122753381729126
it: 1300, train recon loss: 46.3787727355957, local kl: 0.0 global kl: 0.050103891640901566 valid reconstr loss: 749.7607421875
it: 1400, train recon loss: 156.49325561523438, local kl: 0.0 global kl: 4.6592198486905545e-05 valid reconstr loss: 1.2980307340621948
it: 1500, train recon loss: -0.41182181239128113, local kl: 0.0 global kl: 5.2997260354459286e-05 valid reconstr loss: 3.8912956714630127
it: 1600, train recon loss: 0.4759983420372009, local kl: 0.0 global kl: 3.723118788911961e-05 valid reconstr loss: 760.6087036132812
it: 1700, train recon loss: -0.2030496895313263, local kl: 0.0 global kl: 5.955399319645949e-05 valid reconstr loss: 0.9386849403381348
it: 1800, train recon loss: 0.06681569665670395, local kl: 0.0 global kl: 5.991821672068909e-05 valid reconstr loss: 39.863948822021484
it: 1900, train recon loss: 0.3130240738391876, local kl: 0.0 global kl: 1.6532761947019026e-05 valid reconstr loss: -0.13948775827884674
Saving best model with reconstruction loss -0.13948776
it: 2000, train recon loss: 8467.7978515625, local kl: 0.0 global kl: 8.097686077235267e-06 valid reconstr loss: 469.1603698730469
it: 2100, train recon loss: -0.49949464201927185, local kl: 0.0 global kl: 1.8710010408540256e-05 valid reconstr loss: -0.4665374159812927
Saving best model with reconstruction loss -0.46653742
it: 2200, train recon loss: -0.19540652632713318, local kl: 0.0 global kl: 0.00040175748290494084 valid reconstr loss: 0.6430242657661438
it: 2300, train recon loss: 10.861920356750488, local kl: 0.0 global kl: 0.0001995790225919336 valid reconstr loss: 0.07763964682817459
it: 2400, train recon loss: 22.411277770996094, local kl: 0.0 global kl: 3.085427306359634e-05 valid reconstr loss: -0.1819668710231781
it: 2500, train recon loss: 149.77284240722656, local kl: 0.0 global kl: 8.796819020062685e-05 valid reconstr loss: 82.12391662597656
it: 2600, train recon loss: -0.6662352085113525, local kl: 0.0 global kl: 0.0002713404828682542 valid reconstr loss: -0.2651672661304474
it: 2700, train recon loss: 106.36983489990234, local kl: 0.0 global kl: 8.727751264814287e-05 valid reconstr loss: -0.051225826144218445
it: 2800, train recon loss: -0.3729304373264313, local kl: 0.0 global kl: 0.00013698723341803998 valid reconstr loss: -0.5674495100975037
Saving best model with reconstruction loss -0.5674495
it: 2900, train recon loss: -0.4069942533969879, local kl: 0.0 global kl: 0.0004647425375878811 valid reconstr loss: 21.26088523864746
it: 3000, train recon loss: -0.427033007144928, local kl: 0.0 global kl: 3.1112133001443e-05 valid reconstr loss: -0.009553832933306694
it: 3100, train recon loss: -1.0935068130493164, local kl: 0.0 global kl: 0.00024547925568185747 valid reconstr loss: -0.0453915037214756
it: 3200, train recon loss: -0.507560133934021, local kl: 0.0 global kl: 9.310952009400353e-05 valid reconstr loss: 1.4740843772888184
it: 3300, train recon loss: 1.910452961921692, local kl: 0.0 global kl: 1.0318090971850324e-05 valid reconstr loss: 1.8102463483810425
it: 3400, train recon loss: 5.717299461364746, local kl: 0.0 global kl: 0.00047194299986585975 valid reconstr loss: -0.2951104938983917
it: 3500, train recon loss: -0.8669620156288147, local kl: 0.0 global kl: 0.0002713021240197122 valid reconstr loss: 916.6925048828125
it: 3600, train recon loss: -0.8182851672172546, local kl: 0.0 global kl: 1.7104801372624934e-05 valid reconstr loss: 1779.5615234375
it: 3700, train recon loss: -0.6931092739105225, local kl: 0.0 global kl: 6.406526608770946e-06 valid reconstr loss: 0.029841521754860878
it: 3800, train recon loss: -0.2378532737493515, local kl: 0.0 global kl: 9.121412585955113e-05 valid reconstr loss: 5952.94482421875
it: 3900, train recon loss: -0.06934444606304169, local kl: 0.0 global kl: 1.5395575246657245e-05 valid reconstr loss: 4.747864246368408
it: 4000, train recon loss: -0.7467439770698547, local kl: 0.0 global kl: 9.288582987210248e-06 valid reconstr loss: 1.8527038097381592
it: 4100, train recon loss: -1.2245700359344482, local kl: 0.0 global kl: 2.023413617280312e-05 valid reconstr loss: -1.0956515073776245
Saving best model with reconstruction loss -1.0956515
it: 4200, train recon loss: -0.6127234101295471, local kl: 0.0 global kl: 0.0001874340232461691 valid reconstr loss: -0.6677449941635132
it: 4300, train recon loss: -1.1443246603012085, local kl: 0.0 global kl: 2.6755346880236175e-06 valid reconstr loss: -0.9361098408699036
it: 4400, train recon loss: -0.8801137208938599, local kl: 0.0 global kl: 2.6412640181661118e-06 valid reconstr loss: -0.9690253734588623
it: 4500, train recon loss: 43.31295394897461, local kl: 0.0 global kl: 3.585703234421089e-05 valid reconstr loss: 0.937146782875061
it: 4600, train recon loss: 0.042411040514707565, local kl: 0.0 global kl: 7.856475349399261e-06 valid reconstr loss: 0.5809460878372192
it: 4700, train recon loss: -0.45684558153152466, local kl: 0.0 global kl: 4.716983767139027e-06 valid reconstr loss: -0.49587464332580566
it: 4800, train recon loss: 1.8237372636795044, local kl: 0.0 global kl: 5.9660596889443696e-06 valid reconstr loss: 0.9846855401992798
it: 4900, train recon loss: -0.9055657982826233, local kl: 0.0 global kl: 3.2330113754142076e-05 valid reconstr loss: 782.6768188476562
it: 5000, train recon loss: -0.7285730838775635, local kl: 0.0 global kl: 2.5732373615028337e-05 valid reconstr loss: -0.9764475226402283
it: 5100, train recon loss: -0.7230345606803894, local kl: 0.0 global kl: 0.0003869763750117272 valid reconstr loss: -0.547970175743103
it: 5200, train recon loss: -0.42859935760498047, local kl: 0.0 global kl: 8.35353130241856e-05 valid reconstr loss: -0.6873867511749268
it: 5300, train recon loss: 31.345279693603516, local kl: 0.0 global kl: 6.396019307430834e-05 valid reconstr loss: 9.64041519165039
it: 5400, train recon loss: -0.5510979890823364, local kl: 0.0 global kl: 8.188603533199057e-05 valid reconstr loss: -0.2790377736091614
it: 5500, train recon loss: -0.7136335968971252, local kl: 0.0 global kl: 0.00011325208470225334 valid reconstr loss: -0.33699023723602295
it: 5600, train recon loss: -1.447287678718567, local kl: 0.0 global kl: 4.450038250070065e-05 valid reconstr loss: -0.2876851260662079
it: 5700, train recon loss: -1.34453547000885, local kl: 0.0 global kl: 4.7710927901789546e-05 valid reconstr loss: 0.06550875306129456
it: 5800, train recon loss: -1.231734037399292, local kl: 0.0 global kl: 1.549997614347376e-05 valid reconstr loss: -1.1115823984146118
Saving best model with reconstruction loss -1.1115824
it: 5900, train recon loss: -1.4181984663009644, local kl: 0.0 global kl: 1.719677311484702e-05 valid reconstr loss: -1.120776891708374
Saving best model with reconstruction loss -1.1207769
it: 6000, train recon loss: -1.435347318649292, local kl: 0.0 global kl: 2.171170672227163e-05 valid reconstr loss: -0.3802540898323059
it: 6100, train recon loss: -1.1282529830932617, local kl: 0.0 global kl: 1.639927904761862e-05 valid reconstr loss: -0.9217449426651001
it: 6200, train recon loss: -1.2792481184005737, local kl: 0.0 global kl: 5.391278591559967e-06 valid reconstr loss: -1.2277965545654297
Saving best model with reconstruction loss -1.2277966
it: 6300, train recon loss: -1.0462472438812256, local kl: 0.0 global kl: 1.0006137927121017e-06 valid reconstr loss: -0.9820018410682678
it: 6400, train recon loss: 16609.677734375, local kl: 0.0 global kl: 2.8438312256184872e-06 valid reconstr loss: -0.6368967294692993
it: 6500, train recon loss: -1.1768304109573364, local kl: 0.0 global kl: 3.4566612612252356e-06 valid reconstr loss: -1.0346088409423828
it: 6600, train recon loss: 1182.3726806640625, local kl: 0.0 global kl: 0.00010529928840696812 valid reconstr loss: -0.9937556385993958
it: 6700, train recon loss: 2335.23193359375, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.9340077638626099
it: 6800, train recon loss: 1.083561658859253, local kl: 0.0 global kl: 0.0013046141248196363 valid reconstr loss: -0.6564159393310547
it: 6900, train recon loss: 5.503300666809082, local kl: 0.0 global kl: 0.0003007151244673878 valid reconstr loss: 1.9775701761245728
it: 7000, train recon loss: 1.0413222312927246, local kl: 0.0 global kl: 5.116401007398963e-05 valid reconstr loss: 1.504281759262085
it: 7100, train recon loss: 1.4914683103561401, local kl: 0.0 global kl: 1.990233431570232e-05 valid reconstr loss: -0.5186090469360352
it: 7200, train recon loss: -0.6136521697044373, local kl: 0.0 global kl: 3.839410965156276e-06 valid reconstr loss: -0.2143065333366394
it: 7300, train recon loss: -1.1021370887756348, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.0416067838668823
it: 7400, train recon loss: -0.6252458095550537, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.3234025537967682
it: 7500, train recon loss: -1.2917424440383911, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.0982654094696045
it: 7600, train recon loss: -1.2314566373825073, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.001447319984436
it: 7700, train recon loss: -1.0990949869155884, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.6913392543792725
it: 7800, train recon loss: -0.9050388336181641, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.9097891449928284
it: 7900, train recon loss: -0.6794388890266418, local kl: 0.0 global kl: 2.157486233045347e-05 valid reconstr loss: -0.02751011587679386
it: 8000, train recon loss: 1.1200491189956665, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.7197983264923096
it: 8100, train recon loss: -0.9115350842475891, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.0805734395980835
it: 8200, train recon loss: -1.1719762086868286, local kl: 0.0 global kl: 0.0 valid reconstr loss: 8433.353515625
it: 8300, train recon loss: -1.4056378602981567, local kl: 0.0 global kl: 5.7873548939824104e-05 valid reconstr loss: 82.09278869628906
it: 8400, train recon loss: -1.253493309020996, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.0973564386367798
it: 8500, train recon loss: 2.0848100185394287, local kl: 0.0 global kl: 4.84829542983789e-05 valid reconstr loss: 2.417804718017578
it: 8600, train recon loss: 4.9654693603515625, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.42367076873779297
it: 8700, train recon loss: -1.2123337984085083, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.224054217338562
it: 8800, train recon loss: -1.3722091913223267, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.2795192003250122
Saving best model with reconstruction loss -1.2795192
it: 8900, train recon loss: 2.9501168727874756, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.024566888809204
it: 9000, train recon loss: -0.2860139012336731, local kl: 0.0 global kl: 7.783471664879471e-05 valid reconstr loss: -0.013516548089683056
it: 9100, train recon loss: 0.3099435567855835, local kl: 0.0 global kl: 0.0 valid reconstr loss: 4964.67333984375
it: 9200, train recon loss: -1.2591897249221802, local kl: 0.0 global kl: 1.2582047020259779e-05 valid reconstr loss: -1.2757118940353394
it: 9300, train recon loss: 3.0636794567108154, local kl: 0.0 global kl: 6.0509984905365855e-05 valid reconstr loss: 3.8030097484588623
it: 9400, train recon loss: -1.1236003637313843, local kl: 0.0 global kl: 2.2676085791317746e-06 valid reconstr loss: -1.1143242120742798
it: 9500, train recon loss: 0.7272095680236816, local kl: 0.0 global kl: 5.212778887653258e-06 valid reconstr loss: 0.9028248190879822
it: 9600, train recon loss: -0.8546285629272461, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.6149411201477051
it: 9700, train recon loss: 7.93587064743042, local kl: 0.0 global kl: 0.0001677232066867873 valid reconstr loss: -0.4941098392009735
it: 9800, train recon loss: -0.423115611076355, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.8752338886260986
it: 9900, train recon loss: 0.4798186719417572, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.6282520890235901
beta 1.0 temperature 0.5
it: 0, train recon loss: 1129.0802001953125, local kl: 0.0 global kl: 0.014980887994170189 valid reconstr loss: 254.24734497070312
Saving best model with reconstruction loss 254.24734
it: 100, train recon loss: 3.4555678367614746, local kl: 0.0 global kl: 0.0012475942494347692 valid reconstr loss: 3.6294286251068115
Saving best model with reconstruction loss 3.6294286
it: 200, train recon loss: 3.2552638053894043, local kl: 0.0 global kl: 0.0006869734497740865 valid reconstr loss: 3.3218066692352295
Saving best model with reconstruction loss 3.3218067
it: 300, train recon loss: 2.3610732555389404, local kl: 0.0 global kl: 0.0004234682710375637 valid reconstr loss: 3.036569118499756
Saving best model with reconstruction loss 3.036569
it: 400, train recon loss: 2.1715638637542725, local kl: 0.0 global kl: 0.0001676076208241284 valid reconstr loss: 2.1930973529815674
Saving best model with reconstruction loss 2.1930974
it: 500, train recon loss: 3.925016403198242, local kl: 0.0 global kl: 0.00024913204833865166 valid reconstr loss: 14.089076042175293
it: 600, train recon loss: 1.529703140258789, local kl: 0.0 global kl: 0.00020288288942538202 valid reconstr loss: 1.3094089031219482
Saving best model with reconstruction loss 1.3094089
it: 700, train recon loss: 0.6530607342720032, local kl: 0.0 global kl: 0.0002021001128014177 valid reconstr loss: 0.7522913813591003
Saving best model with reconstruction loss 0.7522914
it: 800, train recon loss: 0.7911478281021118, local kl: 0.0 global kl: 0.0013765920884907246 valid reconstr loss: 1.0439001321792603
it: 900, train recon loss: 3.40848970413208, local kl: 0.0 global kl: 0.0007393253035843372 valid reconstr loss: 3.3209540843963623
it: 1000, train recon loss: 0.3363281786441803, local kl: 0.0 global kl: 5.7812227169051766e-05 valid reconstr loss: 0.5102860331535339
Saving best model with reconstruction loss 0.51028603
it: 1100, train recon loss: 2.8727757930755615, local kl: 0.0 global kl: 0.0008478775853291154 valid reconstr loss: 1.0890583992004395
it: 1200, train recon loss: 0.448574960231781, local kl: 0.0 global kl: 2.159747418772895e-05 valid reconstr loss: 0.3678995668888092
Saving best model with reconstruction loss 0.36789957
it: 1300, train recon loss: 0.045851148664951324, local kl: 0.0 global kl: 0.0005202883039601147 valid reconstr loss: 0.5999642014503479
it: 1400, train recon loss: 187.1402587890625, local kl: 0.0 global kl: 6.895767728565261e-05 valid reconstr loss: 9.5277738571167
it: 1500, train recon loss: -0.14673273265361786, local kl: 0.0 global kl: 0.0004773714463226497 valid reconstr loss: 0.18631982803344727
Saving best model with reconstruction loss 0.18631983
it: 1600, train recon loss: 0.3485099673271179, local kl: 0.0 global kl: 6.512628169730306e-05 valid reconstr loss: 8.985310554504395
it: 1700, train recon loss: -0.026781082153320312, local kl: 0.0 global kl: 0.0002893697819672525 valid reconstr loss: -0.45997747778892517
Saving best model with reconstruction loss -0.45997748
it: 1800, train recon loss: -0.44244620203971863, local kl: 0.0 global kl: 0.000370002759154886 valid reconstr loss: -0.34277471899986267
it: 1900, train recon loss: -0.6798412203788757, local kl: 0.0 global kl: 0.0001377078442601487 valid reconstr loss: 0.7077277898788452
it: 2000, train recon loss: -0.5885560512542725, local kl: 0.0 global kl: 0.00020751202828250825 valid reconstr loss: -0.6673235893249512
Saving best model with reconstruction loss -0.6673236
it: 2100, train recon loss: -0.32398682832717896, local kl: 0.0 global kl: 0.00016889111429918557 valid reconstr loss: 2.480922222137451
it: 2200, train recon loss: -0.5811726450920105, local kl: 0.0 global kl: 0.0005456358194351196 valid reconstr loss: 36.524147033691406
it: 2300, train recon loss: 46.70093536376953, local kl: 0.0 global kl: 0.0002124294696841389 valid reconstr loss: 59.575870513916016
it: 2400, train recon loss: -0.7591351270675659, local kl: 0.0 global kl: 0.00031149075948633254 valid reconstr loss: -0.7354997396469116
Saving best model with reconstruction loss -0.73549974
it: 2500, train recon loss: -0.36595669388771057, local kl: 0.0 global kl: 0.00044654618250206113 valid reconstr loss: -0.7098800539970398
it: 2600, train recon loss: -0.7788435220718384, local kl: 0.0 global kl: 0.000833927362691611 valid reconstr loss: -0.8232929706573486
Saving best model with reconstruction loss -0.823293
it: 2700, train recon loss: -0.5798357129096985, local kl: 0.0 global kl: 3.935856148018502e-05 valid reconstr loss: -0.7501456141471863
it: 2800, train recon loss: -0.6538389921188354, local kl: 0.0 global kl: 2.9869243007851765e-05 valid reconstr loss: -0.8354576826095581
Saving best model with reconstruction loss -0.8354577
it: 2900, train recon loss: -0.6865988969802856, local kl: 0.0 global kl: 2.4775834390311502e-05 valid reconstr loss: -0.9548355937004089
Saving best model with reconstruction loss -0.9548356
it: 3000, train recon loss: -0.725492537021637, local kl: 0.0 global kl: 5.282154961605556e-05 valid reconstr loss: -0.8989008665084839
it: 3100, train recon loss: -0.9852608442306519, local kl: 0.0 global kl: 0.0009589944966137409 valid reconstr loss: -0.9641889929771423
Saving best model with reconstruction loss -0.964189
it: 3200, train recon loss: -0.26998189091682434, local kl: 0.0 global kl: 0.00010772743553388864 valid reconstr loss: -0.8725525140762329
it: 3300, train recon loss: -1.238003134727478, local kl: 0.0 global kl: 1.5820460248505697e-05 valid reconstr loss: -0.6519888639450073
it: 3400, train recon loss: 3113.2060546875, local kl: 0.0 global kl: 0.0002623764448799193 valid reconstr loss: -0.9414661526679993
it: 3500, train recon loss: -0.7792393565177917, local kl: 0.0 global kl: 4.53533539257478e-05 valid reconstr loss: -1.1090928316116333
Saving best model with reconstruction loss -1.1090928
it: 3600, train recon loss: -0.9419320225715637, local kl: 0.0 global kl: 0.00020703740301541984 valid reconstr loss: -0.9605777859687805
it: 3700, train recon loss: 19.220991134643555, local kl: 0.0 global kl: 0.00031115958699956536 valid reconstr loss: -0.9506422281265259
it: 3800, train recon loss: -1.0759035348892212, local kl: 0.0 global kl: 0.0001841510966187343 valid reconstr loss: -1.042438268661499
it: 3900, train recon loss: -1.3355575799942017, local kl: 0.0 global kl: 6.999741162871942e-05 valid reconstr loss: -0.9938684105873108
it: 4000, train recon loss: -1.5434048175811768, local kl: 0.0 global kl: 0.0003839277778752148 valid reconstr loss: -0.7536998987197876
it: 4100, train recon loss: 125.33075714111328, local kl: 0.0 global kl: 0.00041128956945613027 valid reconstr loss: -0.9488534927368164
it: 4200, train recon loss: -0.8208073377609253, local kl: 0.0 global kl: 9.908544598147273e-05 valid reconstr loss: -0.8660991191864014
it: 4300, train recon loss: -1.1492831707000732, local kl: 0.0 global kl: 1.5122900549613405e-05 valid reconstr loss: -1.08568274974823
it: 4400, train recon loss: -1.1147828102111816, local kl: 0.0 global kl: 0.0005612287204712629 valid reconstr loss: -1.0870566368103027
it: 4500, train recon loss: -1.1628597974777222, local kl: 0.0 global kl: 0.00033786348649300635 valid reconstr loss: -1.1018413305282593
it: 4600, train recon loss: 5.522411823272705, local kl: 0.0 global kl: 0.0032026071567088366 valid reconstr loss: 6747.9638671875
it: 4700, train recon loss: -1.0763421058654785, local kl: 0.0 global kl: 0.0004865627270191908 valid reconstr loss: -1.0744669437408447
it: 4800, train recon loss: -1.4659733772277832, local kl: 0.0 global kl: 0.00019049705588258803 valid reconstr loss: -1.046142339706421
it: 4900, train recon loss: -1.285244345664978, local kl: 0.0 global kl: 0.0003139916807413101 valid reconstr loss: -0.7448738217353821
it: 5000, train recon loss: -1.2306294441223145, local kl: 0.0 global kl: 0.0002705072402022779 valid reconstr loss: -1.1980596780776978
Saving best model with reconstruction loss -1.1980597
it: 5100, train recon loss: -1.0535025596618652, local kl: 0.0 global kl: 0.001620686613023281 valid reconstr loss: -1.1787848472595215
it: 5200, train recon loss: 118.76822662353516, local kl: 0.0 global kl: 0.03566519170999527 valid reconstr loss: 62887.453125
it: 5300, train recon loss: 19.28758430480957, local kl: 0.0 global kl: 0.001093005994334817 valid reconstr loss: 1.7659231424331665
it: 5400, train recon loss: -1.1503205299377441, local kl: 0.0 global kl: 5.157796749699628e-06 valid reconstr loss: -0.9992532730102539
it: 5500, train recon loss: -1.3987990617752075, local kl: 0.0 global kl: 2.610987939988263e-05 valid reconstr loss: -1.1507344245910645
it: 5600, train recon loss: -0.7852345705032349, local kl: 0.0 global kl: 1.6729292838135734e-05 valid reconstr loss: -1.2362595796585083
Saving best model with reconstruction loss -1.2362596
it: 5700, train recon loss: -0.8484718203544617, local kl: 0.0 global kl: 7.50212202547118e-05 valid reconstr loss: -1.2490283250808716
Saving best model with reconstruction loss -1.2490283
it: 5800, train recon loss: 1.9960733652114868, local kl: 0.0 global kl: 0.023314302787184715 valid reconstr loss: 1.011673092842102
it: 5900, train recon loss: 0.5840165019035339, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.8284039497375488
it: 6000, train recon loss: -1.0309088230133057, local kl: 0.0 global kl: 0.00011998364789178595 valid reconstr loss: 28.034515380859375
it: 6100, train recon loss: -0.5737055540084839, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.48454535007476807
it: 6200, train recon loss: -0.5172519683837891, local kl: 0.0 global kl: 7.40318384373495e-08 valid reconstr loss: -0.8033437728881836
it: 6300, train recon loss: -0.9101806879043579, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.7821019291877747
it: 6400, train recon loss: 8474.51171875, local kl: 0.0 global kl: 5.017035437049344e-05 valid reconstr loss: 1.2634060382843018
it: 6500, train recon loss: -0.5688304305076599, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.522108793258667
it: 6600, train recon loss: -0.6336328983306885, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.9865105152130127
it: 6700, train recon loss: -0.82908034324646, local kl: 0.0 global kl: 5.86690384807298e-06 valid reconstr loss: -0.14634360373020172
it: 6800, train recon loss: -0.7896081209182739, local kl: 0.0 global kl: 0.000477153982501477 valid reconstr loss: 0.5328181982040405
it: 6900, train recon loss: -0.7317001223564148, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.5939544439315796
it: 7000, train recon loss: -0.0445430688560009, local kl: 0.0 global kl: 0.00030344826518557966 valid reconstr loss: 387.6836853027344
it: 7100, train recon loss: -0.938530445098877, local kl: 0.0 global kl: 0.00024888027110137045 valid reconstr loss: -0.9475477933883667
it: 7200, train recon loss: 564.211181640625, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.36165523529052734
it: 7300, train recon loss: -0.6327887773513794, local kl: 0.0 global kl: 9.474238140683156e-06 valid reconstr loss: 281.8141174316406
it: 7400, train recon loss: -0.5609174966812134, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.8006977438926697
it: 7500, train recon loss: -1.125401496887207, local kl: 0.0 global kl: 0.00014603209274355322 valid reconstr loss: -0.9295569658279419
it: 7600, train recon loss: -1.27773118019104, local kl: 0.0 global kl: 0.00011755493324017152 valid reconstr loss: -1.0457080602645874
it: 7700, train recon loss: -1.2883989810943604, local kl: 0.0 global kl: 2.1675654693353863e-07 valid reconstr loss: -1.1165062189102173
it: 7800, train recon loss: -1.1481602191925049, local kl: 0.0 global kl: 8.05030285846442e-05 valid reconstr loss: -1.1228766441345215
it: 7900, train recon loss: -1.085559606552124, local kl: 0.0 global kl: 0.0003959588357247412 valid reconstr loss: 471.72137451171875
it: 8000, train recon loss: -1.0307940244674683, local kl: 0.0 global kl: 8.753091969992965e-05 valid reconstr loss: -1.2385469675064087
it: 8100, train recon loss: -0.9911049604415894, local kl: 0.0 global kl: 5.588014755630866e-05 valid reconstr loss: -0.1710241436958313
it: 8200, train recon loss: -1.6145988702774048, local kl: 0.0 global kl: 0.0022742333821952343 valid reconstr loss: -0.43820399045944214
it: 8300, train recon loss: -0.8051150441169739, local kl: 0.0 global kl: 0.005579798482358456 valid reconstr loss: -0.8164650797843933
it: 8400, train recon loss: -1.2756850719451904, local kl: 0.0 global kl: 7.110272417776287e-05 valid reconstr loss: -1.1560041904449463
it: 8500, train recon loss: -1.153098702430725, local kl: 0.0 global kl: 0.00032940489472821355 valid reconstr loss: -1.2223801612854004
it: 8600, train recon loss: -0.6665204763412476, local kl: 0.0 global kl: 0.0003994639846496284 valid reconstr loss: -0.7132850289344788
it: 8700, train recon loss: -0.5002624988555908, local kl: 0.0 global kl: 0.00039269239641726017 valid reconstr loss: -0.06137649714946747
it: 8800, train recon loss: 16.363540649414062, local kl: 0.0 global kl: 0.0021240573842078447 valid reconstr loss: 8802.9013671875
it: 8900, train recon loss: 1.0423054695129395, local kl: 0.0 global kl: 0.0028259586542844772 valid reconstr loss: 0.6997091770172119
it: 9000, train recon loss: -0.9915813207626343, local kl: 0.0 global kl: 4.491228173719719e-05 valid reconstr loss: -1.061205267906189
it: 9100, train recon loss: -0.9573596715927124, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.0690175294876099
it: 9200, train recon loss: -0.9604341983795166, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.6768127679824829
it: 9300, train recon loss: -0.9241704940795898, local kl: 0.0 global kl: 4.474308298085816e-05 valid reconstr loss: -1.2750977277755737
Saving best model with reconstruction loss -1.2750977
it: 9400, train recon loss: -0.10208382457494736, local kl: 0.0 global kl: 2.768425110843964e-05 valid reconstr loss: -1.1581108570098877
it: 9500, train recon loss: 53.728660583496094, local kl: 0.0 global kl: 0.0002405919076409191 valid reconstr loss: 0.03282240033149719
it: 9600, train recon loss: -1.3102538585662842, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.0537883043289185
it: 9700, train recon loss: 3.5701088905334473, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.7968258857727051
it: 9800, train recon loss: 3.6936612129211426, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.308840274810791
it: 9900, train recon loss: 2.944903612136841, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.153585910797119
beta 1.0 temperature 1.0
it: 0, train recon loss: 370620.96875, local kl: 0.0 global kl: 0.00783311016857624 valid reconstr loss: 2319.78466796875
Saving best model with reconstruction loss 2319.7847
it: 100, train recon loss: 4.046654224395752, local kl: 0.0 global kl: 0.0009707122226245701 valid reconstr loss: 4.055088996887207
Saving best model with reconstruction loss 4.055089
it: 200, train recon loss: 3.945039987564087, local kl: 0.0 global kl: 8.84831533767283e-05 valid reconstr loss: 3.944751024246216
Saving best model with reconstruction loss 3.944751
it: 300, train recon loss: 10.300638198852539, local kl: 0.0 global kl: 3.205409302609041e-05 valid reconstr loss: 3.217543125152588
Saving best model with reconstruction loss 3.2175431
it: 400, train recon loss: 3.0309574604034424, local kl: 0.0 global kl: 4.8953061195788905e-05 valid reconstr loss: 2.954047679901123
Saving best model with reconstruction loss 2.9540477
it: 500, train recon loss: 2.4811830520629883, local kl: 0.0 global kl: 8.044650894589722e-05 valid reconstr loss: 3.86807918548584
it: 600, train recon loss: 3.71474552154541, local kl: 0.0 global kl: 0.00019097464974038303 valid reconstr loss: 5.325711727142334
it: 700, train recon loss: 2.234969139099121, local kl: 0.0 global kl: 0.0006664348184131086 valid reconstr loss: 2.1665124893188477
Saving best model with reconstruction loss 2.1665125
it: 800, train recon loss: 12.599808692932129, local kl: 0.0 global kl: 0.002347758039832115 valid reconstr loss: 11.506266593933105
it: 900, train recon loss: 1.513597846031189, local kl: 0.0 global kl: 2.880846477637533e-05 valid reconstr loss: 1.4654918909072876
Saving best model with reconstruction loss 1.4654919
it: 1000, train recon loss: 1.2126599550247192, local kl: 0.0 global kl: 1.3183505870983936e-05 valid reconstr loss: 1.8265678882598877
it: 1100, train recon loss: 1.377699851989746, local kl: 0.0 global kl: 0.00013406365178525448 valid reconstr loss: 2.1085612773895264
it: 1200, train recon loss: 8.894681930541992, local kl: 0.0 global kl: 9.372221029479988e-06 valid reconstr loss: 1.8212268352508545
it: 1300, train recon loss: 1.2821288108825684, local kl: 0.0 global kl: 4.4791275286115706e-05 valid reconstr loss: 1.0569701194763184
Saving best model with reconstruction loss 1.0569701
it: 1400, train recon loss: 0.6723729968070984, local kl: 0.0 global kl: 0.00012303286348469555 valid reconstr loss: 1.0005673170089722
Saving best model with reconstruction loss 1.0005673
it: 1500, train recon loss: 0.6056379675865173, local kl: 0.0 global kl: 1.5138834896788467e-05 valid reconstr loss: 0.7579613327980042
Saving best model with reconstruction loss 0.75796133
it: 1600, train recon loss: 0.6693468689918518, local kl: 0.0 global kl: 2.1443118384922855e-05 valid reconstr loss: 107.16433715820312
it: 1700, train recon loss: 0.3857056200504303, local kl: 0.0 global kl: 3.7691512261517346e-05 valid reconstr loss: 0.384578138589859
Saving best model with reconstruction loss 0.38457814
it: 1800, train recon loss: 0.566472589969635, local kl: 0.0 global kl: 1.3315075193531811e-05 valid reconstr loss: 0.09555758535861969
Saving best model with reconstruction loss 0.095557585
it: 1900, train recon loss: -0.2878282368183136, local kl: 0.0 global kl: 4.2825035052374005e-05 valid reconstr loss: 0.33740395307540894
it: 2000, train recon loss: -0.08981335908174515, local kl: 0.0 global kl: 6.353297794703394e-05 valid reconstr loss: 0.1348492056131363
it: 2100, train recon loss: -0.2843274176120758, local kl: 0.0 global kl: 0.0002112912479788065 valid reconstr loss: -0.239468514919281
Saving best model with reconstruction loss -0.23946851
it: 2200, train recon loss: 16.709381103515625, local kl: 0.0 global kl: 0.0004219244292471558 valid reconstr loss: 2.0844597816467285
it: 2300, train recon loss: -0.6828799843788147, local kl: 0.0 global kl: 0.00020170742936898023 valid reconstr loss: 0.741460382938385
it: 2400, train recon loss: -0.07048056274652481, local kl: 0.0 global kl: 4.353016265667975e-05 valid reconstr loss: -0.281479150056839
Saving best model with reconstruction loss -0.28147915
it: 2500, train recon loss: -0.12989895045757294, local kl: 0.0 global kl: 2.737044087552931e-05 valid reconstr loss: -0.3851625621318817
Saving best model with reconstruction loss -0.38516256
it: 2600, train recon loss: -0.4249035120010376, local kl: 0.0 global kl: 2.709442014747765e-05 valid reconstr loss: -0.4294317364692688
Saving best model with reconstruction loss -0.42943174
it: 2700, train recon loss: -0.5235800743103027, local kl: 0.0 global kl: 0.0006768596940673888 valid reconstr loss: 0.24751611053943634
it: 2800, train recon loss: -0.5241464376449585, local kl: 0.0 global kl: 3.9772898162482306e-05 valid reconstr loss: -0.3925830125808716
it: 2900, train recon loss: -0.6174306273460388, local kl: 0.0 global kl: 3.506465509417467e-05 valid reconstr loss: -0.31753161549568176
it: 3000, train recon loss: -0.3322027623653412, local kl: 0.0 global kl: 0.00028318463591858745 valid reconstr loss: -0.7260560393333435
Saving best model with reconstruction loss -0.72605604
it: 3100, train recon loss: -0.9264335036277771, local kl: 0.0 global kl: 1.8325799828744493e-05 valid reconstr loss: -0.19015508890151978
it: 3200, train recon loss: 28.206350326538086, local kl: 0.0 global kl: 0.00035609310725703835 valid reconstr loss: -0.3154362142086029
it: 3300, train recon loss: -0.44621479511260986, local kl: 0.0 global kl: 2.2113756131147966e-05 valid reconstr loss: -0.8468708395957947
Saving best model with reconstruction loss -0.84687084
it: 3400, train recon loss: -0.7007856965065002, local kl: 0.0 global kl: 3.416320396354422e-05 valid reconstr loss: -0.05909207835793495
it: 3500, train recon loss: -0.7890440225601196, local kl: 0.0 global kl: 2.1906591427978128e-05 valid reconstr loss: -0.627714991569519
it: 3600, train recon loss: -0.631976306438446, local kl: 0.0 global kl: 0.00022794473625253886 valid reconstr loss: -0.7270151376724243
it: 3700, train recon loss: -0.7899155020713806, local kl: 0.0 global kl: 5.975236945232609e-06 valid reconstr loss: 40.87138748168945
it: 3800, train recon loss: -0.6321081519126892, local kl: 0.0 global kl: 8.63617333379807e-06 valid reconstr loss: 0.10935284197330475
it: 3900, train recon loss: 2.9639439582824707, local kl: 0.0 global kl: 0.00026994984364137053 valid reconstr loss: 2.792144536972046
it: 4000, train recon loss: 1.6418557167053223, local kl: 0.0 global kl: 1.7016043784678914e-05 valid reconstr loss: 3.5156900882720947
it: 4100, train recon loss: 1.622036337852478, local kl: 0.0 global kl: 8.868204895406961e-05 valid reconstr loss: 1.5945557355880737
it: 4200, train recon loss: 1.380090594291687, local kl: 0.0 global kl: 2.710490116442088e-05 valid reconstr loss: 2.0421159267425537
it: 4300, train recon loss: 1.2374192476272583, local kl: 0.0 global kl: 0.00020589923951774836 valid reconstr loss: 1.3766751289367676
it: 4400, train recon loss: 0.9812541007995605, local kl: 0.0 global kl: 4.040844942210242e-05 valid reconstr loss: 0.730302631855011
it: 4500, train recon loss: 0.6266794800758362, local kl: 0.0 global kl: 2.2116799300420098e-05 valid reconstr loss: 0.7825549244880676
it: 4600, train recon loss: 0.46686261892318726, local kl: 0.0 global kl: 0.0012629888951778412 valid reconstr loss: 0.7440236806869507
it: 4700, train recon loss: 1.2724229097366333, local kl: 0.0 global kl: 8.091831841738895e-05 valid reconstr loss: 0.33320721983909607
it: 4800, train recon loss: 0.16036678850650787, local kl: 0.0 global kl: 7.668063335586339e-05 valid reconstr loss: 0.3135402500629425
it: 4900, train recon loss: 0.3112333416938782, local kl: 0.0 global kl: 3.880060830852017e-06 valid reconstr loss: 3.660628080368042
it: 5000, train recon loss: 0.45478737354278564, local kl: 0.0 global kl: 2.704198777792044e-05 valid reconstr loss: 0.2621484398841858
it: 5100, train recon loss: -0.02545146644115448, local kl: 0.0 global kl: 0.0003257674106862396 valid reconstr loss: 20144.26171875
it: 5200, train recon loss: -0.034919366240501404, local kl: 0.0 global kl: 1.6965239410637878e-05 valid reconstr loss: 0.042245298624038696
it: 5300, train recon loss: -0.01999385468661785, local kl: 0.0 global kl: 5.8297289797337726e-05 valid reconstr loss: 0.25275760889053345
it: 5400, train recon loss: 0.22493121027946472, local kl: 0.0 global kl: 5.499479448189959e-05 valid reconstr loss: 0.5696199536323547
it: 5500, train recon loss: -0.3229484260082245, local kl: 0.0 global kl: 8.094316399365198e-06 valid reconstr loss: 0.1492711901664734
it: 5600, train recon loss: 13.529253005981445, local kl: 0.0 global kl: 5.720528861274943e-06 valid reconstr loss: 7837.697265625
it: 5700, train recon loss: 0.47700777649879456, local kl: 0.0 global kl: 3.070623642997816e-05 valid reconstr loss: -0.15489043295383453
it: 5800, train recon loss: 0.7976108193397522, local kl: 0.0 global kl: 9.380239134770818e-06 valid reconstr loss: -0.19997814297676086
it: 5900, train recon loss: 19408.283203125, local kl: 0.0 global kl: 9.127163139055483e-06 valid reconstr loss: 25.79475975036621
it: 6000, train recon loss: -0.42995062470436096, local kl: 0.0 global kl: 3.454547913861461e-05 valid reconstr loss: 2.234743356704712
it: 6100, train recon loss: -0.16199617087841034, local kl: 0.0 global kl: 3.833022128674202e-05 valid reconstr loss: 0.5357853174209595
it: 6200, train recon loss: -0.06238602474331856, local kl: 0.0 global kl: 3.7431632904372236e-07 valid reconstr loss: -0.2974281311035156
it: 6300, train recon loss: 1456480.75, local kl: 0.0 global kl: 0.025476420298218727 valid reconstr loss: 2296.16796875
it: 6400, train recon loss: 0.24197553098201752, local kl: 0.0 global kl: 0.000189418627996929 valid reconstr loss: -0.09238851815462112
it: 6500, train recon loss: 0.017796097323298454, local kl: 0.0 global kl: 2.7631729153654305e-06 valid reconstr loss: -0.2224651277065277
it: 6600, train recon loss: -0.03091750293970108, local kl: 0.0 global kl: 2.4497416234225966e-05 valid reconstr loss: -0.11087999492883682
it: 6700, train recon loss: 6040.8447265625, local kl: 0.0 global kl: 4.79711961816065e-05 valid reconstr loss: 107.60043334960938
it: 6800, train recon loss: -0.5977137088775635, local kl: 0.0 global kl: 0.00014786070096306503 valid reconstr loss: -0.27563273906707764
it: 6900, train recon loss: 0.2814483642578125, local kl: 0.0 global kl: 3.9192300391732715e-06 valid reconstr loss: 0.700715184211731
it: 7000, train recon loss: 0.13194656372070312, local kl: 0.0 global kl: 1.5311215975088999e-06 valid reconstr loss: -0.34112548828125
it: 7100, train recon loss: 0.047332651913166046, local kl: 0.0 global kl: 4.225171869620681e-05 valid reconstr loss: 0.3472859561443329
it: 7200, train recon loss: -0.38365113735198975, local kl: 0.0 global kl: 6.353820208460093e-05 valid reconstr loss: -0.480564147233963
it: 7300, train recon loss: -0.26866981387138367, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.29198917746543884
it: 7400, train recon loss: 13.410037994384766, local kl: 0.0 global kl: 0.0001467377005610615 valid reconstr loss: 0.8888105154037476
it: 7500, train recon loss: -0.2673024535179138, local kl: 0.0 global kl: 5.383655661717057e-06 valid reconstr loss: -0.40555745363235474
it: 7600, train recon loss: 1.1507724523544312, local kl: 0.0 global kl: 0.000291967939119786 valid reconstr loss: 1.835661768913269
it: 7700, train recon loss: 1352.483154296875, local kl: 0.0 global kl: 0.0003567120584193617 valid reconstr loss: 0.5723719596862793
it: 7800, train recon loss: 0.1730211079120636, local kl: 0.0 global kl: 7.831533821445191e-08 valid reconstr loss: 0.7657768130302429
it: 7900, train recon loss: -0.08661980926990509, local kl: 0.0 global kl: 0.0006270912708714604 valid reconstr loss: 0.06507714837789536
it: 8000, train recon loss: -0.04401359334588051, local kl: 0.0 global kl: 5.453855465020752e-06 valid reconstr loss: 0.12644584476947784
it: 8100, train recon loss: -0.5927887558937073, local kl: 0.0 global kl: 7.64660086360891e-08 valid reconstr loss: 0.6893790364265442
it: 8200, train recon loss: -0.13474223017692566, local kl: 0.0 global kl: 0.00019285142479930073 valid reconstr loss: 5272.44384765625
it: 8300, train recon loss: -0.10834066569805145, local kl: 0.0 global kl: 7.875873961893376e-07 valid reconstr loss: -0.13792334496974945
it: 8400, train recon loss: 2161.2275390625, local kl: 0.0 global kl: 0.0 valid reconstr loss: 564.584716796875
it: 8500, train recon loss: 3.3642847537994385, local kl: 0.0 global kl: 1.2319103070979054e-08 valid reconstr loss: 4.109925746917725
it: 8600, train recon loss: 4.848286151885986, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.495053768157959
it: 8700, train recon loss: 3.3144381046295166, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.2764289379119873
it: 8800, train recon loss: 2.9221444129943848, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.572000026702881
it: 8900, train recon loss: 3.087050199508667, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.2980010509490967
it: 9000, train recon loss: 2.416187286376953, local kl: 0.0 global kl: 4.255440580891445e-05 valid reconstr loss: 2.7163469791412354
it: 9100, train recon loss: 1.71451997756958, local kl: 0.0 global kl: 0.0 valid reconstr loss: 37.936153411865234
it: 9200, train recon loss: 1.7260653972625732, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.559996485710144
it: 9300, train recon loss: 1.0965702533721924, local kl: 0.0 global kl: 1.634066029509995e-05 valid reconstr loss: 1.3220534324645996
it: 9400, train recon loss: 0.9750440120697021, local kl: 0.0 global kl: 0.0 valid reconstr loss: 99.2410659790039
it: 9500, train recon loss: 0.3777863681316376, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.11382491141557693
it: 9600, train recon loss: 0.0639866515994072, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.2366989105939865
it: 9700, train recon loss: 0.13649983704090118, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.22502991557121277
it: 9800, train recon loss: -0.039762381464242935, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.35579684376716614
it: 9900, train recon loss: 194.968994140625, local kl: 0.0 global kl: 5.569132099481067e-06 valid reconstr loss: 6348.5458984375
beta 1.0 temperature 2.0
it: 0, train recon loss: 231.4581298828125, local kl: 0.0 global kl: 0.006863579154014587 valid reconstr loss: 80.35620880126953
Saving best model with reconstruction loss 80.35621
it: 100, train recon loss: 3.8091795444488525, local kl: 0.0 global kl: 0.000987629173323512 valid reconstr loss: 4.005043029785156
Saving best model with reconstruction loss 4.005043
it: 200, train recon loss: 4.185156345367432, local kl: 0.0 global kl: 8.592885569669306e-05 valid reconstr loss: 4.203921318054199
it: 300, train recon loss: 3.08107590675354, local kl: 0.0 global kl: 0.000762640091124922 valid reconstr loss: 4.691854476928711
it: 400, train recon loss: 2.9760780334472656, local kl: 0.0 global kl: 0.0005180186708457768 valid reconstr loss: 2.8787662982940674
Saving best model with reconstruction loss 2.8787663
it: 500, train recon loss: 1.9486503601074219, local kl: 0.0 global kl: 0.000246420648181811 valid reconstr loss: 1.9157130718231201
Saving best model with reconstruction loss 1.9157131
it: 600, train recon loss: 1.9858126640319824, local kl: 0.0 global kl: 7.965211823446339e-12 valid reconstr loss: 2.0242185592651367
it: 700, train recon loss: 1.6498960256576538, local kl: 0.0 global kl: 1.0533796057643485e-12 valid reconstr loss: 2.032074451446533
it: 800, train recon loss: 0.9677436947822571, local kl: 0.0 global kl: 1.6068868458063434e-11 valid reconstr loss: 1.7271076440811157
Saving best model with reconstruction loss 1.7271076
it: 900, train recon loss: 1.78170907497406, local kl: 0.0 global kl: 1.6879726782992321e-13 valid reconstr loss: 1.1894382238388062
Saving best model with reconstruction loss 1.1894382
it: 1000, train recon loss: 1.3758758306503296, local kl: 0.0 global kl: 1.5746848269770908e-12 valid reconstr loss: 0.9946359992027283
Saving best model with reconstruction loss 0.994636
it: 1100, train recon loss: 0.5161868929862976, local kl: 0.0 global kl: 1.0417226109504796e-11 valid reconstr loss: 0.6238239407539368
Saving best model with reconstruction loss 0.62382394
it: 1200, train recon loss: 0.26653853058815, local kl: 0.0 global kl: 3.9022118869525e-12 valid reconstr loss: 0.6345577239990234
it: 1300, train recon loss: 0.7046204209327698, local kl: 0.0 global kl: 8.042455590384634e-13 valid reconstr loss: 0.7225190997123718
it: 1400, train recon loss: 0.20102201402187347, local kl: 0.0 global kl: 1.9807558371276457e-12 valid reconstr loss: 0.33021092414855957
Saving best model with reconstruction loss 0.33021092
it: 1500, train recon loss: 2.075324296951294, local kl: 0.0 global kl: 5.485611964672898e-13 valid reconstr loss: 0.7553675174713135
it: 1600, train recon loss: 0.5709726214408875, local kl: 0.0 global kl: 1.6427484372805168e-12 valid reconstr loss: 0.19918188452720642
Saving best model with reconstruction loss 0.19918188
it: 1700, train recon loss: -0.29312998056411743, local kl: 0.0 global kl: 1.897032878006577e-12 valid reconstr loss: 71.70881652832031
it: 1800, train recon loss: 750.367919921875, local kl: 0.0 global kl: 2.6707525080382766e-11 valid reconstr loss: 64.76309204101562
it: 1900, train recon loss: -0.060704175382852554, local kl: 0.0 global kl: 8.705813847598165e-13 valid reconstr loss: -0.10589181631803513
Saving best model with reconstruction loss -0.10589182
it: 2000, train recon loss: -0.07823292911052704, local kl: 0.0 global kl: 3.703294615409192e-12 valid reconstr loss: -0.10267820954322815
it: 2100, train recon loss: -0.3619300425052643, local kl: 0.0 global kl: 8.698153308728251e-12 valid reconstr loss: 213.66331481933594
it: 2200, train recon loss: -0.35386326909065247, local kl: 0.0 global kl: 8.978651155899797e-13 valid reconstr loss: 0.03352969512343407
it: 2300, train recon loss: -0.5347279906272888, local kl: 0.0 global kl: 2.4520080410539435e-11 valid reconstr loss: 1467.431884765625
it: 2400, train recon loss: 484.8941955566406, local kl: 0.0 global kl: 1.0950573781087769e-11 valid reconstr loss: 187.29481506347656
it: 2500, train recon loss: -0.3173994719982147, local kl: 0.0 global kl: 2.805285517770706e-12 valid reconstr loss: -0.18109750747680664
Saving best model with reconstruction loss -0.18109751
it: 2600, train recon loss: 23.174341201782227, local kl: 0.0 global kl: 1.8852697181159783e-12 valid reconstr loss: 0.023298004642128944
it: 2700, train recon loss: -0.18576134741306305, local kl: 0.0 global kl: 4.799452935771997e-12 valid reconstr loss: 1.1498435735702515
it: 2800, train recon loss: -0.8147974610328674, local kl: 0.0 global kl: 3.7132536628847745e-13 valid reconstr loss: 3.7244722843170166
it: 2900, train recon loss: -0.4913342595100403, local kl: 0.0 global kl: 6.095741966749557e-12 valid reconstr loss: 0.1166103258728981
it: 3000, train recon loss: -0.3858395218849182, local kl: 0.0 global kl: 9.350853424905381e-14 valid reconstr loss: -0.5745912790298462
Saving best model with reconstruction loss -0.5745913
it: 3100, train recon loss: -0.5507607460021973, local kl: 0.0 global kl: 1.223425874496975e-12 valid reconstr loss: -0.5004168748855591
it: 3200, train recon loss: 0.9177601933479309, local kl: 0.0 global kl: 3.225475442292236e-13 valid reconstr loss: -0.48393112421035767
it: 3300, train recon loss: -0.7820187211036682, local kl: 0.0 global kl: 2.1871393585115584e-14 valid reconstr loss: 0.009668395854532719
it: 3400, train recon loss: -0.7657209634780884, local kl: 0.0 global kl: 0.0 valid reconstr loss: 112.40817260742188
it: 3500, train recon loss: 30.986902236938477, local kl: 0.0 global kl: 2.6244067682923422e-11 valid reconstr loss: -0.10043134540319443
it: 3600, train recon loss: -0.7041329741477966, local kl: 0.0 global kl: 9.323791738680143e-14 valid reconstr loss: -0.5493743419647217
it: 3700, train recon loss: -0.8256043195724487, local kl: 0.0 global kl: 3.845618268272233e-12 valid reconstr loss: -0.6411857008934021
Saving best model with reconstruction loss -0.6411857
it: 3800, train recon loss: -0.47814077138900757, local kl: 0.0 global kl: 6.9610428532485e-12 valid reconstr loss: -0.8054947257041931
Saving best model with reconstruction loss -0.8054947
it: 3900, train recon loss: -0.8081845641136169, local kl: 0.0 global kl: 7.241429678117584e-14 valid reconstr loss: -0.33819201588630676
it: 4000, train recon loss: -1.1736149787902832, local kl: 0.0 global kl: 9.039990978010337e-14 valid reconstr loss: 1061.5347900390625
it: 4100, train recon loss: -0.6602677702903748, local kl: 0.0 global kl: 4.647715182550319e-14 valid reconstr loss: -0.7646651268005371
it: 4200, train recon loss: -0.41855835914611816, local kl: 0.0 global kl: 2.495781359357352e-13 valid reconstr loss: -0.8037529587745667
it: 4300, train recon loss: 177.77890014648438, local kl: 0.0 global kl: 8.576472865229334e-14 valid reconstr loss: 22.220935821533203
it: 4400, train recon loss: -0.8417769074440002, local kl: 0.0 global kl: 2.5385632831942395e-11 valid reconstr loss: 0.0814364105463028
it: 4500, train recon loss: -0.854798436164856, local kl: 0.0 global kl: 2.504727653583616e-13 valid reconstr loss: -0.8543745875358582
Saving best model with reconstruction loss -0.8543746
it: 4600, train recon loss: -0.25885605812072754, local kl: 0.0 global kl: 7.406368920936934e-12 valid reconstr loss: -0.8300766348838806
it: 4700, train recon loss: -0.9016761183738708, local kl: 0.0 global kl: 1.199040866595169e-14 valid reconstr loss: -0.8119220733642578
it: 4800, train recon loss: -1.3057087659835815, local kl: 0.0 global kl: 0.0 valid reconstr loss: 8.15732192993164
it: 4900, train recon loss: -0.9578191041946411, local kl: 0.0 global kl: 4.379477336335924e-11 valid reconstr loss: -0.8008915781974792
it: 5000, train recon loss: 5.7269415855407715, local kl: 0.0 global kl: 4.890950491831525e-12 valid reconstr loss: 5.746562480926514
it: 5100, train recon loss: -0.9506620764732361, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.34131017327308655
it: 5200, train recon loss: 374875.53125, local kl: 0.0 global kl: 5.6244509050173974e-11 valid reconstr loss: -0.22002175450325012
it: 5300, train recon loss: -0.974804699420929, local kl: 0.0 global kl: 4.172585887918245e-12 valid reconstr loss: -0.7330878376960754
it: 5400, train recon loss: 131.4029998779297, local kl: 0.0 global kl: 3.844537535546699e-14 valid reconstr loss: -0.41906610131263733
it: 5500, train recon loss: -1.1424846649169922, local kl: 0.0 global kl: 9.552053592543075e-12 valid reconstr loss: 6.459362030029297
it: 5600, train recon loss: -0.6268808245658875, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.6370699405670166
it: 5700, train recon loss: 1543.6510009765625, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.15261800587177277
it: 5800, train recon loss: -1.0511136054992676, local kl: 0.0 global kl: 9.872658246479205e-12 valid reconstr loss: -0.9656263589859009
Saving best model with reconstruction loss -0.96562636
it: 5900, train recon loss: 0.6857871413230896, local kl: 0.0 global kl: 4.70613131797748e-14 valid reconstr loss: -1.1324232816696167
Saving best model with reconstruction loss -1.1324233
it: 6000, train recon loss: -1.256562352180481, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.9563120007514954
it: 6100, train recon loss: -0.9142395257949829, local kl: 0.0 global kl: 1.3184592306814125e-13 valid reconstr loss: 180.5493927001953
it: 6200, train recon loss: -1.1533535718917847, local kl: 0.0 global kl: 8.484719871137969e-12 valid reconstr loss: -0.5816130638122559
it: 6300, train recon loss: -1.0398695468902588, local kl: 0.0 global kl: 3.882810739597176e-12 valid reconstr loss: -0.6517542004585266
it: 6400, train recon loss: 51.957542419433594, local kl: 0.0 global kl: 3.771722517642573e-15 valid reconstr loss: 8.549768447875977
it: 6500, train recon loss: -1.0263532400131226, local kl: 0.0 global kl: 4.163336342344337e-16 valid reconstr loss: 5.615143299102783
it: 6600, train recon loss: -1.0187331438064575, local kl: 0.0 global kl: 3.379414803550418e-14 valid reconstr loss: 6.998425483703613
it: 6700, train recon loss: 6011.4482421875, local kl: 0.0 global kl: 3.29625216011209e-13 valid reconstr loss: 3483.854736328125
it: 6800, train recon loss: -1.1790786981582642, local kl: 0.0 global kl: 4.605482661901306e-13 valid reconstr loss: -0.9837039709091187
it: 6900, train recon loss: -0.9246975779533386, local kl: 0.0 global kl: 5.6118738128718704e-14 valid reconstr loss: 5814.69580078125
it: 7000, train recon loss: -0.2030414491891861, local kl: 0.0 global kl: 3.145012028582528e-12 valid reconstr loss: -1.2030949592590332
Saving best model with reconstruction loss -1.203095
it: 7100, train recon loss: -0.8833805322647095, local kl: 0.0 global kl: 1.8385071243187667e-11 valid reconstr loss: -0.2162032574415207
it: 7200, train recon loss: 0.7566622495651245, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.5469518899917603
it: 7300, train recon loss: -0.916418194770813, local kl: 0.0 global kl: 1.732641807805635e-14 valid reconstr loss: -1.135364294052124
it: 7400, train recon loss: -0.5484220385551453, local kl: 0.0 global kl: 4.0368403064761083e-13 valid reconstr loss: -1.0569193363189697
it: 7500, train recon loss: -0.9986667037010193, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.081476092338562
it: 7600, train recon loss: -1.1072016954421997, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.7648978233337402
it: 7700, train recon loss: -1.0289174318313599, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.9953919649124146
it: 7800, train recon loss: -0.9010134339332581, local kl: 0.0 global kl: 0.0 valid reconstr loss: 30.96285629272461
it: 7900, train recon loss: -0.9672262072563171, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.0766226053237915
it: 8000, train recon loss: 114.75077819824219, local kl: 0.0 global kl: 7.429097337308121e-10 valid reconstr loss: 142.8585662841797
it: 8100, train recon loss: 38.03995132446289, local kl: 0.0 global kl: 4.000888509381184e-10 valid reconstr loss: 51.375431060791016
it: 8200, train recon loss: 16.773168563842773, local kl: 0.0 global kl: 8.970335585445355e-10 valid reconstr loss: 24.67294692993164
it: 8300, train recon loss: 11.28787612915039, local kl: 0.0 global kl: 9.945342327455364e-10 valid reconstr loss: 14.674543380737305
it: 8400, train recon loss: 8.606243133544922, local kl: 0.0 global kl: 2.2061517057636593e-09 valid reconstr loss: 9.878843307495117
it: 8500, train recon loss: 7.1320414543151855, local kl: 0.0 global kl: 2.0954971091668995e-09 valid reconstr loss: 7.2851715087890625
it: 8600, train recon loss: 5.397130012512207, local kl: 0.0 global kl: 1.4894609989823948e-09 valid reconstr loss: 5.758101940155029
it: 8700, train recon loss: 4.665398597717285, local kl: 0.0 global kl: 7.54470264041629e-10 valid reconstr loss: 4.846518516540527
it: 8800, train recon loss: 3.516608238220215, local kl: 0.0 global kl: 2.785967012641777e-10 valid reconstr loss: 4.2502217292785645
it: 8900, train recon loss: 3.347446918487549, local kl: 0.0 global kl: 9.932392686096136e-10 valid reconstr loss: 3.8647453784942627
it: 9000, train recon loss: 3.0874712467193604, local kl: 0.0 global kl: 1.0540350814380872e-09 valid reconstr loss: 3.545517683029175
it: 9100, train recon loss: 3.5430757999420166, local kl: 0.0 global kl: 8.651994676256436e-10 valid reconstr loss: 3.3812475204467773
it: 9200, train recon loss: 2.9448344707489014, local kl: 0.0 global kl: 4.627236371845811e-10 valid reconstr loss: 3.217343807220459
it: 9300, train recon loss: 3.0558338165283203, local kl: 0.0 global kl: 4.0632786024730194e-10 valid reconstr loss: 3.1172657012939453
it: 9400, train recon loss: 3.107891082763672, local kl: 0.0 global kl: 6.945946040559647e-10 valid reconstr loss: 3.048311233520508
it: 9500, train recon loss: 3.0898373126983643, local kl: 0.0 global kl: 4.261355712742443e-10 valid reconstr loss: 3.0339157581329346
it: 9600, train recon loss: 2.914278745651245, local kl: 0.0 global kl: 2.4893553884908215e-10 valid reconstr loss: 2.969742774963379
it: 9700, train recon loss: 2.984058380126953, local kl: 0.0 global kl: 1.692372464390246e-09 valid reconstr loss: 2.938586950302124
it: 9800, train recon loss: 2.9899604320526123, local kl: 0.0 global kl: 1.375646263568342e-10 valid reconstr loss: 2.9133028984069824
it: 9900, train recon loss: 2.693040609359741, local kl: 0.0 global kl: 2.087300332576092e-10 valid reconstr loss: 2.939361810684204
beta 1.0 temperature 5.0
it: 0, train recon loss: 2844.452880859375, local kl: 0.0 global kl: 0.024622533470392227 valid reconstr loss: 936.0132446289062
Saving best model with reconstruction loss 936.01324
it: 100, train recon loss: 3.70800518989563, local kl: 0.0 global kl: 0.0008902029949240386 valid reconstr loss: 3.8772668838500977
Saving best model with reconstruction loss 3.877267
it: 200, train recon loss: 4.165043830871582, local kl: 0.0 global kl: 8.599054126534611e-05 valid reconstr loss: 4.331084251403809
it: 300, train recon loss: 3.939291477203369, local kl: 0.0 global kl: 5.4872525652172044e-05 valid reconstr loss: 3.6904919147491455
Saving best model with reconstruction loss 3.690492
it: 400, train recon loss: 3.6449029445648193, local kl: 0.0 global kl: 6.578933971468359e-05 valid reconstr loss: 3.4929757118225098
Saving best model with reconstruction loss 3.4929757
it: 500, train recon loss: 2.9581334590911865, local kl: 0.0 global kl: 0.000139949144795537 valid reconstr loss: 2.8767240047454834
Saving best model with reconstruction loss 2.876724
it: 600, train recon loss: 2.387495756149292, local kl: 0.0 global kl: 0.0004273762460798025 valid reconstr loss: 2.1581039428710938
Saving best model with reconstruction loss 2.158104
it: 700, train recon loss: 2.2687623500823975, local kl: 0.0 global kl: 0.00018694061145652086 valid reconstr loss: 2.511894702911377
it: 800, train recon loss: 1.7394564151763916, local kl: 0.0 global kl: 2.4987366487039253e-05 valid reconstr loss: 2.366469144821167
it: 900, train recon loss: 3.215679883956909, local kl: 0.0 global kl: 3.660183267584216e-12 valid reconstr loss: 1.5994269847869873
Saving best model with reconstruction loss 1.599427
it: 1000, train recon loss: 1.1525105237960815, local kl: 0.0 global kl: 1.912359159916832e-14 valid reconstr loss: 1.1176592111587524
Saving best model with reconstruction loss 1.1176592
it: 1100, train recon loss: 0.7607884407043457, local kl: 0.0 global kl: 4.794414865116892e-12 valid reconstr loss: 1.620458722114563
it: 1200, train recon loss: 1.0660091638565063, local kl: 0.0 global kl: 5.2695972585503625e-12 valid reconstr loss: 0.9589875340461731
Saving best model with reconstruction loss 0.95898753
it: 1300, train recon loss: 0.33915308117866516, local kl: 0.0 global kl: 4.5180525987120745e-13 valid reconstr loss: 0.5181164741516113
Saving best model with reconstruction loss 0.5181165
it: 1400, train recon loss: 0.2630886733531952, local kl: 0.0 global kl: 1.032209690984831e-12 valid reconstr loss: 0.2776622474193573
Saving best model with reconstruction loss 0.27766225
it: 1500, train recon loss: 1.1976090669631958, local kl: 0.0 global kl: 3.843869667008448e-13 valid reconstr loss: 0.4766598641872406
it: 1600, train recon loss: 0.1519719660282135, local kl: 0.0 global kl: 2.0518170495975596e-11 valid reconstr loss: 1072.9310302734375
it: 1700, train recon loss: -0.007484550587832928, local kl: 0.0 global kl: 1.618805783865085e-11 valid reconstr loss: 0.6816651225090027
it: 1800, train recon loss: 0.5723210573196411, local kl: 0.0 global kl: 2.386563169309852e-12 valid reconstr loss: 0.0655883252620697
Saving best model with reconstruction loss 0.065588325
it: 1900, train recon loss: 1.0388011932373047, local kl: 0.0 global kl: 5.236314853940272e-13 valid reconstr loss: 0.14772850275039673
it: 2000, train recon loss: 49.42549514770508, local kl: 0.0 global kl: 4.793054841911726e-12 valid reconstr loss: 95.30657958984375
it: 2100, train recon loss: -0.23496590554714203, local kl: 0.0 global kl: 3.57370383285982e-14 valid reconstr loss: 0.10693276673555374
it: 2200, train recon loss: 0.06165862828493118, local kl: 0.0 global kl: 4.631850458736153e-13 valid reconstr loss: -0.27888092398643494
Saving best model with reconstruction loss -0.27888092
it: 2300, train recon loss: -0.3652017116546631, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.19968917965888977
it: 2400, train recon loss: -0.15757745504379272, local kl: 0.0 global kl: 5.843381334358355e-13 valid reconstr loss: -0.1387719362974167
it: 2500, train recon loss: 0.016764625906944275, local kl: 0.0 global kl: 1.9196422229583732e-11 valid reconstr loss: 785.9191284179688
it: 2600, train recon loss: -0.6257510781288147, local kl: 0.0 global kl: 1.2415179995173276e-11 valid reconstr loss: -0.28371208906173706
Saving best model with reconstruction loss -0.2837121
it: 2700, train recon loss: 7.007343292236328, local kl: 0.0 global kl: 7.073924779277263e-12 valid reconstr loss: 2208.316162109375
it: 2800, train recon loss: -0.5314188599586487, local kl: 0.0 global kl: 2.1077879025499513e-12 valid reconstr loss: 712.3218383789062
it: 2900, train recon loss: -0.5038113594055176, local kl: 0.0 global kl: 1.174893515809572e-13 valid reconstr loss: -0.12605343759059906
it: 3000, train recon loss: -0.5142714381217957, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.40466856956481934
Saving best model with reconstruction loss -0.40466857
it: 3100, train recon loss: -0.7874129414558411, local kl: 0.0 global kl: 7.087817745915492e-12 valid reconstr loss: 0.22091792523860931
it: 3200, train recon loss: 35.60858154296875, local kl: 0.0 global kl: 2.0821084439903714e-11 valid reconstr loss: -0.43878915905952454
Saving best model with reconstruction loss -0.43878916
it: 3300, train recon loss: -0.5905234813690186, local kl: 0.0 global kl: 3.96797568760765e-12 valid reconstr loss: -0.559296727180481
Saving best model with reconstruction loss -0.5592967
it: 3400, train recon loss: -0.7042161822319031, local kl: 0.0 global kl: 3.500810752399275e-13 valid reconstr loss: -0.41803333163261414
it: 3500, train recon loss: 19.29417610168457, local kl: 0.0 global kl: 6.058598067681942e-12 valid reconstr loss: -0.2934204936027527
it: 3600, train recon loss: 105.2372055053711, local kl: 0.0 global kl: 3.372302437298913e-14 valid reconstr loss: 11.138212203979492
it: 3700, train recon loss: -1.3437418937683105, local kl: 0.0 global kl: 3.1282754164863036e-12 valid reconstr loss: -0.4180644452571869
it: 3800, train recon loss: -0.7730335593223572, local kl: 0.0 global kl: 1.5759534302550726e-11 valid reconstr loss: 3.4533567428588867
it: 3900, train recon loss: -0.7313935160636902, local kl: 0.0 global kl: 1.8639647117457692e-13 valid reconstr loss: -0.6439977288246155
Saving best model with reconstruction loss -0.6439977
it: 4000, train recon loss: -0.8412850499153137, local kl: 0.0 global kl: 2.2398749521812533e-14 valid reconstr loss: 0.43146848678588867
it: 4100, train recon loss: 0.6330627202987671, local kl: 0.0 global kl: 3.7238007816187135e-14 valid reconstr loss: -0.6572979688644409
Saving best model with reconstruction loss -0.65729797
it: 4200, train recon loss: 0.3515329658985138, local kl: 0.0 global kl: 4.362621375264553e-12 valid reconstr loss: -0.8966042995452881
Saving best model with reconstruction loss -0.8966043
it: 4300, train recon loss: -0.759873628616333, local kl: 0.0 global kl: 4.756035842934381e-12 valid reconstr loss: -0.5827486515045166
it: 4400, train recon loss: -0.7631624341011047, local kl: 0.0 global kl: 6.755707104844078e-14 valid reconstr loss: -0.8593665957450867
it: 4500, train recon loss: -1.0181376934051514, local kl: 0.0 global kl: 2.853273173286652e-14 valid reconstr loss: -0.2910700738430023
it: 4600, train recon loss: -1.158616542816162, local kl: 0.0 global kl: 2.3456889208794962e-11 valid reconstr loss: -0.5014904141426086
it: 4700, train recon loss: -1.1068015098571777, local kl: 0.0 global kl: 9.578449144953538e-14 valid reconstr loss: -0.22361873090267181
it: 4800, train recon loss: -1.0178401470184326, local kl: 0.0 global kl: 0.0 valid reconstr loss: 79.70734405517578
it: 4900, train recon loss: -1.2070220708847046, local kl: 0.0 global kl: 3.4721010788718587e-13 valid reconstr loss: 55.42023468017578
it: 5000, train recon loss: 25.376989364624023, local kl: 0.0 global kl: 0.0 valid reconstr loss: 35.171897888183594
it: 5100, train recon loss: -0.9429364204406738, local kl: 0.0 global kl: 0.0 valid reconstr loss: 758.2303466796875
it: 5200, train recon loss: -0.05145727097988129, local kl: 0.0 global kl: 4.703522293869611e-12 valid reconstr loss: 106.3570556640625
it: 5300, train recon loss: -0.5876305103302002, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.0431149005889893
Saving best model with reconstruction loss -1.0431149
it: 5400, train recon loss: 6422.01318359375, local kl: 0.0 global kl: 1.399869803409004e-13 valid reconstr loss: -0.7222063541412354
it: 5500, train recon loss: -1.082414984703064, local kl: 0.0 global kl: 4.808653475407709e-14 valid reconstr loss: 9.174144744873047
it: 5600, train recon loss: 1559.568359375, local kl: 0.0 global kl: 0.0 valid reconstr loss: 13.3163423538208
it: 5700, train recon loss: 2.308734655380249, local kl: 0.0 global kl: 2.5254381919026514e-11 valid reconstr loss: 2016.7684326171875
it: 5800, train recon loss: 7.022441864013672, local kl: 0.0 global kl: 4.696520949920568e-13 valid reconstr loss: 6.575688362121582
it: 5900, train recon loss: -1.4890142679214478, local kl: 0.0 global kl: 1.884777056648801e-14 valid reconstr loss: -0.29691773653030396
it: 6000, train recon loss: -1.2763683795928955, local kl: 0.0 global kl: 1.2625764691555352e-13 valid reconstr loss: -0.990578293800354
it: 6100, train recon loss: -1.1992504596710205, local kl: 0.0 global kl: 4.577761780755196e-14 valid reconstr loss: -0.9689176082611084
it: 6200, train recon loss: -0.9897974133491516, local kl: 0.0 global kl: 0.0 valid reconstr loss: 449.2598876953125
it: 6300, train recon loss: 388.57708740234375, local kl: 0.0 global kl: 8.536574225281868e-15 valid reconstr loss: 7.835834980010986
it: 6400, train recon loss: 37854.77734375, local kl: 0.0 global kl: 0.0 valid reconstr loss: 257.85870361328125
it: 6500, train recon loss: 11.659561157226562, local kl: 0.0 global kl: 6.238239091960196e-14 valid reconstr loss: 466.9843444824219
it: 6600, train recon loss: -1.2512625455856323, local kl: 0.0 global kl: 2.4701490852763186e-11 valid reconstr loss: -0.6670327186584473
it: 6700, train recon loss: 0.4333358108997345, local kl: 0.0 global kl: 2.47123502217228e-13 valid reconstr loss: 11.189140319824219
it: 6800, train recon loss: -1.0350035429000854, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.2542717456817627
it: 6900, train recon loss: 886.1097412109375, local kl: 0.0 global kl: 4.974909373345326e-13 valid reconstr loss: 6939.611328125
it: 7000, train recon loss: 3.6280617713928223, local kl: 0.0 global kl: 1.6070700326054066e-11 valid reconstr loss: 3.768780469894409
it: 7100, train recon loss: -0.15392810106277466, local kl: 0.0 global kl: 6.735757784870344e-13 valid reconstr loss: -0.0010953674791380763
it: 7200, train recon loss: 5.655978679656982, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.353198528289795
it: 7300, train recon loss: 50.14788055419922, local kl: 0.0 global kl: 2.0011770018868447e-14 valid reconstr loss: 389.2385559082031
it: 7400, train recon loss: 10.373037338256836, local kl: 0.0 global kl: 2.059638917750739e-12 valid reconstr loss: -0.769048273563385
it: 7500, train recon loss: -0.12825094163417816, local kl: 0.0 global kl: 2.4683380339673988e-14 valid reconstr loss: 863.5009765625
it: 7600, train recon loss: -1.026536226272583, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.0627672672271729
Saving best model with reconstruction loss -1.0627673
it: 7700, train recon loss: 2515.695556640625, local kl: 0.0 global kl: 1.2039802488317264e-10 valid reconstr loss: 4.2535505294799805
it: 7800, train recon loss: -1.0714927911758423, local kl: 0.0 global kl: 1.7333356971960256e-13 valid reconstr loss: -1.1559391021728516
Saving best model with reconstruction loss -1.1559391
it: 7900, train recon loss: -1.081003189086914, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.9019173383712769
it: 8000, train recon loss: -1.0908406972885132, local kl: 0.0 global kl: 1.544597783009749e-14 valid reconstr loss: -0.9774489998817444
it: 8100, train recon loss: -1.1129367351531982, local kl: 0.0 global kl: 1.587618925213974e-14 valid reconstr loss: -1.1542670726776123
it: 8200, train recon loss: -0.8965346217155457, local kl: 0.0 global kl: 2.673555821175455e-14 valid reconstr loss: -0.9092662334442139
it: 8300, train recon loss: -1.189287543296814, local kl: 0.0 global kl: 1.2859158182720876e-11 valid reconstr loss: -0.7300843000411987
it: 8400, train recon loss: -1.1391143798828125, local kl: 0.0 global kl: 4.4748926786297716e-14 valid reconstr loss: -1.1371710300445557
it: 8500, train recon loss: -1.2091069221496582, local kl: 0.0 global kl: 3.7925218521195347e-13 valid reconstr loss: -0.9669251441955566
it: 8600, train recon loss: -1.142205834388733, local kl: 0.0 global kl: 2.7852720130283615e-12 valid reconstr loss: 3075.2373046875
it: 8700, train recon loss: 8.331666946411133, local kl: 0.0 global kl: 4.338640991613518e-14 valid reconstr loss: 19.86954689025879
it: 8800, train recon loss: -1.0367976427078247, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.7699680328369141
it: 8900, train recon loss: 6.180401802062988, local kl: 0.0 global kl: 0.0 valid reconstr loss: 345.7122497558594
it: 9000, train recon loss: -1.364116907119751, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.19957435131073
Saving best model with reconstruction loss -1.1995744
it: 9100, train recon loss: -1.0065258741378784, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.3509564399719238
Saving best model with reconstruction loss -1.3509564
it: 9200, train recon loss: -1.2379605770111084, local kl: 0.0 global kl: 0.0 valid reconstr loss: 30.134109497070312
it: 9300, train recon loss: -0.9357291460037231, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.106593132019043
it: 9400, train recon loss: -0.5601221919059753, local kl: 0.0 global kl: 0.0 valid reconstr loss: 22732568.0
it: 9500, train recon loss: -1.0716731548309326, local kl: 0.0 global kl: 0.0 valid reconstr loss: 857.0199584960938
it: 9600, train recon loss: 3.5974886417388916, local kl: 0.0 global kl: 0.0 valid reconstr loss: 8.606245994567871
it: 9700, train recon loss: -1.3258213996887207, local kl: 0.0 global kl: 0.0 valid reconstr loss: 17.095462799072266
it: 9800, train recon loss: -0.9333859086036682, local kl: 0.0 global kl: 0.0 valid reconstr loss: 43.4586296081543
it: 9900, train recon loss: -1.3522825241088867, local kl: 0.0 global kl: 0.0 valid reconstr loss: 108.81183624267578
beta 1.0 temperature 1000.0
it: 0, train recon loss: 96678.7734375, local kl: 0.0 global kl: 0.011864875443279743 valid reconstr loss: 557.1206665039062
Saving best model with reconstruction loss 557.12067
it: 100, train recon loss: 3.792186737060547, local kl: 0.0 global kl: 0.0015546574722975492 valid reconstr loss: 4.429218769073486
Saving best model with reconstruction loss 4.429219
it: 200, train recon loss: 3.98447585105896, local kl: 0.0 global kl: 5.088742454972817e-06 valid reconstr loss: 4.038256645202637
Saving best model with reconstruction loss 4.0382566
it: 300, train recon loss: 3.797269821166992, local kl: 0.0 global kl: 4.536944288702216e-06 valid reconstr loss: 3.964303731918335
Saving best model with reconstruction loss 3.9643037
it: 400, train recon loss: 3.7620272636413574, local kl: 0.0 global kl: 0.0009992632549256086 valid reconstr loss: 3.9212489128112793
Saving best model with reconstruction loss 3.921249
it: 500, train recon loss: 2.8954334259033203, local kl: 0.0 global kl: 0.0005749693955294788 valid reconstr loss: 3.0295569896698
Saving best model with reconstruction loss 3.029557
it: 600, train recon loss: 2.471883535385132, local kl: 0.0 global kl: 0.00013110041618347168 valid reconstr loss: 2.482603073120117
Saving best model with reconstruction loss 2.482603
it: 700, train recon loss: 1.948744773864746, local kl: 0.0 global kl: 7.168745848676328e-13 valid reconstr loss: 2.081254005432129
Saving best model with reconstruction loss 2.081254
it: 800, train recon loss: 2.9125287532806396, local kl: 0.0 global kl: 1.6900439381295485e-12 valid reconstr loss: 1.9245202541351318
Saving best model with reconstruction loss 1.9245203
it: 900, train recon loss: 2.2941975593566895, local kl: 0.0 global kl: 3.885948854365218e-12 valid reconstr loss: 2.2004878520965576
it: 1000, train recon loss: 1.7212430238723755, local kl: 0.0 global kl: 5.0155227193648955e-12 valid reconstr loss: 1.5941791534423828
Saving best model with reconstruction loss 1.5941792
it: 1100, train recon loss: 1.475156545639038, local kl: 0.0 global kl: 3.078648447285559e-13 valid reconstr loss: 1.8417046070098877
it: 1200, train recon loss: 1.4410264492034912, local kl: 0.0 global kl: 1.2212453270876722e-11 valid reconstr loss: 1.3513721227645874
Saving best model with reconstruction loss 1.3513721
it: 1300, train recon loss: 5.388496398925781, local kl: 0.0 global kl: 4.317612673637727e-12 valid reconstr loss: 1.476723551750183
it: 1400, train recon loss: 7.884854793548584, local kl: 0.0 global kl: 9.636132169976719e-12 valid reconstr loss: 11.849608421325684
it: 1500, train recon loss: 0.8517408967018127, local kl: 0.0 global kl: 4.0488589043985446e-12 valid reconstr loss: 1.0233852863311768
Saving best model with reconstruction loss 1.0233853
it: 1600, train recon loss: 0.9685290455818176, local kl: 0.0 global kl: 5.80176184872272e-12 valid reconstr loss: 0.8546038866043091
Saving best model with reconstruction loss 0.8546039
it: 1700, train recon loss: 1.2402290105819702, local kl: 0.0 global kl: 2.5581564644383548e-11 valid reconstr loss: 1.0451977252960205
it: 1800, train recon loss: 2.8907554149627686, local kl: 0.0 global kl: 7.691389192210352e-12 valid reconstr loss: 0.857481062412262
it: 1900, train recon loss: 0.6606115698814392, local kl: 0.0 global kl: 3.0031532816110484e-13 valid reconstr loss: 0.760978639125824
Saving best model with reconstruction loss 0.76097864
it: 2000, train recon loss: 0.7704179286956787, local kl: 0.0 global kl: 2.1736128522076292e-13 valid reconstr loss: 0.8343008160591125
it: 2100, train recon loss: 0.4846385419368744, local kl: 0.0 global kl: 5.51463388842599e-13 valid reconstr loss: 1.2955691814422607
it: 2200, train recon loss: 0.6376739740371704, local kl: 0.0 global kl: 3.218814104144485e-13 valid reconstr loss: 6.797361850738525
it: 2300, train recon loss: 0.3887166976928711, local kl: 0.0 global kl: 7.807365864920257e-13 valid reconstr loss: 1.189595103263855
it: 2400, train recon loss: 1.492140769958496, local kl: 0.0 global kl: 1.48849821357544e-11 valid reconstr loss: 0.6063063144683838
Saving best model with reconstruction loss 0.6063063
it: 2500, train recon loss: 0.8650485277175903, local kl: 0.0 global kl: 2.898953863419984e-13 valid reconstr loss: 0.6076894998550415
it: 2600, train recon loss: 0.6262710690498352, local kl: 0.0 global kl: 1.5574104506033137e-11 valid reconstr loss: 0.48604002594947815
Saving best model with reconstruction loss 0.48604003
it: 2700, train recon loss: 120.28144073486328, local kl: 0.0 global kl: 3.3098220345029183e-13 valid reconstr loss: 0.6616613268852234
it: 2800, train recon loss: 0.5400341153144836, local kl: 0.0 global kl: 3.028995457232675e-12 valid reconstr loss: 0.5134041905403137
it: 2900, train recon loss: 0.22853699326515198, local kl: 0.0 global kl: 2.1119776066891305e-11 valid reconstr loss: 332.64703369140625
it: 3000, train recon loss: 0.6537483930587769, local kl: 0.0 global kl: 5.921430012989504e-12 valid reconstr loss: 0.5414363145828247
it: 3100, train recon loss: 0.36091873049736023, local kl: 0.0 global kl: 2.1199148339534624e-11 valid reconstr loss: 4.503298282623291
it: 3200, train recon loss: 0.14555658400058746, local kl: 0.0 global kl: 1.5045659162993275e-11 valid reconstr loss: 0.17971080541610718
Saving best model with reconstruction loss 0.1797108
it: 3300, train recon loss: 0.1246446818113327, local kl: 0.0 global kl: 6.950356956636483e-12 valid reconstr loss: 0.4847061336040497
it: 3400, train recon loss: 0.23345842957496643, local kl: 0.0 global kl: 1.365088597715669e-13 valid reconstr loss: 3.955444097518921
it: 3500, train recon loss: 0.03646663948893547, local kl: 0.0 global kl: 1.865174681370263e-14 valid reconstr loss: 0.621124267578125
it: 3600, train recon loss: 0.141365185379982, local kl: 0.0 global kl: 3.0332160394497265e-13 valid reconstr loss: 0.41109201312065125
it: 3700, train recon loss: 0.5780637860298157, local kl: 0.0 global kl: 2.8277727381897932e-14 valid reconstr loss: 0.31379106640815735
it: 3800, train recon loss: 0.32196033000946045, local kl: 0.0 global kl: 8.72635297355373e-13 valid reconstr loss: 0.28367307782173157
it: 3900, train recon loss: 0.1713934689760208, local kl: 0.0 global kl: 2.9271723933632643e-13 valid reconstr loss: 0.40945032238960266
it: 4000, train recon loss: -0.48461124300956726, local kl: 0.0 global kl: 3.1482206164917814e-12 valid reconstr loss: 0.3507240116596222
it: 4100, train recon loss: -0.6252821683883667, local kl: 0.0 global kl: 3.665337889972864e-13 valid reconstr loss: 0.8151065111160278
it: 4200, train recon loss: -0.14084745943546295, local kl: 0.0 global kl: 6.258435610029256e-13 valid reconstr loss: -0.1503782868385315
Saving best model with reconstruction loss -0.15037829
it: 4300, train recon loss: 1698.3543701171875, local kl: 0.0 global kl: 9.53506718026631e-12 valid reconstr loss: 0.1488509625196457
it: 4400, train recon loss: 0.31943243741989136, local kl: 0.0 global kl: 9.860140481876556e-12 valid reconstr loss: 2.403567314147949
it: 4500, train recon loss: 0.4316249191761017, local kl: 0.0 global kl: 8.241324289670615e-12 valid reconstr loss: -0.3711569011211395
Saving best model with reconstruction loss -0.3711569
it: 4600, train recon loss: 36.85808181762695, local kl: 0.0 global kl: 2.619575563411747e-13 valid reconstr loss: 1862.8189697265625
it: 4700, train recon loss: -0.15887384116649628, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.49526602029800415
Saving best model with reconstruction loss -0.49526602
it: 4800, train recon loss: -0.6530165076255798, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.4645960032939911
it: 4900, train recon loss: -0.1741476207971573, local kl: 0.0 global kl: 5.4605625576797934e-14 valid reconstr loss: 188.3696746826172
it: 5000, train recon loss: -0.5435296297073364, local kl: 0.0 global kl: 2.8716751199198143e-12 valid reconstr loss: -0.038937099277973175
it: 5100, train recon loss: -0.44737178087234497, local kl: 0.0 global kl: 4.681691232605312e-12 valid reconstr loss: 0.08560094982385635
it: 5200, train recon loss: -0.47418975830078125, local kl: 0.0 global kl: 1.1193823645783141e-13 valid reconstr loss: 9.587189674377441
it: 5300, train recon loss: -0.5455335378646851, local kl: 0.0 global kl: 2.8185787037671162e-14 valid reconstr loss: 3.7717697620391846
it: 5400, train recon loss: -0.4169463813304901, local kl: 0.0 global kl: 7.121497835882451e-12 valid reconstr loss: 23.676284790039062
it: 5500, train recon loss: -0.12569065392017365, local kl: 0.0 global kl: 1.3414443167381052e-13 valid reconstr loss: -0.37124261260032654
it: 5600, train recon loss: 316.3735046386719, local kl: 0.0 global kl: 6.593683932187844e-15 valid reconstr loss: 7.699995040893555
it: 5700, train recon loss: 25.075275421142578, local kl: 0.0 global kl: 2.0300709897835834e-14 valid reconstr loss: 26.547374725341797
it: 5800, train recon loss: -0.018771525472402573, local kl: 0.0 global kl: 1.951120254944927e-12 valid reconstr loss: -0.48040249943733215
it: 5900, train recon loss: -1.0288310050964355, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.7228908538818359
Saving best model with reconstruction loss -0.72289085
it: 6000, train recon loss: -0.8559548854827881, local kl: 0.0 global kl: 5.507400091531167e-14 valid reconstr loss: -0.7883213758468628
Saving best model with reconstruction loss -0.7883214
it: 6100, train recon loss: -0.6220742464065552, local kl: 0.0 global kl: 6.061123825062964e-14 valid reconstr loss: -0.6451442241668701
it: 6200, train recon loss: -0.7613414525985718, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.5698322057723999
it: 6300, train recon loss: -0.4439217746257782, local kl: 0.0 global kl: 5.201568342716456e-14 valid reconstr loss: -0.7096276879310608
it: 6400, train recon loss: -0.45498695969581604, local kl: 0.0 global kl: 1.657007864253046e-14 valid reconstr loss: 97.8388671875
it: 6500, train recon loss: 0.04109283164143562, local kl: 0.0 global kl: 7.43016759230386e-14 valid reconstr loss: -0.8940405249595642
Saving best model with reconstruction loss -0.8940405
it: 6600, train recon loss: -1.0006928443908691, local kl: 0.0 global kl: 6.390721285498557e-15 valid reconstr loss: 91.0363998413086
it: 6700, train recon loss: -0.4722035527229309, local kl: 0.0 global kl: 1.0292934907174534e-11 valid reconstr loss: 583.6987915039062
it: 6800, train recon loss: 1.5853787660598755, local kl: 0.0 global kl: 4.6281481251575496e-12 valid reconstr loss: 1383.8851318359375
it: 6900, train recon loss: -0.6433066129684448, local kl: 0.0 global kl: 3.991251773527438e-14 valid reconstr loss: 2764.505615234375
it: 7000, train recon loss: -0.4632841944694519, local kl: 0.0 global kl: 4.989758606299688e-14 valid reconstr loss: 1.9607489109039307
it: 7100, train recon loss: -0.33425667881965637, local kl: 0.0 global kl: 8.46801968823474e-12 valid reconstr loss: -0.5524623394012451
it: 7200, train recon loss: 1707.1180419921875, local kl: 0.0 global kl: 6.902285600712821e-12 valid reconstr loss: 54.969459533691406
it: 7300, train recon loss: -0.358405202627182, local kl: 0.0 global kl: 6.564887522486629e-14 valid reconstr loss: -0.8890905380249023
it: 7400, train recon loss: -0.6686990857124329, local kl: 0.0 global kl: 1.7658097206663115e-13 valid reconstr loss: -0.7863251566886902
it: 7500, train recon loss: -0.9753130078315735, local kl: 0.0 global kl: 3.640837631380123e-14 valid reconstr loss: 32.637229919433594
it: 7600, train recon loss: 145.84976196289062, local kl: 0.0 global kl: 3.611989873864019e-11 valid reconstr loss: 4.813383102416992
it: 7700, train recon loss: -1.3101003170013428, local kl: 0.0 global kl: 2.5368596112684827e-14 valid reconstr loss: -0.6572477221488953
it: 7800, train recon loss: -0.3201248049736023, local kl: 0.0 global kl: 8.98812274607863e-14 valid reconstr loss: 9247.310546875
it: 7900, train recon loss: -0.9800942540168762, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.7253918647766113
it: 8000, train recon loss: -1.05014967918396, local kl: 0.0 global kl: 6.025846596875717e-14 valid reconstr loss: -0.9730849266052246
Saving best model with reconstruction loss -0.9730849
it: 8100, train recon loss: -0.9303898811340332, local kl: 0.0 global kl: 3.693637409818429e-12 valid reconstr loss: -1.1980446577072144
Saving best model with reconstruction loss -1.1980447
it: 8200, train recon loss: 26.518075942993164, local kl: 0.0 global kl: 2.2479414163445455e-13 valid reconstr loss: -1.11909019947052
it: 8300, train recon loss: -1.0016324520111084, local kl: 0.0 global kl: 1.0029429137236104e-15 valid reconstr loss: -1.0568206310272217
it: 8400, train recon loss: -1.182935357093811, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.809333324432373
it: 8500, train recon loss: -1.03703773021698, local kl: 0.0 global kl: 2.190150838465854e-12 valid reconstr loss: -0.4466915726661682
it: 8600, train recon loss: 0.4577989876270294, local kl: 0.0 global kl: 0.0 valid reconstr loss: 8.801080703735352
it: 8700, train recon loss: 82.86622619628906, local kl: 0.0 global kl: 1.2791283657276331e-11 valid reconstr loss: 7.875330924987793
it: 8800, train recon loss: -0.9954355955123901, local kl: 0.0 global kl: 2.320263122260191e-13 valid reconstr loss: -0.6459962725639343
it: 8900, train recon loss: -1.0981724262237549, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.0258816480636597
it: 9000, train recon loss: -1.1477382183074951, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.0465233325958252
it: 9100, train recon loss: 18.947956085205078, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.5785835981369019
it: 9200, train recon loss: -1.0087722539901733, local kl: 0.0 global kl: 0.0 valid reconstr loss: 281.2738037109375
it: 9300, train recon loss: -0.9046710729598999, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.0844438076019287
it: 9400, train recon loss: -0.8336183428764343, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.0438371896743774
it: 9500, train recon loss: -1.0721595287322998, local kl: 0.0 global kl: 3.9995784462121264e-14 valid reconstr loss: -1.155644178390503
it: 9600, train recon loss: 3.9189884662628174, local kl: 0.0 global kl: 5.305311745473773e-12 valid reconstr loss: 3.397125720977783
it: 9700, train recon loss: -0.8844397068023682, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.0973871946334839
it: 9800, train recon loss: -0.9380819201469421, local kl: 0.0 global kl: 2.165934098741218e-12 valid reconstr loss: -0.9595049619674683
it: 9900, train recon loss: -0.2094893455505371, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.40921077132225037
beta 2.0 temperature 0.001
it: 0, train recon loss: 124845592.0, local kl: 0.0 global kl: 0.03373324126005173 valid reconstr loss: 109707904.0
Saving best model with reconstruction loss 109707900.0
it: 100, train recon loss: 3.1694693565368652, local kl: 0.0 global kl: 0.0008293673163279891 valid reconstr loss: 3.6301558017730713
Saving best model with reconstruction loss 3.6301558
it: 200, train recon loss: 3.0153183937072754, local kl: 0.0 global kl: 2.387701897532679e-05 valid reconstr loss: 3.142491102218628
Saving best model with reconstruction loss 3.142491
it: 300, train recon loss: 2.565331220626831, local kl: 0.0 global kl: 3.14824101224076e-05 valid reconstr loss: 2.5669543743133545
Saving best model with reconstruction loss 2.5669544
it: 400, train recon loss: 4.264248371124268, local kl: 0.0 global kl: 1.3978064998809714e-05 valid reconstr loss: 5.712296962738037
it: 500, train recon loss: 1.5283827781677246, local kl: 0.0 global kl: 7.096641638781875e-05 valid reconstr loss: 1.4034438133239746
Saving best model with reconstruction loss 1.4034438
it: 600, train recon loss: 1.202120065689087, local kl: 0.0 global kl: 0.00023790537670720369 valid reconstr loss: 4.300086975097656
it: 700, train recon loss: 0.9212583303451538, local kl: 0.0 global kl: 5.870268796570599e-05 valid reconstr loss: 1.2188762426376343
Saving best model with reconstruction loss 1.2188762
it: 800, train recon loss: 19.744287490844727, local kl: 0.0 global kl: 9.928895451594144e-05 valid reconstr loss: 24.956907272338867
it: 900, train recon loss: 0.7903504967689514, local kl: 0.0 global kl: 6.858012784505263e-05 valid reconstr loss: 0.42467647790908813
Saving best model with reconstruction loss 0.42467648
it: 1000, train recon loss: 10.130209922790527, local kl: 0.0 global kl: 0.0016361197922378778 valid reconstr loss: 2.151761770248413
it: 1100, train recon loss: 1.344541072845459, local kl: 0.0 global kl: 0.00039971014484763145 valid reconstr loss: 7.6101250648498535
it: 1200, train recon loss: 2.720548629760742, local kl: 0.0 global kl: 1.1483866728667635e-05 valid reconstr loss: 1.7011542320251465
it: 1300, train recon loss: 0.2086099237203598, local kl: 0.0 global kl: 3.7968613469274715e-05 valid reconstr loss: 0.42766863107681274
it: 1400, train recon loss: 0.15254124999046326, local kl: 0.0 global kl: 4.537486165645532e-05 valid reconstr loss: 0.06055312976241112
Saving best model with reconstruction loss 0.06055313
it: 1500, train recon loss: 31.254241943359375, local kl: 0.0 global kl: 7.910820568213239e-05 valid reconstr loss: 1.064697027206421
it: 1600, train recon loss: 0.7865009903907776, local kl: 0.0 global kl: 0.00018200646445620805 valid reconstr loss: 34.90460968017578
it: 1700, train recon loss: -0.07462575286626816, local kl: 0.0 global kl: 1.9078135665040463e-05 valid reconstr loss: -0.02244945801794529
Saving best model with reconstruction loss -0.022449458
it: 1800, train recon loss: 0.07685984671115875, local kl: 0.0 global kl: 4.165370773989707e-05 valid reconstr loss: 101.03886413574219
it: 1900, train recon loss: -0.577543318271637, local kl: 0.0 global kl: 0.00023823624360375106 valid reconstr loss: 0.23879824578762054
it: 2000, train recon loss: -0.008703935891389847, local kl: 0.0 global kl: 0.000157663831487298 valid reconstr loss: -0.3125721216201782
Saving best model with reconstruction loss -0.31257212
it: 2100, train recon loss: -0.562224268913269, local kl: 0.0 global kl: 0.00013620671234093606 valid reconstr loss: 33.70674514770508
it: 2200, train recon loss: -0.5953431725502014, local kl: 0.0 global kl: 3.569661930669099e-05 valid reconstr loss: -0.011945766396820545
it: 2300, train recon loss: -0.07783137261867523, local kl: 0.0 global kl: 3.173000004608184e-05 valid reconstr loss: -0.45702868700027466
Saving best model with reconstruction loss -0.4570287
it: 2400, train recon loss: 2.6712703704833984, local kl: 0.0 global kl: 4.210070983390324e-05 valid reconstr loss: 0.7579938769340515
it: 2500, train recon loss: -0.27895909547805786, local kl: 0.0 global kl: 1.007416176435072e-05 valid reconstr loss: 108.40281677246094
it: 2600, train recon loss: -0.7201927900314331, local kl: 0.0 global kl: 0.00016234211216215044 valid reconstr loss: -0.3302612006664276
it: 2700, train recon loss: -0.5739248991012573, local kl: 0.0 global kl: 2.4667349862284027e-05 valid reconstr loss: -0.754169762134552
Saving best model with reconstruction loss -0.75416976
it: 2800, train recon loss: -0.7732337713241577, local kl: 0.0 global kl: 8.54150130180642e-05 valid reconstr loss: -0.6871084570884705
it: 2900, train recon loss: -0.6269879341125488, local kl: 0.0 global kl: 0.00010860244219657034 valid reconstr loss: -0.48351284861564636
it: 3000, train recon loss: -0.7025773525238037, local kl: 0.0 global kl: 4.746040576719679e-05 valid reconstr loss: 0.40039142966270447
it: 3100, train recon loss: -0.3904520869255066, local kl: 0.0 global kl: 1.8415810700389557e-05 valid reconstr loss: -0.7856591939926147
Saving best model with reconstruction loss -0.7856592
it: 3200, train recon loss: -0.44461527466773987, local kl: 0.0 global kl: 0.00013398074952419847 valid reconstr loss: -0.6235423684120178
it: 3300, train recon loss: -0.9139378070831299, local kl: 0.0 global kl: 0.0002513132931198925 valid reconstr loss: 3.853695869445801
it: 3400, train recon loss: -0.8392060399055481, local kl: 0.0 global kl: 4.938698839396238e-05 valid reconstr loss: -0.8289964199066162
Saving best model with reconstruction loss -0.8289964
it: 3500, train recon loss: -1.2669991254806519, local kl: 0.0 global kl: 0.00013766286429017782 valid reconstr loss: 2.3380064964294434
it: 3600, train recon loss: -0.48382875323295593, local kl: 0.0 global kl: 3.9670187106821686e-05 valid reconstr loss: -0.8856629729270935
Saving best model with reconstruction loss -0.885663
it: 3700, train recon loss: -1.1288564205169678, local kl: 0.0 global kl: 9.14770134841092e-06 valid reconstr loss: -0.6989102363586426
it: 3800, train recon loss: -1.053454875946045, local kl: 0.0 global kl: 0.00010858957102755085 valid reconstr loss: -0.671229898929596
it: 3900, train recon loss: -1.0628712177276611, local kl: 0.0 global kl: 3.830160858342424e-05 valid reconstr loss: 74.23857116699219
it: 4000, train recon loss: -1.3158093690872192, local kl: 0.0 global kl: 8.692813935340382e-07 valid reconstr loss: 3534.586181640625
it: 4100, train recon loss: -1.2880139350891113, local kl: 0.0 global kl: 6.674677479168167e-06 valid reconstr loss: -1.0343214273452759
Saving best model with reconstruction loss -1.0343214
it: 4200, train recon loss: -0.9413096904754639, local kl: 0.0 global kl: 8.56607755395089e-07 valid reconstr loss: -0.5489954352378845
it: 4300, train recon loss: -1.1776927709579468, local kl: 0.0 global kl: 9.835314813244622e-06 valid reconstr loss: -0.8400159478187561
it: 4400, train recon loss: -0.8176767826080322, local kl: 0.0 global kl: 6.079902050259989e-06 valid reconstr loss: 94.12278747558594
it: 4500, train recon loss: -1.1109797954559326, local kl: 0.0 global kl: 1.877779322967399e-05 valid reconstr loss: -0.8102320432662964
it: 4600, train recon loss: 7842.57177734375, local kl: 0.0 global kl: 2.6033446829387685e-06 valid reconstr loss: -1.028620719909668
it: 4700, train recon loss: -0.7809526920318604, local kl: 0.0 global kl: 1.924551725096535e-05 valid reconstr loss: 1231.09814453125
it: 4800, train recon loss: -1.4043611288070679, local kl: 0.0 global kl: 2.1750576706836e-05 valid reconstr loss: -1.1588553190231323
Saving best model with reconstruction loss -1.1588553
it: 4900, train recon loss: 3.739431619644165, local kl: 0.0 global kl: 0.0017779372865334153 valid reconstr loss: 9810.171875
it: 5000, train recon loss: -1.0323058366775513, local kl: 0.0 global kl: 0.0001158857048721984 valid reconstr loss: -1.011243224143982
it: 5100, train recon loss: -1.0020745992660522, local kl: 0.0 global kl: 1.6348403732990846e-05 valid reconstr loss: 9733.455078125
it: 5200, train recon loss: -1.1878550052642822, local kl: 0.0 global kl: 2.698293792491313e-05 valid reconstr loss: -1.1133763790130615
it: 5300, train recon loss: -0.8888574242591858, local kl: 0.0 global kl: 0.0001298092247452587 valid reconstr loss: 13.826881408691406
it: 5400, train recon loss: -1.3079874515533447, local kl: 0.0 global kl: 0.00020332387066446245 valid reconstr loss: -0.39183226227760315
it: 5500, train recon loss: 18197.58984375, local kl: 0.0 global kl: 0.00012934286496602 valid reconstr loss: 77.98609161376953
it: 5600, train recon loss: -1.2397438287734985, local kl: 0.0 global kl: 5.413923190644709e-06 valid reconstr loss: -1.0651816129684448
it: 5700, train recon loss: -1.2807844877243042, local kl: 0.0 global kl: 9.77857143880101e-06 valid reconstr loss: -0.6491421461105347
it: 5800, train recon loss: -1.148349404335022, local kl: 0.0 global kl: 2.8060353542969096e-06 valid reconstr loss: -1.067090630531311
it: 5900, train recon loss: -1.424493670463562, local kl: 0.0 global kl: 5.339330527931452e-05 valid reconstr loss: -1.0476590394973755
it: 6000, train recon loss: -1.4093576669692993, local kl: 0.0 global kl: 8.590451216150541e-06 valid reconstr loss: -1.1942498683929443
Saving best model with reconstruction loss -1.1942499
it: 6100, train recon loss: 10518.7451171875, local kl: 0.0 global kl: 0.003393098246306181 valid reconstr loss: 155.901611328125
it: 6200, train recon loss: -1.3096481561660767, local kl: 0.0 global kl: 7.914358945981803e-08 valid reconstr loss: -1.2186552286148071
Saving best model with reconstruction loss -1.2186552
it: 6300, train recon loss: -1.0954455137252808, local kl: 0.0 global kl: 1.654081074775604e-06 valid reconstr loss: 3849.836181640625
it: 6400, train recon loss: 4422.9814453125, local kl: 0.0 global kl: 6.716646112181479e-06 valid reconstr loss: -0.7842973470687866
it: 6500, train recon loss: -1.240838885307312, local kl: 0.0 global kl: 1.1486607945698779e-05 valid reconstr loss: -0.9134296178817749
it: 6600, train recon loss: -1.2012158632278442, local kl: 0.0 global kl: 0.00021466289763338864 valid reconstr loss: -1.2773491144180298
Saving best model with reconstruction loss -1.2773491
it: 6700, train recon loss: -1.2753757238388062, local kl: 0.0 global kl: 3.4884549222624628e-06 valid reconstr loss: -0.8714556694030762
it: 6800, train recon loss: -1.277818202972412, local kl: 0.0 global kl: 4.406471362017328e-06 valid reconstr loss: -1.3207398653030396
Saving best model with reconstruction loss -1.3207399
it: 6900, train recon loss: -1.0354281663894653, local kl: 0.0 global kl: 1.6294430679408833e-05 valid reconstr loss: -1.2568163871765137
it: 7000, train recon loss: 6920.9013671875, local kl: 0.0 global kl: 1.0151447895623278e-05 valid reconstr loss: -1.0796257257461548
it: 7100, train recon loss: 104.48783874511719, local kl: 0.0 global kl: 1.0016813121183077e-06 valid reconstr loss: 46.52421188354492
it: 7200, train recon loss: -0.03020254150032997, local kl: 0.0 global kl: 2.74008543783566e-05 valid reconstr loss: 0.044480983167886734
it: 7300, train recon loss: -0.6664497256278992, local kl: 0.0 global kl: 3.6915378586854786e-05 valid reconstr loss: 5085.30615234375
it: 7400, train recon loss: -1.037716031074524, local kl: 0.0 global kl: 1.2533266726677539e-06 valid reconstr loss: -1.2581158876419067
it: 7500, train recon loss: -0.9803129434585571, local kl: 0.0 global kl: 3.4243280424561817e-06 valid reconstr loss: 18.132123947143555
it: 7600, train recon loss: -1.236832857131958, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.1996794939041138
it: 7700, train recon loss: -1.4800010919570923, local kl: 0.0 global kl: 3.6338376503408654e-06 valid reconstr loss: -1.2341676950454712
it: 7800, train recon loss: -1.172013521194458, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.235227108001709
it: 7900, train recon loss: -0.31792253255844116, local kl: 0.0 global kl: 6.51798109174706e-05 valid reconstr loss: -0.5683717131614685
it: 8000, train recon loss: -0.989784300327301, local kl: 0.0 global kl: 3.120669816780719e-06 valid reconstr loss: -0.8084273338317871
it: 8100, train recon loss: -0.8710158467292786, local kl: 0.0 global kl: 3.0230762604332995e-06 valid reconstr loss: -0.9569957256317139
it: 8200, train recon loss: -1.1592601537704468, local kl: 0.0 global kl: 2.2780572180636227e-05 valid reconstr loss: -0.9814577698707581
it: 8300, train recon loss: -1.0362939834594727, local kl: 0.0 global kl: 9.240435474566766e-07 valid reconstr loss: -1.2518759965896606
it: 8400, train recon loss: -1.2034441232681274, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.038946509361267
it: 8500, train recon loss: -0.07315338402986526, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.2320396453142166
it: 8600, train recon loss: -0.5797041058540344, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.43242010474205017
it: 8700, train recon loss: -0.8729492425918579, local kl: 0.0 global kl: 0.00011054112110286951 valid reconstr loss: -0.6239429712295532
it: 8800, train recon loss: -1.2416589260101318, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.196427345275879
it: 8900, train recon loss: -1.0160282850265503, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.0319230556488037
it: 9000, train recon loss: -1.079132080078125, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.8260467052459717
it: 9100, train recon loss: -0.5112488269805908, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.5486705303192139
it: 9200, train recon loss: -0.9142536520957947, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.8753082752227783
it: 9300, train recon loss: -0.6936737298965454, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.0590646266937256
it: 9400, train recon loss: -0.869291365146637, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.044881820678711
it: 9500, train recon loss: -0.7522713541984558, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.5995522141456604
it: 9600, train recon loss: -1.0749988555908203, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.1918833255767822
it: 9700, train recon loss: -1.1778085231781006, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.0278562307357788
it: 9800, train recon loss: -1.1539350748062134, local kl: 0.0 global kl: 1.8519525838200934e-05 valid reconstr loss: 12020.0791015625
it: 9900, train recon loss: -0.965517520904541, local kl: 0.0 global kl: 0.0 valid reconstr loss: 11704.3486328125
beta 2.0 temperature 0.5
it: 0, train recon loss: 3479.92822265625, local kl: 0.0 global kl: 0.034822624176740646 valid reconstr loss: 429.8936462402344
Saving best model with reconstruction loss 429.89365
it: 100, train recon loss: 4.2772369384765625, local kl: 0.0 global kl: 0.003812511218711734 valid reconstr loss: 4.089062213897705
Saving best model with reconstruction loss 4.089062
it: 200, train recon loss: 3.4207406044006348, local kl: 0.0 global kl: 0.0004594691563397646 valid reconstr loss: 3.2546472549438477
Saving best model with reconstruction loss 3.2546473
it: 300, train recon loss: 2.737766981124878, local kl: 0.0 global kl: 7.357960566878319e-05 valid reconstr loss: 2.8534486293792725
Saving best model with reconstruction loss 2.8534486
it: 400, train recon loss: 2.4352450370788574, local kl: 0.0 global kl: 4.117196658626199e-05 valid reconstr loss: 2.4910714626312256
Saving best model with reconstruction loss 2.4910715
it: 500, train recon loss: 1.9119106531143188, local kl: 0.0 global kl: 3.097102671745233e-05 valid reconstr loss: 2.4756057262420654
Saving best model with reconstruction loss 2.4756057
it: 600, train recon loss: 2.03037691116333, local kl: 0.0 global kl: 0.00018625961092766374 valid reconstr loss: 1.787994623184204
Saving best model with reconstruction loss 1.7879946
it: 700, train recon loss: 1.0893667936325073, local kl: 0.0 global kl: 0.00013093711459077895 valid reconstr loss: 1.6163274049758911
Saving best model with reconstruction loss 1.6163274
it: 800, train recon loss: 5.747884273529053, local kl: 0.0 global kl: 4.7159683163044974e-05 valid reconstr loss: 3.136697769165039
it: 900, train recon loss: 0.23539969325065613, local kl: 0.0 global kl: 0.00012777776282746345 valid reconstr loss: 19.641265869140625
it: 1000, train recon loss: 0.2378149926662445, local kl: 0.0 global kl: 3.777798337978311e-05 valid reconstr loss: 0.5775108933448792
Saving best model with reconstruction loss 0.5775109
it: 1100, train recon loss: 0.07110222429037094, local kl: 0.0 global kl: 0.00020192116789985448 valid reconstr loss: 0.2829667925834656
Saving best model with reconstruction loss 0.2829668
it: 1200, train recon loss: 0.019780369475483894, local kl: 0.0 global kl: 0.0003515102725941688 valid reconstr loss: -0.048754677176475525
Saving best model with reconstruction loss -0.048754677
it: 1300, train recon loss: -0.24220599234104156, local kl: 0.0 global kl: 0.0009784277062863111 valid reconstr loss: -0.01933913864195347
it: 1400, train recon loss: -0.3628520369529724, local kl: 0.0 global kl: 6.184837548062205e-05 valid reconstr loss: -0.08772636204957962
Saving best model with reconstruction loss -0.08772636
it: 1500, train recon loss: -0.20170456171035767, local kl: 0.0 global kl: 3.0392297048820183e-05 valid reconstr loss: -0.21255198121070862
Saving best model with reconstruction loss -0.21255198
it: 1600, train recon loss: 22.000438690185547, local kl: 0.0 global kl: 0.0007759145810268819 valid reconstr loss: 4.322342395782471
it: 1700, train recon loss: 1.500939130783081, local kl: 0.0 global kl: 0.0013788997894153 valid reconstr loss: 2.783869981765747
it: 1800, train recon loss: 1.0103598833084106, local kl: 0.0 global kl: 0.0005120888235978782 valid reconstr loss: 1.0783272981643677
it: 1900, train recon loss: 0.3853556513786316, local kl: 0.0 global kl: 5.012886322219856e-05 valid reconstr loss: 1.6454906463623047
it: 2000, train recon loss: 0.8570225834846497, local kl: 0.0 global kl: 0.000980094657279551 valid reconstr loss: 0.25819000601768494
it: 2100, train recon loss: -0.05651337280869484, local kl: 0.0 global kl: 0.0005933664506301284 valid reconstr loss: 0.284955233335495
it: 2200, train recon loss: -0.30465343594551086, local kl: 0.0 global kl: 0.00010272105282638222 valid reconstr loss: -0.2269895076751709
Saving best model with reconstruction loss -0.22698951
it: 2300, train recon loss: 0.13723675906658173, local kl: 0.0 global kl: 2.6712281396612525e-05 valid reconstr loss: -0.16009315848350525
it: 2400, train recon loss: -0.34942150115966797, local kl: 0.0 global kl: 4.090648508281447e-05 valid reconstr loss: 0.03845088183879852
it: 2500, train recon loss: -0.3405052423477173, local kl: 0.0 global kl: 2.0262614270905033e-05 valid reconstr loss: -0.09986056387424469
it: 2600, train recon loss: -0.3560425937175751, local kl: 0.0 global kl: 0.00030660448828712106 valid reconstr loss: -0.3821004033088684
Saving best model with reconstruction loss -0.3821004
it: 2700, train recon loss: 3.9721717834472656, local kl: 0.0 global kl: 0.00011487417941680178 valid reconstr loss: 7.828165054321289
it: 2800, train recon loss: -0.6955471038818359, local kl: 0.0 global kl: 0.0001928347919601947 valid reconstr loss: -0.20949874818325043
it: 2900, train recon loss: 0.04226408526301384, local kl: 0.0 global kl: 0.00023632447118870914 valid reconstr loss: -0.6790309548377991
Saving best model with reconstruction loss -0.67903095
it: 3000, train recon loss: 3.334486484527588, local kl: 0.0 global kl: 0.00027848605532199144 valid reconstr loss: 3.415806531906128
it: 3100, train recon loss: -0.9114851951599121, local kl: 0.0 global kl: 0.0007191433687694371 valid reconstr loss: -0.8276762962341309
Saving best model with reconstruction loss -0.8276763
it: 3200, train recon loss: 1342.6302490234375, local kl: 0.0 global kl: 4.059202183270827e-06 valid reconstr loss: 173.02639770507812
it: 3300, train recon loss: -0.40216052532196045, local kl: 0.0 global kl: 6.69372093398124e-05 valid reconstr loss: -0.5772181153297424
it: 3400, train recon loss: -0.6555606722831726, local kl: 0.0 global kl: 0.001969490200281143 valid reconstr loss: 2.634580612182617
it: 3500, train recon loss: -1.136548638343811, local kl: 0.0 global kl: 4.634785000234842e-05 valid reconstr loss: -0.33849138021469116
it: 3600, train recon loss: -0.633752703666687, local kl: 0.0 global kl: 1.3734877484239405e-06 valid reconstr loss: 1.1160959005355835
it: 3700, train recon loss: -1.0997494459152222, local kl: 0.0 global kl: 0.00017558285617269576 valid reconstr loss: -0.9627536535263062
Saving best model with reconstruction loss -0.96275365
it: 3800, train recon loss: -0.5189030766487122, local kl: 0.0 global kl: 2.5060700863832608e-05 valid reconstr loss: 724.904052734375
it: 3900, train recon loss: -1.040330410003662, local kl: 0.0 global kl: 6.002078407618683e-06 valid reconstr loss: -0.9803085923194885
Saving best model with reconstruction loss -0.9803086
it: 4000, train recon loss: -1.2465609312057495, local kl: 0.0 global kl: 3.274237678851932e-05 valid reconstr loss: 3.2651827335357666
it: 4100, train recon loss: 233.37477111816406, local kl: 0.0 global kl: 5.1555922254920006e-05 valid reconstr loss: 735.1744384765625
it: 4200, train recon loss: -1.045563817024231, local kl: 0.0 global kl: 1.3221731933299452e-05 valid reconstr loss: -0.5078142881393433
it: 4300, train recon loss: -1.070825457572937, local kl: 0.0 global kl: 0.00010376783029641956 valid reconstr loss: -1.040653109550476
Saving best model with reconstruction loss -1.0406531
it: 4400, train recon loss: -0.7819061279296875, local kl: 0.0 global kl: 1.8457807527738623e-05 valid reconstr loss: -0.9897986650466919
it: 4500, train recon loss: -1.1271967887878418, local kl: 0.0 global kl: 0.00014644567272625864 valid reconstr loss: 3574.138671875
it: 4600, train recon loss: -1.5063273906707764, local kl: 0.0 global kl: 0.0014615071704611182 valid reconstr loss: -0.8626999258995056
it: 4700, train recon loss: -1.065288782119751, local kl: 0.0 global kl: 1.0899791959673166e-05 valid reconstr loss: -1.121979832649231
Saving best model with reconstruction loss -1.1219798
it: 4800, train recon loss: 3.0052404403686523, local kl: 0.0 global kl: 0.000721295946277678 valid reconstr loss: 3.3247647285461426
it: 4900, train recon loss: 3.037334442138672, local kl: 0.0 global kl: 2.867905912751212e-11 valid reconstr loss: 3.1641740798950195
it: 5000, train recon loss: 3.0485012531280518, local kl: 0.0 global kl: 4.084165894369107e-09 valid reconstr loss: 3.116520643234253
it: 5100, train recon loss: 2.9821181297302246, local kl: 0.0 global kl: 3.1658744636259284e-12 valid reconstr loss: 3.122853994369507
it: 5200, train recon loss: 3.1054017543792725, local kl: 0.0 global kl: 1.1626819299004332e-11 valid reconstr loss: 3.0958354473114014
it: 5300, train recon loss: 3.0861778259277344, local kl: 0.0 global kl: 7.39123172388556e-12 valid reconstr loss: 3.065411329269409
it: 5400, train recon loss: 3.0625061988830566, local kl: 0.0 global kl: 9.909415302211677e-12 valid reconstr loss: 2.981741189956665
it: 5500, train recon loss: 2.7455735206604004, local kl: 0.0 global kl: 8.05663435271553e-12 valid reconstr loss: 3.0200376510620117
it: 5600, train recon loss: 3.0969481468200684, local kl: 0.0 global kl: 5.3104313481322496e-12 valid reconstr loss: 3.051419973373413
it: 5700, train recon loss: 2.9319682121276855, local kl: 0.0 global kl: 1.8102478066897576e-12 valid reconstr loss: 3.2132811546325684
it: 5800, train recon loss: 2.915457248687744, local kl: 0.0 global kl: 3.325821605962287e-12 valid reconstr loss: 3.0329086780548096
it: 5900, train recon loss: 2.9395997524261475, local kl: 0.0 global kl: 1.0476454773145072e-11 valid reconstr loss: 2.991318941116333
it: 6000, train recon loss: 2.9221084117889404, local kl: 0.0 global kl: 5.378837566322181e-12 valid reconstr loss: 3.0153591632843018
it: 6100, train recon loss: 2.987313985824585, local kl: 0.0 global kl: 2.47300509063908e-12 valid reconstr loss: 3.0364737510681152
it: 6200, train recon loss: 3.024320363998413, local kl: 0.0 global kl: 1.934685701573957e-12 valid reconstr loss: 2.98452091217041
it: 6300, train recon loss: 3.1472344398498535, local kl: 0.0 global kl: 3.0061190082336653e-12 valid reconstr loss: 2.9711074829101562
it: 6400, train recon loss: 2.8398983478546143, local kl: 0.0 global kl: 2.937519585216597e-12 valid reconstr loss: 2.9319589138031006
it: 6500, train recon loss: 2.916703701019287, local kl: 0.0 global kl: 8.969087625376737e-12 valid reconstr loss: 2.993753433227539
it: 6600, train recon loss: 2.7517189979553223, local kl: 0.0 global kl: 3.2214161893584503e-12 valid reconstr loss: 2.9804017543792725
it: 6700, train recon loss: 2.9398019313812256, local kl: 0.0 global kl: 4.205647115285149e-12 valid reconstr loss: 2.9725852012634277
it: 6800, train recon loss: 2.908334493637085, local kl: 0.0 global kl: 1.7926354844988168e-12 valid reconstr loss: 2.9763870239257812
it: 6900, train recon loss: 2.9562299251556396, local kl: 0.0 global kl: 3.1527475942427774e-12 valid reconstr loss: 2.981706142425537
it: 7000, train recon loss: 2.9804365634918213, local kl: 0.0 global kl: 5.290066995566889e-12 valid reconstr loss: 2.9555654525756836
it: 7100, train recon loss: 2.9091475009918213, local kl: 0.0 global kl: 1.5603088482710192e-12 valid reconstr loss: 2.9477829933166504
it: 7200, train recon loss: 2.837125062942505, local kl: 0.0 global kl: 1.024137979283024e-11 valid reconstr loss: 2.8903262615203857
it: 7300, train recon loss: 3.121633291244507, local kl: 0.0 global kl: 2.2970555579177043e-12 valid reconstr loss: 2.9032676219940186
it: 7400, train recon loss: 3.1874959468841553, local kl: 0.0 global kl: 5.0894969662318434e-12 valid reconstr loss: 2.95261812210083
it: 7500, train recon loss: 2.9727985858917236, local kl: 0.0 global kl: 4.534418847346178e-12 valid reconstr loss: 2.940411329269409
it: 7600, train recon loss: 3.049224615097046, local kl: 0.0 global kl: 5.399164622332808e-12 valid reconstr loss: 2.932044506072998
it: 7700, train recon loss: 2.786759853363037, local kl: 0.0 global kl: 3.607195054808332e-12 valid reconstr loss: 2.942007303237915
it: 7800, train recon loss: 2.877431631088257, local kl: 0.0 global kl: 5.828808356117543e-12 valid reconstr loss: 2.9384260177612305
it: 7900, train recon loss: 2.75622296333313, local kl: 0.0 global kl: 6.9774091018826034e-12 valid reconstr loss: 2.9201369285583496
it: 8000, train recon loss: 2.8956751823425293, local kl: 0.0 global kl: 1.357482182218206e-11 valid reconstr loss: 2.924245834350586
it: 8100, train recon loss: 2.8577866554260254, local kl: 0.0 global kl: 5.778452803056888e-12 valid reconstr loss: 2.902543544769287
it: 8200, train recon loss: 2.8871681690216064, local kl: 0.0 global kl: 6.514433523868712e-12 valid reconstr loss: 2.9290449619293213
it: 8300, train recon loss: 2.9401848316192627, local kl: 0.0 global kl: 1.2250670096414229e-11 valid reconstr loss: 2.9089879989624023
it: 8400, train recon loss: 2.6816508769989014, local kl: 0.0 global kl: 2.965723586850766e-12 valid reconstr loss: 3.042982816696167
it: 8500, train recon loss: 3.024717092514038, local kl: 0.0 global kl: 2.6151845446131805e-12 valid reconstr loss: 2.9243133068084717
it: 8600, train recon loss: 3.1380300521850586, local kl: 0.0 global kl: 3.2177205778333162e-12 valid reconstr loss: 2.92868709564209
it: 8700, train recon loss: 2.8865184783935547, local kl: 0.0 global kl: 2.5118818483549354e-11 valid reconstr loss: 2.9396541118621826
it: 8800, train recon loss: 2.7038300037384033, local kl: 0.0 global kl: 4.665677566517701e-12 valid reconstr loss: 3.031081199645996
it: 8900, train recon loss: 2.7383861541748047, local kl: 0.0 global kl: 1.9523462707615735e-11 valid reconstr loss: 3.0994932651519775
it: 9000, train recon loss: 2.6868526935577393, local kl: 0.0 global kl: 9.992521133456167e-13 valid reconstr loss: 2.931532144546509
it: 9100, train recon loss: 3.0800952911376953, local kl: 0.0 global kl: 5.004142933362488e-12 valid reconstr loss: 2.9986112117767334
it: 9200, train recon loss: 2.8628156185150146, local kl: 0.0 global kl: 1.3248679497229743e-12 valid reconstr loss: 2.9362730979919434
it: 9300, train recon loss: 2.9048116207122803, local kl: 0.0 global kl: 9.959563555816953e-12 valid reconstr loss: 2.8906688690185547
it: 9400, train recon loss: 2.9986233711242676, local kl: 0.0 global kl: 6.716014029628514e-12 valid reconstr loss: 2.915255308151245
it: 9500, train recon loss: 3.0020806789398193, local kl: 0.0 global kl: 5.718885000977059e-12 valid reconstr loss: 2.93237566947937
it: 9600, train recon loss: 2.8610150814056396, local kl: 0.0 global kl: 1.2791337000023217e-12 valid reconstr loss: 2.930579662322998
it: 9700, train recon loss: 2.933994770050049, local kl: 0.0 global kl: 4.232089505229464e-12 valid reconstr loss: 2.9281368255615234
it: 9800, train recon loss: 3.0737504959106445, local kl: 0.0 global kl: 3.4598753983344777e-12 valid reconstr loss: 2.9195549488067627
it: 9900, train recon loss: 2.657559871673584, local kl: 0.0 global kl: 4.611532076342906e-12 valid reconstr loss: 2.9183106422424316
beta 2.0 temperature 1.0
it: 0, train recon loss: 473.41619873046875, local kl: 0.0 global kl: 0.04242920130491257 valid reconstr loss: 348.8146667480469
Saving best model with reconstruction loss 348.81467
it: 100, train recon loss: 4.106132984161377, local kl: 0.0 global kl: 0.0018257724586874247 valid reconstr loss: 4.137568950653076
Saving best model with reconstruction loss 4.137569
it: 200, train recon loss: 3.730304479598999, local kl: 0.0 global kl: 0.00013898310135118663 valid reconstr loss: 3.6578452587127686
Saving best model with reconstruction loss 3.6578453
it: 300, train recon loss: 2.685791254043579, local kl: 0.0 global kl: 0.00012898133718408644 valid reconstr loss: 2.908521890640259
Saving best model with reconstruction loss 2.908522
it: 400, train recon loss: 2.983893394470215, local kl: 0.0 global kl: 0.0011521440465003252 valid reconstr loss: 3.1040737628936768
it: 500, train recon loss: 2.3911399841308594, local kl: 0.0 global kl: 0.0003930552920792252 valid reconstr loss: 2.717578411102295
Saving best model with reconstruction loss 2.7175784
it: 600, train recon loss: 2.6678874492645264, local kl: 0.0 global kl: 0.002559622749686241 valid reconstr loss: 3.288088083267212
it: 700, train recon loss: 1.9409419298171997, local kl: 0.0 global kl: 0.00047564823762513697 valid reconstr loss: 1.9561272859573364
Saving best model with reconstruction loss 1.9561273
it: 800, train recon loss: 15.963857650756836, local kl: 0.0 global kl: 0.00012830100604332983 valid reconstr loss: 2.2107083797454834
it: 900, train recon loss: 1.9262378215789795, local kl: 0.0 global kl: 7.303111487999558e-05 valid reconstr loss: 1.8636947870254517
Saving best model with reconstruction loss 1.8636948
it: 1000, train recon loss: 1.7116498947143555, local kl: 0.0 global kl: 1.81897576112533e-05 valid reconstr loss: 1.5741420984268188
Saving best model with reconstruction loss 1.5741421
it: 1100, train recon loss: 1.7542171478271484, local kl: 0.0 global kl: 6.387829489540309e-05 valid reconstr loss: 3.4368112087249756
it: 1200, train recon loss: 5.7176361083984375, local kl: 0.0 global kl: 4.202689160592854e-05 valid reconstr loss: 1.9162728786468506
it: 1300, train recon loss: 1.2040232419967651, local kl: 0.0 global kl: 0.00013327255146577954 valid reconstr loss: 1.2266403436660767
Saving best model with reconstruction loss 1.2266403
it: 1400, train recon loss: 1.3560935258865356, local kl: 0.0 global kl: 0.0004373026604298502 valid reconstr loss: 395.7652893066406
it: 1500, train recon loss: 1.3029745817184448, local kl: 0.0 global kl: 0.00027870465419255197 valid reconstr loss: 1.8015302419662476
it: 1600, train recon loss: 3.059861421585083, local kl: 0.0 global kl: 0.00017312643467448652 valid reconstr loss: 1.2433570623397827
it: 1700, train recon loss: 0.7985058426856995, local kl: 0.0 global kl: 1.876308124337811e-05 valid reconstr loss: 1.274170160293579
it: 1800, train recon loss: 0.7252258658409119, local kl: 0.0 global kl: 4.0293038182426244e-05 valid reconstr loss: 0.44985532760620117
Saving best model with reconstruction loss 0.44985533
it: 1900, train recon loss: 0.4255373477935791, local kl: 0.0 global kl: 0.00010926055256277323 valid reconstr loss: 0.47577983140945435
it: 2000, train recon loss: 0.47316470742225647, local kl: 0.0 global kl: 6.531666440423578e-05 valid reconstr loss: 1299.3648681640625
it: 2100, train recon loss: 0.2873322069644928, local kl: 0.0 global kl: 2.9497099603759125e-05 valid reconstr loss: 88.67807006835938
it: 2200, train recon loss: 0.29411813616752625, local kl: 0.0 global kl: 0.0018994059646502137 valid reconstr loss: 0.35800787806510925
Saving best model with reconstruction loss 0.35800788
it: 2300, train recon loss: 0.17173896729946136, local kl: 0.0 global kl: 7.711887883488089e-05 valid reconstr loss: 0.1794721782207489
Saving best model with reconstruction loss 0.17947218
it: 2400, train recon loss: 0.5262936949729919, local kl: 0.0 global kl: 3.5905177355743945e-05 valid reconstr loss: 0.12042900174856186
Saving best model with reconstruction loss 0.120429
it: 2500, train recon loss: 0.3914509415626526, local kl: 0.0 global kl: 1.5787443771841936e-05 valid reconstr loss: 0.165385439991951
it: 2600, train recon loss: 0.24696925282478333, local kl: 0.0 global kl: 2.539840807003202e-06 valid reconstr loss: 0.09960008412599564
Saving best model with reconstruction loss 0.099600084
it: 2700, train recon loss: 2.121674060821533, local kl: 0.0 global kl: 0.00015332855400629342 valid reconstr loss: -0.0071776811964809895
Saving best model with reconstruction loss -0.007177681
it: 2800, train recon loss: 0.08974563330411911, local kl: 0.0 global kl: 4.014695150544867e-05 valid reconstr loss: 0.2756063938140869
it: 2900, train recon loss: 0.12591327726840973, local kl: 0.0 global kl: 5.7068642490776256e-05 valid reconstr loss: 0.006728004664182663
it: 3000, train recon loss: 3.8413352966308594, local kl: 0.0 global kl: 3.856794864987023e-05 valid reconstr loss: 3.81955623626709
it: 3100, train recon loss: 3.1283960342407227, local kl: 0.0 global kl: 2.4637786282255547e-06 valid reconstr loss: 3.1958494186401367
it: 3200, train recon loss: 3.5743038654327393, local kl: 0.0 global kl: 1.873263766327682e-09 valid reconstr loss: 3.279576301574707
it: 3300, train recon loss: 2.9886183738708496, local kl: 0.0 global kl: 7.690163172924258e-09 valid reconstr loss: 3.4160032272338867
it: 3400, train recon loss: 3.1557021141052246, local kl: 0.0 global kl: 6.146823494646014e-09 valid reconstr loss: 3.1737935543060303
it: 3500, train recon loss: 3.002372980117798, local kl: 0.0 global kl: 1.5164169919756887e-09 valid reconstr loss: 3.119764566421509
it: 3600, train recon loss: 3.2994468212127686, local kl: 0.0 global kl: 8.20984891181098e-10 valid reconstr loss: 3.18156099319458
it: 3700, train recon loss: 2.9439730644226074, local kl: 0.0 global kl: 8.033362308701442e-10 valid reconstr loss: 3.154550075531006
it: 3800, train recon loss: 3.2341485023498535, local kl: 0.0 global kl: 1.3083008287839704e-10 valid reconstr loss: 3.1230764389038086
it: 3900, train recon loss: 2.7986626625061035, local kl: 0.0 global kl: 4.236549722147487e-10 valid reconstr loss: 3.1819143295288086
it: 4000, train recon loss: 2.7119388580322266, local kl: 0.0 global kl: 2.795584319592592e-10 valid reconstr loss: 3.1646926403045654
it: 4100, train recon loss: 2.9449727535247803, local kl: 0.0 global kl: 2.3065389875842612e-11 valid reconstr loss: 3.1999659538269043
it: 4200, train recon loss: 3.123720169067383, local kl: 0.0 global kl: 7.168426963133356e-11 valid reconstr loss: 3.0988810062408447
it: 4300, train recon loss: 2.804212808609009, local kl: 0.0 global kl: 9.746888712802715e-11 valid reconstr loss: 3.1850671768188477
it: 4400, train recon loss: 3.046642541885376, local kl: 0.0 global kl: 7.926204137476134e-11 valid reconstr loss: 3.0869410037994385
it: 4500, train recon loss: 2.921335458755493, local kl: 0.0 global kl: 8.740014788288786e-12 valid reconstr loss: 3.1268672943115234
it: 4600, train recon loss: 3.026934862136841, local kl: 0.0 global kl: 1.8944373847418206e-11 valid reconstr loss: 3.0880019664764404
it: 4700, train recon loss: 3.0534298419952393, local kl: 0.0 global kl: 1.1596674141800545e-11 valid reconstr loss: 3.088618516921997
it: 4800, train recon loss: 2.887911081314087, local kl: 0.0 global kl: 9.943551190771949e-12 valid reconstr loss: 3.079914093017578
it: 4900, train recon loss: 2.92842435836792, local kl: 0.0 global kl: 6.487389618559103e-12 valid reconstr loss: 3.091498851776123
it: 5000, train recon loss: 2.9348785877227783, local kl: 0.0 global kl: 4.705966519247262e-12 valid reconstr loss: 3.05731201171875
it: 5100, train recon loss: 2.8963139057159424, local kl: 0.0 global kl: 4.18688174408377e-12 valid reconstr loss: 3.0560622215270996
it: 5200, train recon loss: 3.0224196910858154, local kl: 0.0 global kl: 4.108339102942837e-12 valid reconstr loss: 3.0673599243164062
it: 5300, train recon loss: 2.999591112136841, local kl: 0.0 global kl: 2.4380059603090753e-12 valid reconstr loss: 3.062183141708374
it: 5400, train recon loss: 3.0910685062408447, local kl: 0.0 global kl: 2.701567051663356e-12 valid reconstr loss: 3.051542043685913
it: 5500, train recon loss: 2.7438712120056152, local kl: 0.0 global kl: 4.175756095070593e-12 valid reconstr loss: 3.044067859649658
it: 5600, train recon loss: 3.101297378540039, local kl: 0.0 global kl: 3.876112972256429e-12 valid reconstr loss: 3.0199296474456787
it: 5700, train recon loss: 2.9848132133483887, local kl: 0.0 global kl: 3.068221674992766e-12 valid reconstr loss: 3.0237321853637695
it: 5800, train recon loss: 2.9074549674987793, local kl: 0.0 global kl: 3.062978039605757e-12 valid reconstr loss: 3.033041477203369
it: 5900, train recon loss: 2.9311654567718506, local kl: 0.0 global kl: 4.4893411904611824e-12 valid reconstr loss: 3.0217807292938232
it: 6000, train recon loss: 2.9175007343292236, local kl: 0.0 global kl: 1.6632173069353051e-12 valid reconstr loss: 3.0179319381713867
it: 6100, train recon loss: 2.98962664604187, local kl: 0.0 global kl: 1.221184611766013e-12 valid reconstr loss: 3.019099712371826
it: 6200, train recon loss: 3.0082144737243652, local kl: 0.0 global kl: 2.0370365550609337e-12 valid reconstr loss: 2.9889776706695557
it: 6300, train recon loss: 3.062274694442749, local kl: 0.0 global kl: 1.4166767802262226e-12 valid reconstr loss: 2.999570369720459
it: 6400, train recon loss: 2.8571372032165527, local kl: 0.0 global kl: 1.1455537039795072e-12 valid reconstr loss: 2.983428955078125
it: 6500, train recon loss: 2.919177770614624, local kl: 0.0 global kl: 1.9280646957470227e-12 valid reconstr loss: 2.995232343673706
it: 6600, train recon loss: 2.757932662963867, local kl: 0.0 global kl: 1.7506861845623112e-12 valid reconstr loss: 2.9897689819335938
it: 6700, train recon loss: 3.001145601272583, local kl: 0.0 global kl: 1.9679219191715003e-12 valid reconstr loss: 2.9758455753326416
it: 6800, train recon loss: 2.948993444442749, local kl: 0.0 global kl: 2.0726367670753643e-12 valid reconstr loss: 2.9902539253234863
it: 6900, train recon loss: 2.911680221557617, local kl: 0.0 global kl: 1.0962657647980989e-12 valid reconstr loss: 3.020735263824463
it: 7000, train recon loss: 3.0037190914154053, local kl: 0.0 global kl: 1.8565561414607412e-12 valid reconstr loss: 2.9693715572357178
it: 7100, train recon loss: 2.8486249446868896, local kl: 0.0 global kl: 2.6454047286073035e-12 valid reconstr loss: 3.001376152038574
it: 7200, train recon loss: 2.903456449508667, local kl: 0.0 global kl: 9.633859465382755e-13 valid reconstr loss: 2.995840072631836
it: 7300, train recon loss: 3.1859829425811768, local kl: 0.0 global kl: 1.6515825165819287e-12 valid reconstr loss: 2.978590488433838
it: 7400, train recon loss: 3.1786746978759766, local kl: 0.0 global kl: 1.1668377852477874e-12 valid reconstr loss: 2.992344856262207
it: 7500, train recon loss: 3.0147266387939453, local kl: 0.0 global kl: 9.440531039198174e-13 valid reconstr loss: 2.964717388153076
it: 7600, train recon loss: 3.1025545597076416, local kl: 0.0 global kl: 1.3656764521335907e-12 valid reconstr loss: 2.988548755645752
it: 7700, train recon loss: 2.8148226737976074, local kl: 0.0 global kl: 1.1181598182086194e-12 valid reconstr loss: 2.9915950298309326
it: 7800, train recon loss: 2.880678415298462, local kl: 0.0 global kl: 1.2310867386275404e-12 valid reconstr loss: 2.991375684738159
it: 7900, train recon loss: 2.7801568508148193, local kl: 0.0 global kl: 1.5695538401958031e-12 valid reconstr loss: 2.9817676544189453
it: 8000, train recon loss: 2.923403263092041, local kl: 0.0 global kl: 1.5931631014431957e-12 valid reconstr loss: 3.0232744216918945
it: 8100, train recon loss: 2.9164280891418457, local kl: 0.0 global kl: 1.4817088528351618e-12 valid reconstr loss: 2.9998185634613037
it: 8200, train recon loss: 2.900041341781616, local kl: 0.0 global kl: 1.7555456871196085e-12 valid reconstr loss: 2.9847733974456787
it: 8300, train recon loss: 2.9858381748199463, local kl: 0.0 global kl: 1.1097003307578013e-12 valid reconstr loss: 2.986499547958374
it: 8400, train recon loss: 2.7037601470947266, local kl: 0.0 global kl: 1.3366323022359627e-12 valid reconstr loss: 2.989020347595215
it: 8500, train recon loss: 3.0343642234802246, local kl: 0.0 global kl: 1.588637533155024e-12 valid reconstr loss: 2.9940764904022217
it: 8600, train recon loss: 3.1508426666259766, local kl: 0.0 global kl: 1.6050153827523728e-12 valid reconstr loss: 2.972618579864502
it: 8700, train recon loss: 2.8828108310699463, local kl: 0.0 global kl: 8.967135844625829e-13 valid reconstr loss: 2.9861884117126465
it: 8800, train recon loss: 2.7213685512542725, local kl: 0.0 global kl: 1.7385669726782682e-12 valid reconstr loss: 2.977853775024414
it: 8900, train recon loss: 2.7911343574523926, local kl: 0.0 global kl: 9.452964669712238e-13 valid reconstr loss: 2.975092887878418
it: 9000, train recon loss: 2.735618829727173, local kl: 0.0 global kl: 9.825671092728028e-13 valid reconstr loss: 2.9616146087646484
it: 9100, train recon loss: 3.1025919914245605, local kl: 0.0 global kl: 1.2784504357932214e-12 valid reconstr loss: 2.985675573348999
it: 9200, train recon loss: 2.8934762477874756, local kl: 0.0 global kl: 1.26257706546673e-12 valid reconstr loss: 2.9854941368103027
it: 9300, train recon loss: 2.9707460403442383, local kl: 0.0 global kl: 1.128701082250827e-12 valid reconstr loss: 2.9622185230255127
it: 9400, train recon loss: 3.033576488494873, local kl: 0.0 global kl: 1.2237278248020123e-12 valid reconstr loss: 2.982511520385742
it: 9500, train recon loss: 3.0133838653564453, local kl: 0.0 global kl: 8.849209342728925e-13 valid reconstr loss: 2.9801337718963623
it: 9600, train recon loss: 2.896315336227417, local kl: 0.0 global kl: 9.885813955640144e-13 valid reconstr loss: 2.994873285293579
it: 9700, train recon loss: 2.9339306354522705, local kl: 0.0 global kl: 1.321921196638376e-12 valid reconstr loss: 2.9851372241973877
it: 9800, train recon loss: 3.031851291656494, local kl: 0.0 global kl: 1.7875145807977333e-12 valid reconstr loss: 2.9664995670318604
it: 9900, train recon loss: 2.716717481613159, local kl: 0.0 global kl: 1.554644976121955e-12 valid reconstr loss: 2.9684958457946777
beta 2.0 temperature 2.0
it: 0, train recon loss: 438.3956604003906, local kl: 0.0 global kl: 0.02388199046254158 valid reconstr loss: 108.70185089111328
Saving best model with reconstruction loss 108.70185
it: 100, train recon loss: 3.685605049133301, local kl: 0.0 global kl: 0.0007641698466613889 valid reconstr loss: 4.009809970855713
Saving best model with reconstruction loss 4.00981
it: 200, train recon loss: 3.8874778747558594, local kl: 0.0 global kl: 1.7991172853726312e-06 valid reconstr loss: 3.9793901443481445
Saving best model with reconstruction loss 3.9793901
it: 300, train recon loss: 3.250917911529541, local kl: 0.0 global kl: 0.00022951056598685682 valid reconstr loss: 3.326263904571533
Saving best model with reconstruction loss 3.326264
it: 400, train recon loss: 2.7339706420898438, local kl: 0.0 global kl: 6.517449219245464e-05 valid reconstr loss: 3.1208767890930176
Saving best model with reconstruction loss 3.1208768
it: 500, train recon loss: 2.3236563205718994, local kl: 0.0 global kl: 0.0002883505658246577 valid reconstr loss: 2.727752208709717
Saving best model with reconstruction loss 2.7277522
it: 600, train recon loss: 2.5616626739501953, local kl: 0.0 global kl: 7.864427607273683e-05 valid reconstr loss: 2.6678593158721924
Saving best model with reconstruction loss 2.6678593
it: 700, train recon loss: 2.5044262409210205, local kl: 0.0 global kl: 6.134787122746843e-12 valid reconstr loss: 2.297321081161499
Saving best model with reconstruction loss 2.297321
it: 800, train recon loss: 1.9377995729446411, local kl: 0.0 global kl: 2.1974540242997165e-12 valid reconstr loss: 2.197413682937622
Saving best model with reconstruction loss 2.1974137
it: 900, train recon loss: 1.8396555185317993, local kl: 0.0 global kl: 2.3964163986534004e-13 valid reconstr loss: 2.0078611373901367
Saving best model with reconstruction loss 2.0078611
it: 1000, train recon loss: 1.8406693935394287, local kl: 0.0 global kl: 2.382538610845586e-12 valid reconstr loss: 1.6818886995315552
Saving best model with reconstruction loss 1.6818887
it: 1100, train recon loss: 1.7356325387954712, local kl: 0.0 global kl: 4.725109192804666e-13 valid reconstr loss: 1.7335458993911743
it: 1200, train recon loss: 1.8014298677444458, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.245959758758545
it: 1300, train recon loss: 1.892845630645752, local kl: 0.0 global kl: 5.275779813018744e-13 valid reconstr loss: 1.847152590751648
it: 1400, train recon loss: 1.2480238676071167, local kl: 0.0 global kl: 1.2031986518223903e-11 valid reconstr loss: 6.9047932624816895
it: 1500, train recon loss: 1.0481483936309814, local kl: 0.0 global kl: 1.1772416375066541e-11 valid reconstr loss: 1.3580399751663208
Saving best model with reconstruction loss 1.35804
it: 1600, train recon loss: 2.406759262084961, local kl: 0.0 global kl: 6.1816836371964e-12 valid reconstr loss: 1.3433598279953003
Saving best model with reconstruction loss 1.3433598
it: 1700, train recon loss: 1.3327934741973877, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.1983526945114136
Saving best model with reconstruction loss 1.1983527
it: 1800, train recon loss: 0.7345567345619202, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.7841075658798218
Saving best model with reconstruction loss 0.78410757
it: 1900, train recon loss: 0.2935350239276886, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.3308461010456085
Saving best model with reconstruction loss 0.3308461
it: 2000, train recon loss: 0.4208921492099762, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.2692890763282776
Saving best model with reconstruction loss 0.26928908
it: 2100, train recon loss: 0.07587160170078278, local kl: 0.0 global kl: 1.3664937237312103e-12 valid reconstr loss: 0.19339953362941742
Saving best model with reconstruction loss 0.19339953
it: 2200, train recon loss: 0.033360060304403305, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.2721254825592041
it: 2300, train recon loss: 0.06054913252592087, local kl: 0.0 global kl: 3.8308245464691026e-13 valid reconstr loss: 0.020764948800206184
Saving best model with reconstruction loss 0.020764949
it: 2400, train recon loss: 0.12577080726623535, local kl: 0.0 global kl: 1.346089906206771e-12 valid reconstr loss: 0.2885952293872833
it: 2500, train recon loss: 0.0752343162894249, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.19923043251037598
Saving best model with reconstruction loss -0.19923043
it: 2600, train recon loss: -0.06274230778217316, local kl: 0.0 global kl: 1.1888823259198489e-11 valid reconstr loss: 0.18359167873859406
it: 2700, train recon loss: 0.1923210769891739, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.27639612555503845
it: 2800, train recon loss: 3.6853296756744385, local kl: 0.0 global kl: 4.388253649345586e-12 valid reconstr loss: 3.7714273929595947
it: 2900, train recon loss: 3.12013578414917, local kl: 0.0 global kl: 3.901353545776587e-11 valid reconstr loss: 2.950625419616699
it: 3000, train recon loss: 1.3899766206741333, local kl: 0.0 global kl: 3.5012270860335093e-13 valid reconstr loss: 2.5573201179504395
it: 3100, train recon loss: 0.6397281289100647, local kl: 0.0 global kl: 0.0 valid reconstr loss: 21.614961624145508
it: 3200, train recon loss: -0.04745281860232353, local kl: 0.0 global kl: 2.1471713296250527e-13 valid reconstr loss: 0.422997385263443
it: 3300, train recon loss: -0.027440369129180908, local kl: 0.0 global kl: 2.974498902394429e-13 valid reconstr loss: 1470.104736328125
it: 3400, train recon loss: 8.227889060974121, local kl: 0.0 global kl: 7.751577157932843e-13 valid reconstr loss: 0.18870364129543304
it: 3500, train recon loss: -0.47568780183792114, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.2886410057544708
Saving best model with reconstruction loss -0.288641
it: 3600, train recon loss: -0.3278254270553589, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.339717835187912
Saving best model with reconstruction loss -0.33971784
it: 3700, train recon loss: -0.9633083343505859, local kl: 0.0 global kl: 1.0267342531733448e-12 valid reconstr loss: 0.4345667362213135
it: 3800, train recon loss: -0.40034446120262146, local kl: 0.0 global kl: 2.3104299723408772e-11 valid reconstr loss: -0.25168198347091675
it: 3900, train recon loss: -0.40468472242355347, local kl: 0.0 global kl: 7.884339882358038e-13 valid reconstr loss: -0.4569620192050934
Saving best model with reconstruction loss -0.45696202
it: 4000, train recon loss: -0.958814263343811, local kl: 0.0 global kl: 6.97043811559439e-11 valid reconstr loss: -0.38173526525497437
it: 4100, train recon loss: 3.8756628036499023, local kl: 0.0 global kl: 7.940457180666272e-10 valid reconstr loss: 3.7045705318450928
it: 4200, train recon loss: 0.17527706921100616, local kl: 0.0 global kl: 2.1049828546892968e-13 valid reconstr loss: -0.7836927771568298
Saving best model with reconstruction loss -0.7836928
it: 4300, train recon loss: -0.6093140244483948, local kl: 0.0 global kl: 9.14601727686204e-12 valid reconstr loss: -0.2742437720298767
it: 4400, train recon loss: 1110098.75, local kl: 0.0 global kl: 8.14527112424912e-10 valid reconstr loss: 3.944262742996216
it: 4500, train recon loss: -0.7299113869667053, local kl: 0.0 global kl: 1.1682488310071903e-11 valid reconstr loss: -0.4712482690811157
it: 4600, train recon loss: 159.64822387695312, local kl: 0.0 global kl: 2.0003443346183758e-13 valid reconstr loss: 3803.21435546875
it: 4700, train recon loss: -0.3089148998260498, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.5126715898513794
it: 4800, train recon loss: -0.974757730960846, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.6317870020866394
it: 4900, train recon loss: -0.9859675765037537, local kl: 0.0 global kl: 8.921795247029962e-11 valid reconstr loss: 259.056884765625
it: 5000, train recon loss: 294.34417724609375, local kl: 0.0 global kl: 4.6109782658732e-11 valid reconstr loss: -0.8196624517440796
Saving best model with reconstruction loss -0.81966245
it: 5100, train recon loss: -0.33836764097213745, local kl: 0.0 global kl: 3.389388769647894e-11 valid reconstr loss: -0.1536961793899536
it: 5200, train recon loss: 0.005333862267434597, local kl: 0.0 global kl: 4.0971295800495966e-11 valid reconstr loss: -0.6371747851371765
it: 5300, train recon loss: 0.5584211349487305, local kl: 0.0 global kl: 0.0 valid reconstr loss: 292.9892272949219
it: 5400, train recon loss: 3.660154342651367, local kl: 0.0 global kl: 1.8839774185153146e-10 valid reconstr loss: 21.584178924560547
it: 5500, train recon loss: -0.8327533006668091, local kl: 0.0 global kl: 1.973608082517231e-11 valid reconstr loss: 0.12015421688556671
it: 5600, train recon loss: -1.112270712852478, local kl: 0.0 global kl: 1.7552626019323725e-13 valid reconstr loss: 0.6197397112846375
it: 5700, train recon loss: 1463.9739990234375, local kl: 0.0 global kl: 1.924189974022994e-13 valid reconstr loss: -0.4657357931137085
it: 5800, train recon loss: 338.9805908203125, local kl: 0.0 global kl: 8.285178099143309e-13 valid reconstr loss: 29.217296600341797
it: 5900, train recon loss: -1.115667700767517, local kl: 0.0 global kl: 9.975353876257032e-14 valid reconstr loss: 6.255167007446289
it: 6000, train recon loss: 2.5028510093688965, local kl: 0.0 global kl: 6.603051438958119e-14 valid reconstr loss: -0.6642089486122131
it: 6100, train recon loss: 1706.427734375, local kl: 0.0 global kl: 2.3929302983560774e-10 valid reconstr loss: 76.05484008789062
it: 6200, train recon loss: -0.8698297142982483, local kl: 0.0 global kl: 9.24603796303991e-12 valid reconstr loss: 228.24049377441406
it: 6300, train recon loss: -0.8490074872970581, local kl: 0.0 global kl: 7.28583859910259e-17 valid reconstr loss: -0.6558793783187866
it: 6400, train recon loss: -0.7259258031845093, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.45121636986732483
it: 6500, train recon loss: 4.084377765655518, local kl: 0.0 global kl: 0.0 valid reconstr loss: 776.029296875
it: 6600, train recon loss: -0.9591597318649292, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.6884571313858032
it: 6700, train recon loss: -1.0618189573287964, local kl: 0.0 global kl: 0.0 valid reconstr loss: 12009.31640625
it: 6800, train recon loss: -1.0075570344924927, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.11344198882579803
it: 6900, train recon loss: 1.7068830728530884, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.7011886239051819
it: 7000, train recon loss: -0.9928121566772461, local kl: 0.0 global kl: 3.3861802251067274e-15 valid reconstr loss: 12037.560546875
it: 7100, train recon loss: -0.8858652114868164, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.9995047450065613
Saving best model with reconstruction loss -0.99950475
it: 7200, train recon loss: -1.0087090730667114, local kl: 0.0 global kl: 0.0 valid reconstr loss: 744.0137939453125
it: 7300, train recon loss: -0.5920501947402954, local kl: 0.0 global kl: 0.0 valid reconstr loss: 159.1537322998047
it: 7400, train recon loss: -0.843347430229187, local kl: 0.0 global kl: 0.0 valid reconstr loss: 22.701961517333984
it: 7500, train recon loss: -1.1288998126983643, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.8206303119659424
it: 7600, train recon loss: 956.2858276367188, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.7313961386680603
it: 7700, train recon loss: -1.2489269971847534, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.6686680912971497
it: 7800, train recon loss: -1.022283673286438, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.5805781483650208
it: 7900, train recon loss: -0.5303409695625305, local kl: 0.0 global kl: 0.0 valid reconstr loss: 30.670385360717773
it: 8000, train recon loss: -1.1220481395721436, local kl: 0.0 global kl: 0.0 valid reconstr loss: 546.3582153320312
it: 8100, train recon loss: 3.437304973602295, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.0652778148651123
Saving best model with reconstruction loss -1.0652778
it: 8200, train recon loss: 52.352874755859375, local kl: 0.0 global kl: 1.7771513427522478e-12 valid reconstr loss: 204716.578125
it: 8300, train recon loss: -1.0435301065444946, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.9071870446205139
it: 8400, train recon loss: -1.2091749906539917, local kl: 0.0 global kl: 0.0 valid reconstr loss: 541.017333984375
it: 8500, train recon loss: -0.9661857485771179, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.129925012588501
Saving best model with reconstruction loss -1.129925
it: 8600, train recon loss: 8.38628101348877, local kl: 0.0 global kl: 6.033173960418026e-12 valid reconstr loss: -1.0311774015426636
it: 8700, train recon loss: 3.615013599395752, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.1855664253234863
Saving best model with reconstruction loss -1.1855664
it: 8800, train recon loss: -1.1584722995758057, local kl: 0.0 global kl: 0.0 valid reconstr loss: 8.97598934173584
it: 8900, train recon loss: -1.2353442907333374, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.1165881156921387
it: 9000, train recon loss: -1.3276445865631104, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.1025218963623047
it: 9100, train recon loss: 0.243483766913414, local kl: 0.0 global kl: 0.0 valid reconstr loss: 8.368810653686523
it: 9200, train recon loss: -1.1088216304779053, local kl: 0.0 global kl: 0.0 valid reconstr loss: 18.702938079833984
it: 9300, train recon loss: -0.7670968174934387, local kl: 0.0 global kl: 3.2370217617483377e-12 valid reconstr loss: -1.2694331407546997
Saving best model with reconstruction loss -1.2694331
it: 9400, train recon loss: -1.051276683807373, local kl: 0.0 global kl: 1.0236256287043943e-13 valid reconstr loss: 8.00633430480957
it: 9500, train recon loss: -1.0195894241333008, local kl: 0.0 global kl: 7.90106868819862e-12 valid reconstr loss: -0.8306186199188232
it: 9600, train recon loss: -0.0018796691438183188, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.6365747451782227
it: 9700, train recon loss: -1.2436637878417969, local kl: 0.0 global kl: 8.056735834038875e-12 valid reconstr loss: 117.19267272949219
it: 9800, train recon loss: -0.966148316860199, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.1364318132400513
it: 9900, train recon loss: -1.241819977760315, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.1543285846710205
beta 2.0 temperature 5.0
it: 0, train recon loss: 85140.796875, local kl: 0.0 global kl: 0.028951769694685936 valid reconstr loss: 437.13067626953125
Saving best model with reconstruction loss 437.13068
it: 100, train recon loss: 3.7604050636291504, local kl: 0.0 global kl: 0.0014947920572012663 valid reconstr loss: 3.9951508045196533
Saving best model with reconstruction loss 3.9951508
it: 200, train recon loss: 4.597578048706055, local kl: 0.0 global kl: 0.00019580460502766073 valid reconstr loss: 4.000144004821777
it: 300, train recon loss: 3.734633684158325, local kl: 0.0 global kl: 0.000213091800105758 valid reconstr loss: 3.741239547729492
Saving best model with reconstruction loss 3.7412395
it: 400, train recon loss: 2.977856159210205, local kl: 0.0 global kl: 0.0003102059999946505 valid reconstr loss: 4.056236267089844
it: 500, train recon loss: 2.851325750350952, local kl: 0.0 global kl: 0.00016262703866232187 valid reconstr loss: 2.9920339584350586
Saving best model with reconstruction loss 2.992034
it: 600, train recon loss: 1.957223653793335, local kl: 0.0 global kl: 0.0004354362899903208 valid reconstr loss: 2.4197475910186768
Saving best model with reconstruction loss 2.4197476
it: 700, train recon loss: 2.326871156692505, local kl: 0.0 global kl: 2.1538326677728037e-11 valid reconstr loss: 1.723851203918457
Saving best model with reconstruction loss 1.7238512
it: 800, train recon loss: 2.447885751724243, local kl: 0.0 global kl: 5.196412744545853e-12 valid reconstr loss: 22.5230770111084
it: 900, train recon loss: 1.4049279689788818, local kl: 0.0 global kl: 1.6880580266942502e-11 valid reconstr loss: 1.6841652393341064
Saving best model with reconstruction loss 1.6841652
it: 1000, train recon loss: 1.3421921730041504, local kl: 0.0 global kl: 6.068479052601106e-13 valid reconstr loss: 1.062984824180603
Saving best model with reconstruction loss 1.0629848
it: 1100, train recon loss: 0.8645076751708984, local kl: 0.0 global kl: 5.969003069594692e-12 valid reconstr loss: 0.7590396404266357
Saving best model with reconstruction loss 0.75903964
it: 1200, train recon loss: 0.794952392578125, local kl: 0.0 global kl: 1.558492224162933e-11 valid reconstr loss: 0.8426814079284668
it: 1300, train recon loss: 0.7557502388954163, local kl: 0.0 global kl: 4.902234174153364e-11 valid reconstr loss: 0.5678497552871704
Saving best model with reconstruction loss 0.56784976
it: 1400, train recon loss: 0.35438811779022217, local kl: 0.0 global kl: 6.0689370195987635e-12 valid reconstr loss: 0.3274955749511719
Saving best model with reconstruction loss 0.32749557
it: 1500, train recon loss: 0.19401296973228455, local kl: 0.0 global kl: 1.2621570455451092e-12 valid reconstr loss: 0.9769689440727234
it: 1600, train recon loss: 0.4866153597831726, local kl: 0.0 global kl: 2.240430063693566e-13 valid reconstr loss: 0.2127961814403534
Saving best model with reconstruction loss 0.21279618
it: 1700, train recon loss: 0.2383173108100891, local kl: 0.0 global kl: 1.2079226507921703e-13 valid reconstr loss: 0.2914642095565796
it: 1800, train recon loss: 0.5534586906433105, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.24583300948143005
it: 1900, train recon loss: 0.21340294182300568, local kl: 0.0 global kl: 1.140337824168114e-13 valid reconstr loss: 1.5511544942855835
it: 2000, train recon loss: 2.735257387161255, local kl: 0.0 global kl: 1.6986648199157628e-11 valid reconstr loss: 50.06181335449219
it: 2100, train recon loss: -0.15828116238117218, local kl: 0.0 global kl: 2.47452058843578e-12 valid reconstr loss: 304.1993408203125
it: 2200, train recon loss: -0.1390807330608368, local kl: 0.0 global kl: 8.175460308734728e-12 valid reconstr loss: 0.1629457026720047
Saving best model with reconstruction loss 0.1629457
it: 2300, train recon loss: 273.38958740234375, local kl: 0.0 global kl: 3.813325002988144e-11 valid reconstr loss: 0.41077831387519836
it: 2400, train recon loss: 67.25569152832031, local kl: 0.0 global kl: 2.3206458456270784e-13 valid reconstr loss: 0.08902468532323837
Saving best model with reconstruction loss 0.089024685
it: 2500, train recon loss: 0.06580882519483566, local kl: 0.0 global kl: 2.5870985909115518e-11 valid reconstr loss: -0.16020864248275757
Saving best model with reconstruction loss -0.16020864
it: 2600, train recon loss: -0.4668685495853424, local kl: 0.0 global kl: 1.777000768754533e-11 valid reconstr loss: 54.18556213378906
it: 2700, train recon loss: 26.768388748168945, local kl: 0.0 global kl: 6.672179475586404e-11 valid reconstr loss: 1.4559539556503296
it: 2800, train recon loss: -0.1943177878856659, local kl: 0.0 global kl: 6.249962353210847e-12 valid reconstr loss: -0.21570050716400146
Saving best model with reconstruction loss -0.2157005
it: 2900, train recon loss: -0.24974440038204193, local kl: 0.0 global kl: 1.8924556366428646e-11 valid reconstr loss: 0.651809811592102
it: 3000, train recon loss: -0.36591383814811707, local kl: 0.0 global kl: 1.769730195722019e-13 valid reconstr loss: -0.3742855191230774
Saving best model with reconstruction loss -0.37428552
it: 3100, train recon loss: -0.48540520668029785, local kl: 0.0 global kl: 7.587999673042134e-12 valid reconstr loss: -0.2428101748228073
it: 3200, train recon loss: -0.5679230093955994, local kl: 0.0 global kl: 5.914171929966017e-12 valid reconstr loss: 0.14323386549949646
it: 3300, train recon loss: 619.4994506835938, local kl: 0.0 global kl: 2.354338946020107e-12 valid reconstr loss: 0.3891688883304596
it: 3400, train recon loss: 40211.91015625, local kl: 0.0 global kl: 2.9348801167117244e-11 valid reconstr loss: 1.7155396938323975
it: 3500, train recon loss: 2.550461530685425, local kl: 0.0 global kl: 5.637101896383001e-12 valid reconstr loss: 0.9227864742279053
it: 3600, train recon loss: -0.2436048835515976, local kl: 0.0 global kl: 2.999417381133185e-11 valid reconstr loss: -0.530765950679779
Saving best model with reconstruction loss -0.53076595
it: 3700, train recon loss: -0.23582060635089874, local kl: 0.0 global kl: 1.13657305789161e-10 valid reconstr loss: 140.44943237304688
it: 3800, train recon loss: 5288.943359375, local kl: 0.0 global kl: 4.858086155579144e-12 valid reconstr loss: 131.13479614257812
it: 3900, train recon loss: -0.8694828748703003, local kl: 0.0 global kl: 1.3907985874084261e-11 valid reconstr loss: 406.6860656738281
it: 4000, train recon loss: -0.3712559640407562, local kl: 0.0 global kl: 3.7896942528536925e-13 valid reconstr loss: -0.5255964994430542
it: 4100, train recon loss: 105.02035522460938, local kl: 0.0 global kl: 9.865099709349678e-11 valid reconstr loss: 15590.125
it: 4200, train recon loss: -0.3285931944847107, local kl: 0.0 global kl: 7.105427357601002e-15 valid reconstr loss: -0.5844053030014038
Saving best model with reconstruction loss -0.5844053
it: 4300, train recon loss: -0.8498156666755676, local kl: 0.0 global kl: 2.596013681799292e-13 valid reconstr loss: 9.012273788452148
it: 4400, train recon loss: -0.849435567855835, local kl: 0.0 global kl: 7.340600349792226e-12 valid reconstr loss: 1.6330561637878418
it: 4500, train recon loss: 0.6604689955711365, local kl: 0.0 global kl: 2.303105622880608e-14 valid reconstr loss: 286.5027770996094
it: 4600, train recon loss: -0.7718663811683655, local kl: 0.0 global kl: 2.4913404672588513e-13 valid reconstr loss: -0.845072329044342
Saving best model with reconstruction loss -0.8450723
it: 4700, train recon loss: -0.7079839110374451, local kl: 0.0 global kl: 7.448819339117563e-12 valid reconstr loss: 4.704981803894043
it: 4800, train recon loss: -0.9413385987281799, local kl: 0.0 global kl: 7.053006182561328e-15 valid reconstr loss: -0.7601185441017151
it: 4900, train recon loss: -1.1480525732040405, local kl: 0.0 global kl: 3.1585845050585704e-14 valid reconstr loss: -0.7877874970436096
it: 5000, train recon loss: -0.6264027953147888, local kl: 0.0 global kl: 1.789542125596455e-11 valid reconstr loss: -0.7431290745735168
it: 5100, train recon loss: 15.76171875, local kl: 0.0 global kl: 5.545848780208473e-10 valid reconstr loss: 1503.2962646484375
it: 5200, train recon loss: -1.1390490531921387, local kl: 0.0 global kl: 4.275962917077436e-11 valid reconstr loss: -0.7851356267929077
it: 5300, train recon loss: -0.9125944972038269, local kl: 0.0 global kl: 2.0102142173072934e-11 valid reconstr loss: 140.93382263183594
it: 5400, train recon loss: -1.0114707946777344, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.9440481066703796
Saving best model with reconstruction loss -0.9440481
it: 5500, train recon loss: 20.509218215942383, local kl: 0.0 global kl: 1.101857494134606e-11 valid reconstr loss: 112.91773986816406
it: 5600, train recon loss: 4.839324474334717, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.48954352736473083
it: 5700, train recon loss: 10.806143760681152, local kl: 0.0 global kl: 1.138031422570629e-11 valid reconstr loss: 5.373775005340576
it: 5800, train recon loss: -0.9181203842163086, local kl: 0.0 global kl: 0.0 valid reconstr loss: 72.65264892578125
it: 5900, train recon loss: -1.0262792110443115, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.0038502216339111
Saving best model with reconstruction loss -1.0038502
it: 6000, train recon loss: -0.3154846131801605, local kl: 0.0 global kl: 3.867753339825697e-12 valid reconstr loss: -0.7981382012367249
it: 6100, train recon loss: -0.4315643012523651, local kl: 0.0 global kl: 6.457334666976067e-14 valid reconstr loss: -0.9513270258903503
it: 6200, train recon loss: -0.48925817012786865, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.1288824081420898
Saving best model with reconstruction loss -1.1288824
it: 6300, train recon loss: -0.9485753178596497, local kl: 0.0 global kl: 0.0 valid reconstr loss: 82.54092407226562
it: 6400, train recon loss: 461.7582092285156, local kl: 0.0 global kl: 1.31469002351281e-11 valid reconstr loss: 9.938773155212402
it: 6500, train recon loss: 3.835930347442627, local kl: 0.0 global kl: 0.0 valid reconstr loss: 22819.630859375
it: 6600, train recon loss: -1.0774462223052979, local kl: 0.0 global kl: 5.5566662382489085e-14 valid reconstr loss: -1.185401201248169
Saving best model with reconstruction loss -1.1854012
it: 6700, train recon loss: -0.41410696506500244, local kl: 0.0 global kl: 8.944234242136417e-14 valid reconstr loss: 666.5711059570312
it: 6800, train recon loss: -1.1783835887908936, local kl: 0.0 global kl: 9.09690118372497e-12 valid reconstr loss: -1.0034767389297485
it: 6900, train recon loss: -0.47504809498786926, local kl: 0.0 global kl: 1.2037093544137178e-11 valid reconstr loss: -0.8116359710693359
it: 7000, train recon loss: -0.9585418701171875, local kl: 0.0 global kl: 1.8485213360008856e-14 valid reconstr loss: -1.0054296255111694
it: 7100, train recon loss: -0.8426029086112976, local kl: 0.0 global kl: 2.78388423424758e-14 valid reconstr loss: 1.451566219329834
it: 7200, train recon loss: 3.09670090675354, local kl: 0.0 global kl: 3.2487960238691826e-11 valid reconstr loss: 3.117985486984253
it: 7300, train recon loss: 3.3032219409942627, local kl: 0.0 global kl: 2.685457758944132e-11 valid reconstr loss: 3.219027042388916
it: 7400, train recon loss: 3.3283791542053223, local kl: 0.0 global kl: 5.354122093598201e-13 valid reconstr loss: 3.1221001148223877
it: 7500, train recon loss: 3.1251790523529053, local kl: 0.0 global kl: 2.7598168975823123e-12 valid reconstr loss: 3.1217830181121826
it: 7600, train recon loss: 3.1828882694244385, local kl: 0.0 global kl: 1.1725143425622697e-11 valid reconstr loss: 3.1037333011627197
it: 7700, train recon loss: 2.9047133922576904, local kl: 0.0 global kl: 3.085618566212034e-11 valid reconstr loss: 3.084597587585449
it: 7800, train recon loss: 2.9916908740997314, local kl: 0.0 global kl: 2.4361849343401687e-13 valid reconstr loss: 3.070047378540039
it: 7900, train recon loss: 2.8206982612609863, local kl: 0.0 global kl: 1.3658518760450988e-12 valid reconstr loss: 3.131854772567749
it: 8000, train recon loss: 3.0264816284179688, local kl: 0.0 global kl: 8.425598760353203e-12 valid reconstr loss: 3.063819169998169
it: 8100, train recon loss: 2.9657464027404785, local kl: 0.0 global kl: 1.785822358046918e-12 valid reconstr loss: 3.06162166595459
it: 8200, train recon loss: 2.9382803440093994, local kl: 0.0 global kl: 1.320783066249831e-11 valid reconstr loss: 3.054178237915039
it: 8300, train recon loss: 3.0099077224731445, local kl: 0.0 global kl: 3.0071106195406205e-13 valid reconstr loss: 3.0185272693634033
it: 8400, train recon loss: 2.801119804382324, local kl: 0.0 global kl: 3.617695726321202e-11 valid reconstr loss: 2.989645004272461
it: 8500, train recon loss: 3.085616111755371, local kl: 0.0 global kl: 1.957823313192275e-11 valid reconstr loss: 3.003953218460083
it: 8600, train recon loss: 3.1643319129943848, local kl: 0.0 global kl: 1.244752998591503e-13 valid reconstr loss: 2.9673027992248535
it: 8700, train recon loss: 2.8967442512512207, local kl: 0.0 global kl: 9.117706589734098e-15 valid reconstr loss: 2.9430320262908936
it: 8800, train recon loss: 2.72815203666687, local kl: 0.0 global kl: 1.1560197243909442e-14 valid reconstr loss: 2.9988672733306885
it: 8900, train recon loss: 2.7300384044647217, local kl: 0.0 global kl: 2.6720162540905257e-14 valid reconstr loss: 3.0203299522399902
it: 9000, train recon loss: 2.7189242839813232, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.9931516647338867
it: 9100, train recon loss: 3.1088027954101562, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.0147852897644043
it: 9200, train recon loss: 2.8529982566833496, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.9434282779693604
it: 9300, train recon loss: 2.9126031398773193, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.983724594116211
it: 9400, train recon loss: 2.9592361450195312, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.945312261581421
it: 9500, train recon loss: 3.007513999938965, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.9841675758361816
it: 9600, train recon loss: 2.876600980758667, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.9130806922912598
it: 9700, train recon loss: 2.9516396522521973, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.9663755893707275
it: 9800, train recon loss: 3.014317512512207, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.952216386795044
it: 9900, train recon loss: 2.6736910343170166, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.9538307189941406
beta 2.0 temperature 1000.0
it: 0, train recon loss: 273.5511474609375, local kl: 0.0 global kl: 0.041017837822437286 valid reconstr loss: 55.957374572753906
Saving best model with reconstruction loss 55.957375
it: 100, train recon loss: 3.7080845832824707, local kl: 0.0 global kl: 0.0012589736143127084 valid reconstr loss: 4.361564636230469
Saving best model with reconstruction loss 4.3615646
it: 200, train recon loss: 3.996220588684082, local kl: 0.0 global kl: 3.104179631918669e-05 valid reconstr loss: 4.1198577880859375
Saving best model with reconstruction loss 4.119858
it: 300, train recon loss: 30.64837074279785, local kl: 0.0 global kl: 0.00017158262198790908 valid reconstr loss: 3.924497127532959
Saving best model with reconstruction loss 3.9244971
it: 400, train recon loss: 2.7081964015960693, local kl: 0.0 global kl: 3.469772491371259e-05 valid reconstr loss: 2.8117780685424805
Saving best model with reconstruction loss 2.811778
it: 500, train recon loss: 2.172191619873047, local kl: 0.0 global kl: 1.7331525103969625e-11 valid reconstr loss: 2.7507901191711426
Saving best model with reconstruction loss 2.75079
it: 600, train recon loss: 1.9652193784713745, local kl: 0.0 global kl: 4.308447695833273e-12 valid reconstr loss: 2.344057559967041
Saving best model with reconstruction loss 2.3440576
it: 700, train recon loss: 1.188845157623291, local kl: 0.0 global kl: 2.7113644662790648e-11 valid reconstr loss: 1.2285966873168945
Saving best model with reconstruction loss 1.2285967
it: 800, train recon loss: 1.0374807119369507, local kl: 0.0 global kl: 2.5085836186100607e-11 valid reconstr loss: 1.3995459079742432
it: 900, train recon loss: 0.8337496519088745, local kl: 0.0 global kl: 1.4427181671550215e-11 valid reconstr loss: 0.793329656124115
Saving best model with reconstruction loss 0.79332966
it: 1000, train recon loss: 0.9242186546325684, local kl: 0.0 global kl: 1.375344282905644e-12 valid reconstr loss: 0.9501907825469971
it: 1100, train recon loss: 0.4673355519771576, local kl: 0.0 global kl: 1.924033848910156e-12 valid reconstr loss: 0.45600709319114685
Saving best model with reconstruction loss 0.4560071
it: 1200, train recon loss: 0.3358215093612671, local kl: 0.0 global kl: 4.385492836933569e-12 valid reconstr loss: 0.5373203158378601
it: 1300, train recon loss: 0.49748969078063965, local kl: 0.0 global kl: 1.861427678662153e-11 valid reconstr loss: 1.2351878881454468
it: 1400, train recon loss: 19.083892822265625, local kl: 0.0 global kl: 2.714989344454466e-11 valid reconstr loss: 0.5924952030181885
it: 1500, train recon loss: 25.56846046447754, local kl: 0.0 global kl: 7.982900798730874e-12 valid reconstr loss: 0.5258238315582275
it: 1600, train recon loss: 0.8387285470962524, local kl: 0.0 global kl: 7.670364343681513e-12 valid reconstr loss: 0.5329538583755493
it: 1700, train recon loss: -0.12487191706895828, local kl: 0.0 global kl: 4.973799150320701e-14 valid reconstr loss: 0.1741560399532318
Saving best model with reconstruction loss 0.17415604
it: 1800, train recon loss: -0.0005113219958730042, local kl: 0.0 global kl: 2.8372365221129847e-11 valid reconstr loss: -0.0028083878569304943
Saving best model with reconstruction loss -0.0028083879
it: 1900, train recon loss: -0.15740668773651123, local kl: 0.0 global kl: 1.5365486660812167e-13 valid reconstr loss: 0.1279693841934204
it: 2000, train recon loss: 1311.112548828125, local kl: 0.0 global kl: 2.7133850721838826e-13 valid reconstr loss: 77.3113784790039
it: 2100, train recon loss: 0.15014280378818512, local kl: 0.0 global kl: 4.913847106990943e-13 valid reconstr loss: 14.688756942749023
it: 2200, train recon loss: 4297.64990234375, local kl: 0.0 global kl: 2.5672131087617345e-11 valid reconstr loss: 16.466934204101562
it: 2300, train recon loss: 3.7001795768737793, local kl: 0.0 global kl: 1.700782292779479e-10 valid reconstr loss: 3.940844774246216
it: 2400, train recon loss: 0.1568353921175003, local kl: 0.0 global kl: 2.5740801851137363e-11 valid reconstr loss: 0.1460268199443817
it: 2500, train recon loss: 12.877752304077148, local kl: 0.0 global kl: 8.480438573599258e-13 valid reconstr loss: 4.51823091506958
it: 2600, train recon loss: -0.5199601054191589, local kl: 0.0 global kl: 7.329688245238941e-11 valid reconstr loss: -0.16279542446136475
Saving best model with reconstruction loss -0.16279542
it: 2700, train recon loss: 0.13977347314357758, local kl: 0.0 global kl: 1.1403690491906815e-12 valid reconstr loss: -0.1924206167459488
Saving best model with reconstruction loss -0.19242062
it: 2800, train recon loss: -0.013317355886101723, local kl: 0.0 global kl: 1.122495933009171e-11 valid reconstr loss: -0.2181711345911026
Saving best model with reconstruction loss -0.21817113
it: 2900, train recon loss: 0.22446857392787933, local kl: 0.0 global kl: 7.793904410746677e-13 valid reconstr loss: 2.2089040279388428
it: 3000, train recon loss: -0.2681756913661957, local kl: 0.0 global kl: 3.780760426952412e-13 valid reconstr loss: 0.05840136855840683
it: 3100, train recon loss: -0.7276709079742432, local kl: 0.0 global kl: 1.8990364836213303e-13 valid reconstr loss: -0.11143407225608826
it: 3200, train recon loss: 0.021246347576379776, local kl: 0.0 global kl: 8.310019339319297e-14 valid reconstr loss: -0.44179439544677734
Saving best model with reconstruction loss -0.4417944
it: 3300, train recon loss: 0.15960729122161865, local kl: 0.0 global kl: 4.305222844891432e-12 valid reconstr loss: -0.02709687501192093
it: 3400, train recon loss: 9.00356388092041, local kl: 0.0 global kl: 1.261726834123067e-12 valid reconstr loss: -0.2386627048254013
it: 3500, train recon loss: -0.81775963306427, local kl: 0.0 global kl: 1.741651267650468e-11 valid reconstr loss: 0.008660347200930119
it: 3600, train recon loss: -0.4570850431919098, local kl: 0.0 global kl: 1.658234660695257e-11 valid reconstr loss: -0.18454307317733765
it: 3700, train recon loss: -0.7639191150665283, local kl: 0.0 global kl: 1.372080227213246e-11 valid reconstr loss: -0.418034166097641
it: 3800, train recon loss: 20.495006561279297, local kl: 0.0 global kl: 6.971534460831208e-12 valid reconstr loss: -0.1827535331249237
it: 3900, train recon loss: -0.5450685024261475, local kl: 0.0 global kl: 1.372790769949006e-13 valid reconstr loss: -0.16935451328754425
it: 4000, train recon loss: -1.0648720264434814, local kl: 0.0 global kl: 1.1560710722058332e-11 valid reconstr loss: -0.15438427031040192
it: 4100, train recon loss: 0.3499854803085327, local kl: 0.0 global kl: 3.113793944908849e-13 valid reconstr loss: 0.8120570182800293
it: 4200, train recon loss: -0.28470033407211304, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.5764948725700378
Saving best model with reconstruction loss -0.5764949
it: 4300, train recon loss: -0.21286901831626892, local kl: 0.0 global kl: 4.841177458914103e-11 valid reconstr loss: -0.548292875289917
it: 4400, train recon loss: -0.5955395102500916, local kl: 0.0 global kl: 1.648433473055988e-11 valid reconstr loss: -0.4527018070220947
it: 4500, train recon loss: -0.7568438649177551, local kl: 0.0 global kl: 2.109014352047467e-11 valid reconstr loss: 0.108382947742939
it: 4600, train recon loss: -0.9596541523933411, local kl: 0.0 global kl: 2.220515438189352e-12 valid reconstr loss: -0.6093368530273438
Saving best model with reconstruction loss -0.60933685
it: 4700, train recon loss: -0.7863677144050598, local kl: 0.0 global kl: 4.1300296516055823e-14 valid reconstr loss: -0.45604854822158813
it: 4800, train recon loss: -0.9945675134658813, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.5717729330062866
it: 4900, train recon loss: -0.7967756390571594, local kl: 0.0 global kl: 7.510658761589184e-14 valid reconstr loss: -0.7514516115188599
Saving best model with reconstruction loss -0.7514516
it: 5000, train recon loss: -0.7370600700378418, local kl: 0.0 global kl: 3.085461053320415e-11 valid reconstr loss: -0.6980941891670227
it: 5100, train recon loss: -0.9179794788360596, local kl: 0.0 global kl: 2.4929058817235727e-11 valid reconstr loss: -0.14707253873348236
it: 5200, train recon loss: -0.6658609509468079, local kl: 0.0 global kl: 2.5937474390502757e-11 valid reconstr loss: -0.6948496699333191
it: 5300, train recon loss: -0.9177632927894592, local kl: 0.0 global kl: 6.929179452441758e-14 valid reconstr loss: -0.4454303979873657
it: 5400, train recon loss: 0.01956975646317005, local kl: 0.0 global kl: 1.3947176746853529e-14 valid reconstr loss: -0.7610694766044617
Saving best model with reconstruction loss -0.7610695
it: 5500, train recon loss: -0.8673810362815857, local kl: 0.0 global kl: 9.8879238130678e-14 valid reconstr loss: -0.8046419620513916
Saving best model with reconstruction loss -0.80464196
it: 5600, train recon loss: -0.709989607334137, local kl: 0.0 global kl: 1.8514356714405267e-13 valid reconstr loss: -0.8070074319839478
Saving best model with reconstruction loss -0.80700743
it: 5700, train recon loss: -0.4306710362434387, local kl: 0.0 global kl: 4.1806558215284895e-11 valid reconstr loss: -0.9286119937896729
Saving best model with reconstruction loss -0.928612
it: 5800, train recon loss: -1.0090594291687012, local kl: 0.0 global kl: 1.2975137630988343e-10 valid reconstr loss: -0.7151694297790527
it: 5900, train recon loss: -1.1569406986236572, local kl: 0.0 global kl: 4.6407322429331543e-14 valid reconstr loss: -0.2594369649887085
it: 6000, train recon loss: -1.1903951168060303, local kl: 0.0 global kl: 1.3322676295501878e-13 valid reconstr loss: -0.23839685320854187
it: 6100, train recon loss: -0.9602544903755188, local kl: 0.0 global kl: 4.257504505195131e-12 valid reconstr loss: -1.0451292991638184
Saving best model with reconstruction loss -1.0451293
it: 6200, train recon loss: -0.931484580039978, local kl: 0.0 global kl: 2.0317081350640365e-14 valid reconstr loss: -0.7514935731887817
it: 6300, train recon loss: -0.9215602278709412, local kl: 0.0 global kl: 2.070333487980136e-11 valid reconstr loss: -0.2916601300239563
it: 6400, train recon loss: 3763.877197265625, local kl: 0.0 global kl: 5.877609510207549e-11 valid reconstr loss: 39.03963851928711
it: 6500, train recon loss: -0.5847756266593933, local kl: 0.0 global kl: 3.304012965998915e-11 valid reconstr loss: 3.430875778198242
it: 6600, train recon loss: -0.523804783821106, local kl: 0.0 global kl: 3.1308289294429414e-14 valid reconstr loss: -0.9772639274597168
it: 6700, train recon loss: -1.1405730247497559, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.0015956163406372
it: 6800, train recon loss: -0.6396287679672241, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.7421069741249084
it: 6900, train recon loss: -0.7716290950775146, local kl: 0.0 global kl: 7.292742798536977e-12 valid reconstr loss: -0.9729628562927246
it: 7000, train recon loss: -0.9158915877342224, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.5362305641174316
it: 7100, train recon loss: -0.8030939698219299, local kl: 0.0 global kl: 1.5437651157412802e-13 valid reconstr loss: 0.01599203422665596
it: 7200, train recon loss: 307.37200927734375, local kl: 0.0 global kl: 1.8206269825071786e-10 valid reconstr loss: 60.06147766113281
it: 7300, train recon loss: 726.627685546875, local kl: 0.0 global kl: 1.726951914804431e-13 valid reconstr loss: -1.0656282901763916
Saving best model with reconstruction loss -1.0656283
it: 7400, train recon loss: -0.6855469942092896, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.0145909786224365
it: 7500, train recon loss: -1.2274391651153564, local kl: 0.0 global kl: 2.7748636721725006e-14 valid reconstr loss: -1.1433961391448975
Saving best model with reconstruction loss -1.1433961
it: 7600, train recon loss: -0.8855065703392029, local kl: 0.0 global kl: 2.2343238370581275e-15 valid reconstr loss: -0.9157219529151917
it: 7700, train recon loss: -1.1429892778396606, local kl: 0.0 global kl: 9.120537658446892e-12 valid reconstr loss: -0.9392346143722534
it: 7800, train recon loss: -0.9531734585762024, local kl: 0.0 global kl: 4.370729472791268e-12 valid reconstr loss: 511.2565612792969
it: 7900, train recon loss: -0.9591718912124634, local kl: 0.0 global kl: 4.4809767008047174e-11 valid reconstr loss: -1.060954213142395
it: 8000, train recon loss: -1.121068000793457, local kl: 0.0 global kl: 3.008450433217291e-11 valid reconstr loss: 10.396621704101562
it: 8100, train recon loss: -1.1114702224731445, local kl: 0.0 global kl: 3.372441215176991e-13 valid reconstr loss: -0.7377600073814392
it: 8200, train recon loss: -1.1547046899795532, local kl: 0.0 global kl: 7.238237786921786e-13 valid reconstr loss: -0.9609857201576233
it: 8300, train recon loss: -1.0745106935501099, local kl: 0.0 global kl: 3.076705556992465e-14 valid reconstr loss: -0.4126821458339691
it: 8400, train recon loss: -1.0774939060211182, local kl: 0.0 global kl: 2.3965551765314785e-13 valid reconstr loss: -1.0322917699813843
it: 8500, train recon loss: -0.5315225124359131, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3286.4765625
it: 8600, train recon loss: -0.9741727113723755, local kl: 0.0 global kl: 9.159398933755725e-12 valid reconstr loss: 966.4612426757812
it: 8700, train recon loss: -0.972200334072113, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.1222740411758423
it: 8800, train recon loss: 11.710315704345703, local kl: 0.0 global kl: 0.0 valid reconstr loss: 14131.80078125
it: 8900, train recon loss: -1.1381701231002808, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.1016186475753784
it: 9000, train recon loss: 12.001160621643066, local kl: 0.0 global kl: 0.0 valid reconstr loss: 15.832006454467773
it: 9100, train recon loss: 20.523664474487305, local kl: 0.0 global kl: 1.6249588480343746e-13 valid reconstr loss: 30653816.0
it: 9200, train recon loss: -1.182982325553894, local kl: 0.0 global kl: 7.94031507211912e-13 valid reconstr loss: 4381.345703125
it: 9300, train recon loss: -1.0200873613357544, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.1367602348327637
it: 9400, train recon loss: -1.1854530572891235, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.0041841268539429
it: 9500, train recon loss: -0.8766471743583679, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.9047335386276245
it: 9600, train recon loss: -0.6932564377784729, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.5508652329444885
it: 9700, train recon loss: -1.0020772218704224, local kl: 0.0 global kl: 4.111266882489417e-12 valid reconstr loss: -1.096269965171814
it: 9800, train recon loss: -0.9270992279052734, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.3967649936676025
it: 9900, train recon loss: 2415.58935546875, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.264918565750122
beta 5.0 temperature 0.001
it: 0, train recon loss: 35619640.0, local kl: 0.0 global kl: 0.06147420406341553 valid reconstr loss: 1704753.125
Saving best model with reconstruction loss 1704753.1
it: 100, train recon loss: 3.616788387298584, local kl: 0.0 global kl: 0.0012255738256499171 valid reconstr loss: 3.8435585498809814
Saving best model with reconstruction loss 3.8435585
it: 200, train recon loss: 2.804863214492798, local kl: 0.0 global kl: 0.00016739127750042826 valid reconstr loss: 2.847165822982788
Saving best model with reconstruction loss 2.8471658
it: 300, train recon loss: 2.6702282428741455, local kl: 0.0 global kl: 0.00010804247722262517 valid reconstr loss: 4.011561870574951
it: 400, train recon loss: 2.1657896041870117, local kl: 0.0 global kl: 1.7623508028918877e-05 valid reconstr loss: 2.391528367996216
Saving best model with reconstruction loss 2.3915284
it: 500, train recon loss: 1.5556910037994385, local kl: 0.0 global kl: 7.198985986178741e-05 valid reconstr loss: 2.0016348361968994
Saving best model with reconstruction loss 2.0016348
it: 600, train recon loss: 1.2841193675994873, local kl: 0.0 global kl: 0.0001440294145140797 valid reconstr loss: 1.6129573583602905
Saving best model with reconstruction loss 1.6129574
it: 700, train recon loss: 1.3565549850463867, local kl: 0.0 global kl: 0.0002744827652350068 valid reconstr loss: 1.9762862920761108
it: 800, train recon loss: 47.40007400512695, local kl: 0.0 global kl: 0.0003891803207807243 valid reconstr loss: 134.6719207763672
it: 900, train recon loss: 0.9976480603218079, local kl: 0.0 global kl: 0.00014829685096628964 valid reconstr loss: 0.8634203672409058
Saving best model with reconstruction loss 0.86342037
it: 1000, train recon loss: 0.8353842496871948, local kl: 0.0 global kl: 0.00035384477814659476 valid reconstr loss: 132.4068603515625
it: 1100, train recon loss: 1.796445369720459, local kl: 0.0 global kl: 0.017405269667506218 valid reconstr loss: 5.939880847930908
it: 1200, train recon loss: 1.3704607486724854, local kl: 0.0 global kl: 7.529311551479623e-05 valid reconstr loss: 0.5072533488273621
Saving best model with reconstruction loss 0.50725335
it: 1300, train recon loss: 1.1224454641342163, local kl: 0.0 global kl: 0.0003562368801794946 valid reconstr loss: 3.3793399333953857
it: 1400, train recon loss: 0.0038321304600685835, local kl: 0.0 global kl: 6.956214929232374e-05 valid reconstr loss: 0.22754783928394318
Saving best model with reconstruction loss 0.22754784
it: 1500, train recon loss: 13.572121620178223, local kl: 0.0 global kl: 0.00013534261961467564 valid reconstr loss: 14.744488716125488
it: 1600, train recon loss: 0.4027003347873688, local kl: 0.0 global kl: 0.0002924629661720246 valid reconstr loss: 0.3009091913700104
it: 1700, train recon loss: 0.3443690240383148, local kl: 0.0 global kl: 7.451070996467024e-05 valid reconstr loss: 189.706787109375
it: 1800, train recon loss: 0.293920636177063, local kl: 0.0 global kl: 3.914283661288209e-05 valid reconstr loss: 1127.3819580078125
it: 1900, train recon loss: -0.23750613629817963, local kl: 0.0 global kl: 4.218520552967675e-05 valid reconstr loss: 31.98974609375
it: 2000, train recon loss: 4.309775352478027, local kl: 0.0 global kl: 2.3986611267901026e-05 valid reconstr loss: 2.8200912475585938
it: 2100, train recon loss: 1.2480831146240234, local kl: 0.0 global kl: 0.000433244596933946 valid reconstr loss: 1.1328212022781372
it: 2200, train recon loss: -0.017542805522680283, local kl: 0.0 global kl: 0.0001348337100353092 valid reconstr loss: 1.8876450061798096
it: 2300, train recon loss: 0.8757598400115967, local kl: 0.0 global kl: 0.0002888249000534415 valid reconstr loss: 0.613294780254364
it: 2400, train recon loss: 1.8396947383880615, local kl: 0.0 global kl: 1.6420674000983126e-05 valid reconstr loss: 911.0170288085938
it: 2500, train recon loss: -0.27242207527160645, local kl: 0.0 global kl: 3.4123695513699204e-06 valid reconstr loss: -0.29948827624320984
Saving best model with reconstruction loss -0.29948828
it: 2600, train recon loss: -0.6247120499610901, local kl: 0.0 global kl: 4.964626714354381e-05 valid reconstr loss: -0.4020543694496155
Saving best model with reconstruction loss -0.40205437
it: 2700, train recon loss: -0.350920170545578, local kl: 0.0 global kl: 1.4615834516007453e-05 valid reconstr loss: -0.4222639501094818
Saving best model with reconstruction loss -0.42226395
it: 2800, train recon loss: -0.10168977081775665, local kl: 0.0 global kl: 4.955900749337161e-06 valid reconstr loss: 314.8345947265625
it: 2900, train recon loss: 2032.6337890625, local kl: 0.0 global kl: 0.0005476981750689447 valid reconstr loss: 904.0816040039062
it: 3000, train recon loss: 0.003475734731182456, local kl: 0.0 global kl: 1.3465418305713683e-05 valid reconstr loss: -0.26960110664367676
it: 3100, train recon loss: -0.7410905957221985, local kl: 0.0 global kl: 0.0010045666713267565 valid reconstr loss: 82.62285614013672
it: 3200, train recon loss: -0.4341035783290863, local kl: 0.0 global kl: 1.4231738532544114e-05 valid reconstr loss: -0.43039020895957947
Saving best model with reconstruction loss -0.4303902
it: 3300, train recon loss: 76.09778594970703, local kl: 0.0 global kl: 7.287962944246829e-05 valid reconstr loss: -0.3922320008277893
it: 3400, train recon loss: 108.95565795898438, local kl: 0.0 global kl: 3.161386848660186e-05 valid reconstr loss: -0.10896353423595428
it: 3500, train recon loss: -0.6745980978012085, local kl: 0.0 global kl: 8.99694896361325e-06 valid reconstr loss: 902.9730224609375
it: 3600, train recon loss: -0.7168179750442505, local kl: 0.0 global kl: 7.749066753603984e-06 valid reconstr loss: -0.1364513486623764
it: 3700, train recon loss: -0.9230367541313171, local kl: 0.0 global kl: 2.076228520309087e-05 valid reconstr loss: -0.5824180245399475
Saving best model with reconstruction loss -0.582418
it: 3800, train recon loss: -0.5753132104873657, local kl: 0.0 global kl: 0.0005695143481716514 valid reconstr loss: -0.7180395722389221
Saving best model with reconstruction loss -0.7180396
it: 3900, train recon loss: -1.025532841682434, local kl: 0.0 global kl: 2.5357663616887294e-05 valid reconstr loss: -0.549857497215271
it: 4000, train recon loss: -0.8261401653289795, local kl: 0.0 global kl: 1.605558054507128e-06 valid reconstr loss: -0.6534824967384338
it: 4100, train recon loss: -0.7165498733520508, local kl: 0.0 global kl: 6.966698856558651e-05 valid reconstr loss: 2113.331298828125
it: 4200, train recon loss: -0.6039716601371765, local kl: 0.0 global kl: 6.891521479701623e-05 valid reconstr loss: 5.810895919799805
it: 4300, train recon loss: -1.0944796800613403, local kl: 0.0 global kl: 6.40308144284063e-06 valid reconstr loss: -0.8254587054252625
Saving best model with reconstruction loss -0.8254587
it: 4400, train recon loss: -0.9531213045120239, local kl: 0.0 global kl: 1.49983443407109e-05 valid reconstr loss: 0.18043841421604156
it: 4500, train recon loss: -1.0864416360855103, local kl: 0.0 global kl: 9.816617421165574e-06 valid reconstr loss: 0.1005127876996994
it: 4600, train recon loss: -0.9498661160469055, local kl: 0.0 global kl: 0.00011525295121828094 valid reconstr loss: -0.8982389569282532
Saving best model with reconstruction loss -0.89823896
it: 4700, train recon loss: -0.8131788969039917, local kl: 0.0 global kl: 2.137947376468219e-05 valid reconstr loss: 4577.07861328125
it: 4800, train recon loss: -0.8346642851829529, local kl: 0.0 global kl: 0.0001494210446253419 valid reconstr loss: -1.0328036546707153
Saving best model with reconstruction loss -1.0328037
it: 4900, train recon loss: 2.805234432220459, local kl: 0.0 global kl: 5.610828520730138e-05 valid reconstr loss: -0.21965448558330536
it: 5000, train recon loss: -0.002814331091940403, local kl: 0.0 global kl: 3.142768764519133e-05 valid reconstr loss: -0.8101975321769714
it: 5100, train recon loss: -0.5384467244148254, local kl: 0.0 global kl: 2.176367888750974e-05 valid reconstr loss: -0.8892692923545837
it: 5200, train recon loss: -1.268417477607727, local kl: 0.0 global kl: 9.79696778813377e-06 valid reconstr loss: 28703.1484375
it: 5300, train recon loss: -0.8381857872009277, local kl: 0.0 global kl: 5.9469795814948156e-05 valid reconstr loss: -1.018293857574463
it: 5400, train recon loss: -1.072350263595581, local kl: 0.0 global kl: 0.000460443930933252 valid reconstr loss: 18726.212890625
it: 5500, train recon loss: -1.3683422803878784, local kl: 0.0 global kl: 9.277787466999143e-05 valid reconstr loss: -0.7645642757415771
it: 5600, train recon loss: -1.4634495973587036, local kl: 0.0 global kl: 3.162576467730105e-05 valid reconstr loss: -0.9387518167495728
it: 5700, train recon loss: 7.837559700012207, local kl: 0.0 global kl: 7.109389116521925e-05 valid reconstr loss: -0.48407256603240967
it: 5800, train recon loss: -1.3375544548034668, local kl: 0.0 global kl: 1.996088758460246e-05 valid reconstr loss: -0.45157483220100403
it: 5900, train recon loss: -1.160746455192566, local kl: 0.0 global kl: 2.348492489545606e-05 valid reconstr loss: -0.743255615234375
it: 6000, train recon loss: 4.656132221221924, local kl: 0.0 global kl: 1.8160224499297328e-05 valid reconstr loss: 3.863212823867798
it: 6100, train recon loss: 273.0610656738281, local kl: 0.0 global kl: 9.264059190172702e-05 valid reconstr loss: 1.1357706785202026
it: 6200, train recon loss: 1.1091194152832031, local kl: 0.0 global kl: 4.636889570974745e-05 valid reconstr loss: 65.02570343017578
it: 6300, train recon loss: 0.8335791826248169, local kl: 0.0 global kl: 4.2802756070159376e-05 valid reconstr loss: 1.3710979223251343
it: 6400, train recon loss: 45.357032775878906, local kl: 0.0 global kl: 0.00015133750275708735 valid reconstr loss: 0.2337842434644699
it: 6500, train recon loss: 0.286469429731369, local kl: 0.0 global kl: 2.1483274394995533e-06 valid reconstr loss: 0.18540510535240173
it: 6600, train recon loss: 0.09457942098379135, local kl: 0.0 global kl: 1.0568322977633215e-05 valid reconstr loss: -0.09252037107944489
it: 6700, train recon loss: 2.5419445037841797, local kl: 0.0 global kl: 0.0002394020848441869 valid reconstr loss: 0.1671801507472992
it: 6800, train recon loss: 0.8137074708938599, local kl: 0.0 global kl: 0.00034089150722138584 valid reconstr loss: 0.7607678174972534
it: 6900, train recon loss: 8.587017059326172, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.4231750965118408
it: 7000, train recon loss: -0.26181232929229736, local kl: 0.0 global kl: 1.9595665889937663e-06 valid reconstr loss: 0.39457571506500244
it: 7100, train recon loss: 2.134300708770752, local kl: 0.0 global kl: 3.8338998820108827e-07 valid reconstr loss: 0.12523987889289856
it: 7200, train recon loss: -0.30822551250457764, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.21127460896968842
it: 7300, train recon loss: 1.205967903137207, local kl: 0.0 global kl: 7.642066339030862e-05 valid reconstr loss: -0.4360300898551941
it: 7400, train recon loss: -0.11060402542352676, local kl: 0.0 global kl: 0.0 valid reconstr loss: 11786.8056640625
it: 7500, train recon loss: -0.014269876293838024, local kl: 0.0 global kl: 0.00032062470563687384 valid reconstr loss: 3373.01806640625
it: 7600, train recon loss: 0.15948247909545898, local kl: 0.0 global kl: 0.00035364003269933164 valid reconstr loss: 9.470982551574707
it: 7700, train recon loss: -0.4074900448322296, local kl: 0.0 global kl: 0.00027484347810968757 valid reconstr loss: 1319.0537109375
it: 7800, train recon loss: -0.49448662996292114, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.5006232857704163
it: 7900, train recon loss: 4.194014549255371, local kl: 0.0 global kl: 0.00013279126142151654 valid reconstr loss: -0.5567134022712708
it: 8000, train recon loss: -0.3493652939796448, local kl: 0.0 global kl: 0.00019566845730878413 valid reconstr loss: 0.4156753122806549
it: 8100, train recon loss: -0.34052759408950806, local kl: 0.0 global kl: 5.058282476966269e-05 valid reconstr loss: -0.5221006870269775
it: 8200, train recon loss: -0.3512011170387268, local kl: 0.0 global kl: 7.446455128956586e-05 valid reconstr loss: 1.5009621381759644
it: 8300, train recon loss: -0.9228959083557129, local kl: 0.0 global kl: 0.0003807430330198258 valid reconstr loss: -0.30419763922691345
it: 8400, train recon loss: -0.7446043491363525, local kl: 0.0 global kl: 2.034583485510666e-05 valid reconstr loss: -0.16748012602329254
it: 8500, train recon loss: -0.2621789574623108, local kl: 0.0 global kl: 0.0014645010232925415 valid reconstr loss: -0.6509400606155396
it: 8600, train recon loss: -0.28454530239105225, local kl: 0.0 global kl: 9.862102388069616e-07 valid reconstr loss: -0.8922404646873474
it: 8700, train recon loss: -0.9496239423751831, local kl: 0.0 global kl: 1.4515414477500599e-05 valid reconstr loss: -0.7121867537498474
it: 8800, train recon loss: 4168.78955078125, local kl: 0.0 global kl: 0.00044815466389991343 valid reconstr loss: 3.9630239009857178
it: 8900, train recon loss: 0.17778058350086212, local kl: 0.0 global kl: 0.0007144958944991231 valid reconstr loss: 998.451416015625
it: 9000, train recon loss: 47.000667572021484, local kl: 0.0 global kl: 0.0001522630627732724 valid reconstr loss: 25.121042251586914
it: 9100, train recon loss: 3.414522171020508, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.32506754994392395
it: 9200, train recon loss: 0.3114039897918701, local kl: 0.0 global kl: 3.627655360105564e-06 valid reconstr loss: 0.8392578959465027
it: 9300, train recon loss: -0.5754729509353638, local kl: 0.0 global kl: 2.080664489767514e-05 valid reconstr loss: -0.7360352873802185
it: 9400, train recon loss: -0.08107542991638184, local kl: 0.0 global kl: 0.0001906715042423457 valid reconstr loss: 7027.5458984375
it: 9500, train recon loss: -0.4078889489173889, local kl: 0.0 global kl: 3.7198292375251185e-06 valid reconstr loss: 685.1614379882812
it: 9600, train recon loss: -0.953900933265686, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.980514645576477
it: 9700, train recon loss: -0.9147209525108337, local kl: 0.0 global kl: 2.9745001484116074e-07 valid reconstr loss: -0.96791672706604
it: 9800, train recon loss: 0.7671995162963867, local kl: 0.0 global kl: 5.452298864838667e-05 valid reconstr loss: 542.7655029296875
it: 9900, train recon loss: 308.3265075683594, local kl: 0.0 global kl: 1.1389231246994314e-07 valid reconstr loss: 353.0931396484375
beta 5.0 temperature 0.5
it: 0, train recon loss: 229.8505096435547, local kl: 0.0 global kl: 0.058804549276828766 valid reconstr loss: 41.986995697021484
Saving best model with reconstruction loss 41.986996
it: 100, train recon loss: 3.624347448348999, local kl: 0.0 global kl: 0.0028764386661350727 valid reconstr loss: 3.905827283859253
Saving best model with reconstruction loss 3.9058273
it: 200, train recon loss: 3.096052885055542, local kl: 0.0 global kl: 0.00040316846570931375 valid reconstr loss: 3.0785632133483887
Saving best model with reconstruction loss 3.0785632
it: 300, train recon loss: 2.834062099456787, local kl: 0.0 global kl: 0.0006774527137167752 valid reconstr loss: 3.89892840385437
it: 400, train recon loss: 2.3694891929626465, local kl: 0.0 global kl: 0.00019797837012447417 valid reconstr loss: 2.850163221359253
Saving best model with reconstruction loss 2.8501632
it: 500, train recon loss: 1.1234729290008545, local kl: 0.0 global kl: 0.0003160459455102682 valid reconstr loss: 1.8936322927474976
Saving best model with reconstruction loss 1.8936323
it: 600, train recon loss: 1.331855297088623, local kl: 0.0 global kl: 7.268627814482898e-05 valid reconstr loss: 1.060638189315796
Saving best model with reconstruction loss 1.0606382
it: 700, train recon loss: 0.7849045395851135, local kl: 0.0 global kl: 0.0004016616730950773 valid reconstr loss: 0.9915487170219421
Saving best model with reconstruction loss 0.9915487
it: 800, train recon loss: 0.6158046722412109, local kl: 0.0 global kl: 3.6519020795822144e-05 valid reconstr loss: 1.611817479133606
it: 900, train recon loss: 1.0007636547088623, local kl: 0.0 global kl: 9.118585148826241e-05 valid reconstr loss: 234.0219268798828
it: 1000, train recon loss: 0.4363456964492798, local kl: 0.0 global kl: 0.00011345847451593727 valid reconstr loss: 0.23764708638191223
Saving best model with reconstruction loss 0.23764709
it: 1100, train recon loss: 0.05788729712367058, local kl: 0.0 global kl: 0.00024091680825222284 valid reconstr loss: 0.37587064504623413
it: 1200, train recon loss: 3.9019572734832764, local kl: 0.0 global kl: 0.00023230505757965147 valid reconstr loss: 3.986924886703491
it: 1300, train recon loss: 3.8645873069763184, local kl: 0.0 global kl: 0.015233218669891357 valid reconstr loss: 2.8960025310516357
it: 1400, train recon loss: 2.564592123031616, local kl: 0.0 global kl: 0.0004645240551326424 valid reconstr loss: 7.320764541625977
it: 1500, train recon loss: 1.5307377576828003, local kl: 0.0 global kl: 0.0001385780778946355 valid reconstr loss: 1.9520128965377808
it: 1600, train recon loss: 2.678788900375366, local kl: 0.0 global kl: 0.0006516436696983874 valid reconstr loss: 2.6409244537353516
it: 1700, train recon loss: 5.282701015472412, local kl: 0.0 global kl: 0.0006750720203854144 valid reconstr loss: 47.209312438964844
it: 1800, train recon loss: 6.187495231628418, local kl: 0.0 global kl: 0.0007522684172727168 valid reconstr loss: 1.1208091974258423
it: 1900, train recon loss: 1.6077111959457397, local kl: 0.0 global kl: 1.9781615264946595e-05 valid reconstr loss: 1.0840176343917847
it: 2000, train recon loss: 1.3954757452011108, local kl: 0.0 global kl: 0.0005004378035664558 valid reconstr loss: 0.771578311920166
it: 2100, train recon loss: 0.9682698845863342, local kl: 0.0 global kl: 2.7107282221550122e-05 valid reconstr loss: 1.151597023010254
it: 2200, train recon loss: 0.7175027132034302, local kl: 0.0 global kl: 1.9953538867412135e-05 valid reconstr loss: 1.0886776447296143
it: 2300, train recon loss: 2.1916799545288086, local kl: 0.0 global kl: 0.00018617774185258895 valid reconstr loss: 0.9917781949043274
it: 2400, train recon loss: 0.9002651572227478, local kl: 0.0 global kl: 4.948607966070995e-05 valid reconstr loss: 0.5399529933929443
it: 2500, train recon loss: 0.7430101037025452, local kl: 0.0 global kl: 9.104060154641047e-05 valid reconstr loss: 3.877915620803833
it: 2600, train recon loss: 0.6845508217811584, local kl: 0.0 global kl: 4.7364628699142486e-05 valid reconstr loss: 0.6034857630729675
it: 2700, train recon loss: 0.7745735049247742, local kl: 0.0 global kl: 0.0001293648238060996 valid reconstr loss: 1.2616618871688843
it: 2800, train recon loss: 0.5004393458366394, local kl: 0.0 global kl: 0.0004118080250918865 valid reconstr loss: 29.396085739135742
it: 2900, train recon loss: 0.5928455591201782, local kl: 0.0 global kl: 9.838245750870556e-05 valid reconstr loss: 1.2684144973754883
it: 3000, train recon loss: 8.889413833618164, local kl: 0.0 global kl: 0.00098237837664783 valid reconstr loss: 0.6491523385047913
it: 3100, train recon loss: 0.5417429208755493, local kl: 0.0 global kl: 0.0006743749836459756 valid reconstr loss: 0.7310086488723755
it: 3200, train recon loss: 0.7311554551124573, local kl: 0.0 global kl: 0.0004160665557719767 valid reconstr loss: 0.539290726184845
it: 3300, train recon loss: 0.5428767800331116, local kl: 0.0 global kl: 1.0409962669655215e-05 valid reconstr loss: 0.8685897588729858
it: 3400, train recon loss: 0.3718603849411011, local kl: 0.0 global kl: 2.37800177274039e-05 valid reconstr loss: 0.5294874310493469
it: 3500, train recon loss: 0.09883464127779007, local kl: 0.0 global kl: 8.670582610648125e-05 valid reconstr loss: 0.8536007404327393
it: 3600, train recon loss: 0.8738467693328857, local kl: 0.0 global kl: 0.00013976366608403623 valid reconstr loss: 0.4912422001361847
it: 3700, train recon loss: 0.3829348087310791, local kl: 0.0 global kl: 0.0002940995618700981 valid reconstr loss: 0.6040334105491638
it: 3800, train recon loss: 0.8442939519882202, local kl: 0.0 global kl: 0.00022197890211828053 valid reconstr loss: 0.428562730550766
it: 3900, train recon loss: -0.09865248948335648, local kl: 0.0 global kl: 9.584444342181087e-06 valid reconstr loss: 0.926394522190094
it: 4000, train recon loss: 0.28505340218544006, local kl: 0.0 global kl: 0.00012670858995988965 valid reconstr loss: 0.38904571533203125
it: 4100, train recon loss: 0.20887967944145203, local kl: 0.0 global kl: 4.4342839828459546e-05 valid reconstr loss: 0.40287652611732483
it: 4200, train recon loss: 0.6040640473365784, local kl: 0.0 global kl: 1.0306925979364223e-08 valid reconstr loss: 0.26198652386665344
it: 4300, train recon loss: 0.5931218266487122, local kl: 0.0 global kl: 2.091047281282954e-05 valid reconstr loss: 0.47566893696784973
it: 4400, train recon loss: 0.32854020595550537, local kl: 0.0 global kl: 3.469657531240955e-05 valid reconstr loss: 0.7595886588096619
it: 4500, train recon loss: 0.20160064101219177, local kl: 0.0 global kl: 7.206136160675669e-06 valid reconstr loss: 0.6046111583709717
it: 4600, train recon loss: 0.23226162791252136, local kl: 0.0 global kl: 7.246651421155548e-07 valid reconstr loss: 0.4913569390773773
it: 4700, train recon loss: 0.3920895755290985, local kl: 0.0 global kl: 1.3521163964469451e-05 valid reconstr loss: 0.438493549823761
it: 4800, train recon loss: 0.11959174275398254, local kl: 0.0 global kl: 3.333674612804316e-05 valid reconstr loss: 0.3931882381439209
it: 4900, train recon loss: 0.13249275088310242, local kl: 0.0 global kl: 0.0003742655389942229 valid reconstr loss: 0.3277691900730133
it: 5000, train recon loss: 1.2107493877410889, local kl: 0.0 global kl: 3.170695208609686e-07 valid reconstr loss: 1.520992636680603
it: 5100, train recon loss: 0.5665735602378845, local kl: 0.0 global kl: 0.00018574538989923894 valid reconstr loss: 0.8952128291130066
it: 5200, train recon loss: 0.3380688726902008, local kl: 0.0 global kl: 2.1487332560354844e-05 valid reconstr loss: 0.5051103830337524
it: 5300, train recon loss: 2.6110281944274902, local kl: 0.0 global kl: 0.00023437620257027447 valid reconstr loss: 1.7387580871582031
it: 5400, train recon loss: 0.1073828861117363, local kl: 0.0 global kl: 8.90815044840565e-06 valid reconstr loss: 0.3160332143306732
it: 5500, train recon loss: 0.3957802653312683, local kl: 0.0 global kl: 3.1118220249481965e-06 valid reconstr loss: 0.17359454929828644
Saving best model with reconstruction loss 0.17359455
it: 5600, train recon loss: 0.08759884536266327, local kl: 0.0 global kl: 4.258090484654531e-05 valid reconstr loss: 0.9994960427284241
it: 5700, train recon loss: 2.4063358306884766, local kl: 0.0 global kl: 4.3010850276914425e-06 valid reconstr loss: 2.9137368202209473
it: 5800, train recon loss: 3.21773099899292, local kl: 0.0 global kl: 4.870844804827357e-06 valid reconstr loss: 1.940725564956665
it: 5900, train recon loss: 1.2513331174850464, local kl: 0.0 global kl: 3.049822771572508e-05 valid reconstr loss: 1.1710915565490723
it: 6000, train recon loss: 0.8046160936355591, local kl: 0.0 global kl: 2.1463860321091488e-05 valid reconstr loss: 0.9203201532363892
it: 6100, train recon loss: 0.4327527582645416, local kl: 0.0 global kl: 7.059025847411249e-06 valid reconstr loss: 1.236099362373352
it: 6200, train recon loss: 0.7888680696487427, local kl: 0.0 global kl: 2.7126492341267294e-07 valid reconstr loss: 0.7920671105384827
it: 6300, train recon loss: 2.494999885559082, local kl: 0.0 global kl: 1.1648048712231684e-06 valid reconstr loss: 3.0393896102905273
it: 6400, train recon loss: 1.2461456060409546, local kl: 0.0 global kl: 1.6153055071299605e-07 valid reconstr loss: 11.008037567138672
it: 6500, train recon loss: 1.5583534240722656, local kl: 0.0 global kl: 1.3647621017298661e-05 valid reconstr loss: 1.4691388607025146
it: 6600, train recon loss: 1.1433830261230469, local kl: 0.0 global kl: 0.00016855160356499255 valid reconstr loss: 0.8555760383605957
it: 6700, train recon loss: 0.6960185766220093, local kl: 0.0 global kl: 2.8274068881728454e-06 valid reconstr loss: 4.852327823638916
it: 6800, train recon loss: 0.44009196758270264, local kl: 0.0 global kl: 1.7621472352402634e-06 valid reconstr loss: 0.7378169298171997
it: 6900, train recon loss: 0.589522123336792, local kl: 0.0 global kl: 1.8866794562200084e-05 valid reconstr loss: 0.7238398194313049
it: 7000, train recon loss: 2.662631034851074, local kl: 0.0 global kl: 3.0462208542303415e-06 valid reconstr loss: 0.9700077176094055
it: 7100, train recon loss: 0.2909298241138458, local kl: 0.0 global kl: 6.04717286023515e-07 valid reconstr loss: 0.26669609546661377
it: 7200, train recon loss: 558.0845947265625, local kl: 0.0 global kl: 0.000839675310999155 valid reconstr loss: 9.101435661315918
it: 7300, train recon loss: 0.5365064740180969, local kl: 0.0 global kl: 6.510200910270214e-05 valid reconstr loss: 0.4310680329799652
it: 7400, train recon loss: 0.6078348755836487, local kl: 0.0 global kl: 2.4797827791189775e-08 valid reconstr loss: 0.7040795087814331
it: 7500, train recon loss: 1.4396907091140747, local kl: 0.0 global kl: 7.628856837982312e-05 valid reconstr loss: 1.278325080871582
it: 7600, train recon loss: 1.1538900136947632, local kl: 0.0 global kl: 1.3572197531175334e-05 valid reconstr loss: 2.1121742725372314
it: 7700, train recon loss: 1.6665500402450562, local kl: 0.0 global kl: 0.00028076162561774254 valid reconstr loss: 2.803731918334961
it: 7800, train recon loss: 0.9465357661247253, local kl: 0.0 global kl: 7.09713390278921e-08 valid reconstr loss: 0.8733001351356506
it: 7900, train recon loss: 0.5817191004753113, local kl: 0.0 global kl: 0.0001731246302369982 valid reconstr loss: 0.9751858711242676
it: 8000, train recon loss: 0.19452518224716187, local kl: 0.0 global kl: 1.1651120530586923e-06 valid reconstr loss: 0.527662456035614
it: 8100, train recon loss: 1.4713565111160278, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.7578303813934326
it: 8200, train recon loss: -0.04914563149213791, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.6270885467529297
it: 8300, train recon loss: 1.527498722076416, local kl: 0.0 global kl: 1.0316682619304629e-07 valid reconstr loss: 1.2595961093902588
it: 8400, train recon loss: 1.070241093635559, local kl: 0.0 global kl: 1.460380531170813e-06 valid reconstr loss: 1.008752703666687
it: 8500, train recon loss: 1.4827691316604614, local kl: 0.0 global kl: 5.093267100164667e-05 valid reconstr loss: 0.886582612991333
it: 8600, train recon loss: 0.9585766792297363, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.8218339085578918
it: 8700, train recon loss: 0.9495123028755188, local kl: 0.0 global kl: 2.0094952560612e-05 valid reconstr loss: 0.5939041972160339
it: 8800, train recon loss: 1.3139045238494873, local kl: 0.0 global kl: 4.510275539360009e-05 valid reconstr loss: 0.8663922548294067
it: 8900, train recon loss: 0.4475018084049225, local kl: 0.0 global kl: 1.4120547575657838e-07 valid reconstr loss: 0.8491684794425964
it: 9000, train recon loss: 0.4762764275074005, local kl: 0.0 global kl: 8.362803782802075e-06 valid reconstr loss: 0.7241015434265137
it: 9100, train recon loss: 14.445029258728027, local kl: 0.0 global kl: 8.799477768661745e-08 valid reconstr loss: 0.5741092562675476
it: 9200, train recon loss: 0.7795984745025635, local kl: 0.0 global kl: 1.0520290743443184e-05 valid reconstr loss: 0.5643964409828186
it: 9300, train recon loss: 0.2053813487291336, local kl: 0.0 global kl: 2.2667229586659232e-06 valid reconstr loss: 0.26106971502304077
it: 9400, train recon loss: 0.5785466432571411, local kl: 0.0 global kl: 5.838787728862371e-06 valid reconstr loss: 0.32894667983055115
it: 9500, train recon loss: 0.4793620705604553, local kl: 0.0 global kl: 1.792437842595973e-07 valid reconstr loss: 0.44039666652679443
it: 9600, train recon loss: 0.34658271074295044, local kl: 0.0 global kl: 1.003687799538966e-09 valid reconstr loss: 0.48743608593940735
it: 9700, train recon loss: 0.434529185295105, local kl: 0.0 global kl: 2.3565096853417344e-05 valid reconstr loss: 0.27477380633354187
it: 9800, train recon loss: 0.6360170245170593, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.25128257274627686
it: 9900, train recon loss: 544533.0, local kl: 0.0 global kl: 1.2217572475492489e-05 valid reconstr loss: 0.35291972756385803
beta 5.0 temperature 1.0
it: 0, train recon loss: 302.55828857421875, local kl: 0.0 global kl: 0.044660378247499466 valid reconstr loss: 246.5460968017578
Saving best model with reconstruction loss 246.5461
it: 100, train recon loss: 3.7560079097747803, local kl: 0.0 global kl: 0.0005917690577916801 valid reconstr loss: 4.170399188995361
Saving best model with reconstruction loss 4.170399
it: 200, train recon loss: 301.9827575683594, local kl: 0.0 global kl: 3.434220707276836e-05 valid reconstr loss: 4.346340656280518
it: 300, train recon loss: 6.824217319488525, local kl: 0.0 global kl: 0.00010136594937648624 valid reconstr loss: 3.3724021911621094
Saving best model with reconstruction loss 3.3724022
it: 400, train recon loss: 2.66550612449646, local kl: 0.0 global kl: 0.0001493888266850263 valid reconstr loss: 4.201319217681885
it: 500, train recon loss: 2.303938388824463, local kl: 0.0 global kl: 0.00023941505060065538 valid reconstr loss: 2.197618246078491
Saving best model with reconstruction loss 2.1976182
it: 600, train recon loss: 2.0125184059143066, local kl: 0.0 global kl: 2.9605780582642183e-05 valid reconstr loss: 2.1854727268218994
Saving best model with reconstruction loss 2.1854727
it: 700, train recon loss: 1.5521036386489868, local kl: 0.0 global kl: 0.00012684155080933124 valid reconstr loss: 38.30723571777344
it: 800, train recon loss: 2.0060482025146484, local kl: 0.0 global kl: 0.0001187130474136211 valid reconstr loss: 230.70828247070312
it: 900, train recon loss: 2.6991591453552246, local kl: 0.0 global kl: 0.0009606455569155514 valid reconstr loss: 1.3635776042938232
Saving best model with reconstruction loss 1.3635776
it: 1000, train recon loss: 1.2183486223220825, local kl: 0.0 global kl: 0.00010036743333330378 valid reconstr loss: 1.205951452255249
Saving best model with reconstruction loss 1.2059515
it: 1100, train recon loss: 2.5376136302948, local kl: 0.0 global kl: 6.665530963800848e-05 valid reconstr loss: 1.2706915140151978
it: 1200, train recon loss: 6.210448265075684, local kl: 0.0 global kl: 0.0001459907798562199 valid reconstr loss: 5.612042427062988
it: 1300, train recon loss: 0.5599180459976196, local kl: 0.0 global kl: 0.002228020690381527 valid reconstr loss: 4.407437801361084
it: 1400, train recon loss: 1.144668459892273, local kl: 0.0 global kl: 0.0003001828445121646 valid reconstr loss: 1.122666358947754
Saving best model with reconstruction loss 1.1226664
it: 1500, train recon loss: 0.5662272572517395, local kl: 0.0 global kl: 0.0004911859286949039 valid reconstr loss: 0.9297377467155457
Saving best model with reconstruction loss 0.92973775
it: 1600, train recon loss: 0.9812794327735901, local kl: 0.0 global kl: 1.3644899809150957e-05 valid reconstr loss: 1.194677472114563
it: 1700, train recon loss: 94.34839630126953, local kl: 0.0 global kl: 0.0032494135666638613 valid reconstr loss: 5.7740478515625
it: 1800, train recon loss: 0.9069685339927673, local kl: 0.0 global kl: 0.00013948911509942263 valid reconstr loss: 0.9681081175804138
it: 1900, train recon loss: 0.5194085240364075, local kl: 0.0 global kl: 5.77773098484613e-05 valid reconstr loss: 0.8518491387367249
Saving best model with reconstruction loss 0.85184914
it: 2000, train recon loss: 1.0989700555801392, local kl: 0.0 global kl: 0.00015821013948880136 valid reconstr loss: 0.612867534160614
Saving best model with reconstruction loss 0.61286753
it: 2100, train recon loss: 0.6407143473625183, local kl: 0.0 global kl: 3.0088593121035956e-05 valid reconstr loss: 1.0632123947143555
it: 2200, train recon loss: 0.6238855719566345, local kl: 0.0 global kl: 9.875565592665225e-05 valid reconstr loss: 0.6128876805305481
it: 2300, train recon loss: 1.2878365516662598, local kl: 0.0 global kl: 5.219254671828821e-05 valid reconstr loss: 1.9417253732681274
it: 2400, train recon loss: 0.8155848383903503, local kl: 0.0 global kl: 4.283137968741357e-05 valid reconstr loss: 1.0583679676055908
it: 2500, train recon loss: 0.6444426774978638, local kl: 0.0 global kl: 4.1802727537287865e-06 valid reconstr loss: 0.8203815817832947
it: 2600, train recon loss: 0.6285789012908936, local kl: 0.0 global kl: 5.471784243127331e-05 valid reconstr loss: 0.8474738597869873
it: 2700, train recon loss: 0.850427508354187, local kl: 0.0 global kl: 0.00017962725542020053 valid reconstr loss: 0.6406234502792358
it: 2800, train recon loss: 0.7673081755638123, local kl: 0.0 global kl: 0.0001786204520612955 valid reconstr loss: 0.5572347044944763
Saving best model with reconstruction loss 0.5572347
it: 2900, train recon loss: 0.42902740836143494, local kl: 0.0 global kl: 3.786136949202046e-05 valid reconstr loss: 0.723507821559906
it: 3000, train recon loss: 0.5469946265220642, local kl: 0.0 global kl: 3.60139092663303e-05 valid reconstr loss: 0.4440809488296509
Saving best model with reconstruction loss 0.44408095
it: 3100, train recon loss: 1.4614851474761963, local kl: 0.0 global kl: 2.061602117464645e-06 valid reconstr loss: 1.7109782695770264
it: 3200, train recon loss: 4.755972862243652, local kl: 0.0 global kl: 0.0004371994291432202 valid reconstr loss: 3.484463930130005
it: 3300, train recon loss: 3.902076244354248, local kl: 0.0 global kl: 0.0001681887952145189 valid reconstr loss: 3.727522134780884
it: 3400, train recon loss: 3.33905029296875, local kl: 0.0 global kl: 7.303986058104783e-05 valid reconstr loss: 3.4080302715301514
it: 3500, train recon loss: 3.541747570037842, local kl: 0.0 global kl: 5.636788955598604e-07 valid reconstr loss: 3.496896743774414
it: 3600, train recon loss: 3.337648868560791, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.964118480682373
it: 3700, train recon loss: 3.139516592025757, local kl: 0.0 global kl: 0.00013093921006657183 valid reconstr loss: 3.2602343559265137
it: 3800, train recon loss: 3.3478872776031494, local kl: 0.0 global kl: 0.00010786451457533985 valid reconstr loss: 3.2451529502868652
it: 3900, train recon loss: 2.8829615116119385, local kl: 0.0 global kl: 3.3907108445419e-05 valid reconstr loss: 3.209808349609375
it: 4000, train recon loss: 2.82613468170166, local kl: 0.0 global kl: 8.882297152013052e-06 valid reconstr loss: 3.214651346206665
it: 4100, train recon loss: 3.0700345039367676, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.1488847732543945
it: 4200, train recon loss: 3.193753480911255, local kl: 0.0 global kl: 1.7429909348720685e-05 valid reconstr loss: 3.205897808074951
it: 4300, train recon loss: 2.769359827041626, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.1465675830841064
it: 4400, train recon loss: 3.4412002563476562, local kl: 0.0 global kl: 5.186449925531633e-05 valid reconstr loss: 3.077585458755493
it: 4500, train recon loss: 3.011535167694092, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.125447988510132
it: 4600, train recon loss: 3.3425567150115967, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.8549745082855225
it: 4700, train recon loss: 3.1614153385162354, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.056819438934326
it: 4800, train recon loss: 2.596025228500366, local kl: 0.0 global kl: 0.0001548580767121166 valid reconstr loss: 12.159916877746582
it: 4900, train recon loss: 2.9074888229370117, local kl: 0.0 global kl: 0.0002458858070895076 valid reconstr loss: 4.269291400909424
it: 5000, train recon loss: 2.719703197479248, local kl: 0.0 global kl: 0.00189984031021595 valid reconstr loss: 152.60923767089844
it: 5100, train recon loss: 2.833313465118408, local kl: 0.0 global kl: 0.00010297136759618297 valid reconstr loss: 4.575492858886719
it: 5200, train recon loss: 3.316124677658081, local kl: 0.0 global kl: 0.0 valid reconstr loss: 23268090.0
it: 5300, train recon loss: 2.9483377933502197, local kl: 0.0 global kl: 0.0008522424614056945 valid reconstr loss: 3.4574222564697266
it: 5400, train recon loss: 3.161417007446289, local kl: 0.0 global kl: 2.977881558763329e-05 valid reconstr loss: 3.7151741981506348
it: 5500, train recon loss: 2.7015504837036133, local kl: 0.0 global kl: 8.582742339058314e-06 valid reconstr loss: 4.8378753662109375
it: 5600, train recon loss: 3.2059826850891113, local kl: 0.0 global kl: 7.830624235793948e-05 valid reconstr loss: 887.153564453125
it: 5700, train recon loss: 2.917353630065918, local kl: 0.0 global kl: 0.0005909927422180772 valid reconstr loss: 3.3567309379577637
it: 5800, train recon loss: 684.8995971679688, local kl: 0.0 global kl: 3.512196053634398e-05 valid reconstr loss: 2.8885273933410645
it: 5900, train recon loss: 3.159959077835083, local kl: 0.0 global kl: 6.045247857855429e-09 valid reconstr loss: 3.1405255794525146
it: 6000, train recon loss: 4.719505310058594, local kl: 0.0 global kl: 0.00041915540350601077 valid reconstr loss: 7.146515369415283
it: 6100, train recon loss: 20.159469604492188, local kl: 0.0 global kl: 0.0 valid reconstr loss: 23.608610153198242
it: 6200, train recon loss: 13.732211112976074, local kl: 0.0 global kl: 0.0 valid reconstr loss: 14.312729835510254
it: 6300, train recon loss: 9.887533187866211, local kl: 0.0 global kl: 0.0 valid reconstr loss: 9.712549209594727
it: 6400, train recon loss: 6.199582099914551, local kl: 0.0 global kl: 0.0 valid reconstr loss: 7.19179105758667
it: 6500, train recon loss: 5.150568008422852, local kl: 0.0 global kl: 0.0 valid reconstr loss: 5.736087322235107
it: 6600, train recon loss: 4.59501314163208, local kl: 0.0 global kl: 0.0 valid reconstr loss: 4.8213210105896
it: 6700, train recon loss: 3.5808265209198, local kl: 0.0 global kl: 0.0 valid reconstr loss: 4.229001045227051
it: 6800, train recon loss: 3.2567007541656494, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.7982068061828613
it: 6900, train recon loss: 3.708090305328369, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.5433223247528076
it: 7000, train recon loss: 3.298466682434082, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.3363559246063232
it: 7100, train recon loss: 2.966489791870117, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.2277376651763916
it: 7200, train recon loss: 3.043142080307007, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.1095731258392334
it: 7300, train recon loss: 3.2971489429473877, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.0724520683288574
it: 7400, train recon loss: 3.2615749835968018, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.001422882080078
it: 7500, train recon loss: 2.9988090991973877, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.964792013168335
it: 7600, train recon loss: 3.0229008197784424, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.9246416091918945
it: 7700, train recon loss: 2.770214796066284, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.9313745498657227
it: 7800, train recon loss: 2.855484962463379, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.91884183883667
it: 7900, train recon loss: 2.710216522216797, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.9165127277374268
it: 8000, train recon loss: 2.948009490966797, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.912348985671997
it: 8100, train recon loss: 2.84905743598938, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.88961124420166
it: 8200, train recon loss: 2.8490710258483887, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.9379384517669678
it: 8300, train recon loss: 2.9410669803619385, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.8899106979370117
it: 8400, train recon loss: 2.7245495319366455, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.9131393432617188
it: 8500, train recon loss: 3.0008695125579834, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.941415548324585
it: 8600, train recon loss: 3.0972976684570312, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.889005422592163
it: 8700, train recon loss: 2.8447539806365967, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.9129889011383057
it: 8800, train recon loss: 2.632880926132202, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.89351224899292
it: 8900, train recon loss: 2.7186694145202637, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.9626216888427734
it: 9000, train recon loss: 2.644913911819458, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.888032913208008
it: 9100, train recon loss: 3.0707263946533203, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.9294116497039795
it: 9200, train recon loss: 2.8622970581054688, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.905103921890259
it: 9300, train recon loss: 2.8742387294769287, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.8950414657592773
it: 9400, train recon loss: 2.984283685684204, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.894468069076538
it: 9500, train recon loss: 2.9801502227783203, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.944749593734741
it: 9600, train recon loss: 2.871767044067383, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.9044814109802246
it: 9700, train recon loss: 2.928199529647827, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.9046881198883057
it: 9800, train recon loss: 2.9899144172668457, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.9140102863311768
it: 9900, train recon loss: 2.667433500289917, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.924683094024658
beta 5.0 temperature 2.0
it: 0, train recon loss: 194.62774658203125, local kl: 0.0 global kl: 0.03097410872578621 valid reconstr loss: 92.25021362304688
Saving best model with reconstruction loss 92.25021
it: 100, train recon loss: 4.002897262573242, local kl: 0.0 global kl: 0.002357450546696782 valid reconstr loss: 3.984733819961548
Saving best model with reconstruction loss 3.9847338
it: 200, train recon loss: 4.055633068084717, local kl: 0.0 global kl: 3.037194619537331e-05 valid reconstr loss: 4.02206563949585
it: 300, train recon loss: 3.1854588985443115, local kl: 0.0 global kl: 7.120363443391398e-05 valid reconstr loss: 3.3641154766082764
Saving best model with reconstruction loss 3.3641155
it: 400, train recon loss: 2.7541615962982178, local kl: 0.0 global kl: 0.0001396296574966982 valid reconstr loss: 2.7358479499816895
Saving best model with reconstruction loss 2.735848
it: 500, train recon loss: 1.8929215669631958, local kl: 0.0 global kl: 6.187424787640339e-06 valid reconstr loss: 1.8912196159362793
Saving best model with reconstruction loss 1.8912196
it: 600, train recon loss: 1.5386518239974976, local kl: 0.0 global kl: 1.3428713145913207e-06 valid reconstr loss: 2.8663687705993652
it: 700, train recon loss: 0.9658165574073792, local kl: 0.0 global kl: 1.0574874309554616e-12 valid reconstr loss: 1.5607843399047852
Saving best model with reconstruction loss 1.5607843
it: 800, train recon loss: 0.7103286385536194, local kl: 0.0 global kl: 2.3945045599105264e-11 valid reconstr loss: 0.8870070576667786
Saving best model with reconstruction loss 0.88700706
it: 900, train recon loss: 0.9341216087341309, local kl: 0.0 global kl: 5.351274978693255e-13 valid reconstr loss: 0.7683536410331726
Saving best model with reconstruction loss 0.76835364
it: 1000, train recon loss: 0.4514123499393463, local kl: 0.0 global kl: 5.543818876185824e-11 valid reconstr loss: 0.46247580647468567
Saving best model with reconstruction loss 0.4624758
it: 1100, train recon loss: 1.845913290977478, local kl: 0.0 global kl: 4.5478204535598366e-11 valid reconstr loss: 0.5233162045478821
it: 1200, train recon loss: 2.728841543197632, local kl: 0.0 global kl: 1.3432463474849499e-10 valid reconstr loss: 0.6066211462020874
it: 1300, train recon loss: 0.1892436146736145, local kl: 0.0 global kl: 2.19824158875781e-12 valid reconstr loss: 0.1747417151927948
Saving best model with reconstruction loss 0.17474172
it: 1400, train recon loss: -0.15323922038078308, local kl: 0.0 global kl: 9.845492476845408e-12 valid reconstr loss: 1.094139575958252
it: 1500, train recon loss: 0.08760262280702591, local kl: 0.0 global kl: 3.2862601528904634e-13 valid reconstr loss: -0.05022607743740082
Saving best model with reconstruction loss -0.050226077
it: 1600, train recon loss: -0.015172534622251987, local kl: 0.0 global kl: 2.0872192862952943e-13 valid reconstr loss: -0.1827169507741928
Saving best model with reconstruction loss -0.18271695
it: 1700, train recon loss: -0.07441520690917969, local kl: 0.0 global kl: 1.6860748908165135e-11 valid reconstr loss: 0.4316646158695221
it: 1800, train recon loss: 0.3682153820991516, local kl: 0.0 global kl: 5.027089855502709e-12 valid reconstr loss: 0.869158148765564
it: 1900, train recon loss: 0.3922898769378662, local kl: 0.0 global kl: 4.760102902123808e-13 valid reconstr loss: -0.051633693277835846
it: 2000, train recon loss: 0.24446944892406464, local kl: 0.0 global kl: 3.608335852334221e-11 valid reconstr loss: 0.5693119764328003
it: 2100, train recon loss: -0.15823829174041748, local kl: 0.0 global kl: 1.2390088954816747e-12 valid reconstr loss: -0.13922777771949768
it: 2200, train recon loss: -0.4184701442718506, local kl: 0.0 global kl: 1.1052270210143433e-12 valid reconstr loss: -0.3842126429080963
Saving best model with reconstruction loss -0.38421264
it: 2300, train recon loss: -0.13028699159622192, local kl: 0.0 global kl: 6.866729407306593e-13 valid reconstr loss: -0.25470128655433655
it: 2400, train recon loss: 2018.3321533203125, local kl: 0.0 global kl: 5.347278175804604e-11 valid reconstr loss: 265946.875
it: 2500, train recon loss: -0.22404663264751434, local kl: 0.0 global kl: 8.956572462859302e-13 valid reconstr loss: -0.2620652914047241
it: 2600, train recon loss: 9.305304527282715, local kl: 0.0 global kl: 6.208977776367419e-11 valid reconstr loss: -0.3284892141819
it: 2700, train recon loss: -0.2817184031009674, local kl: 0.0 global kl: 1.9239543291860173e-10 valid reconstr loss: 0.3838076889514923
it: 2800, train recon loss: -0.511530876159668, local kl: 0.0 global kl: 3.5309255519422322e-12 valid reconstr loss: 230.9773406982422
it: 2900, train recon loss: 0.009443012066185474, local kl: 0.0 global kl: 1.7985612998927536e-13 valid reconstr loss: -0.4285193681716919
Saving best model with reconstruction loss -0.42851937
it: 3000, train recon loss: -0.6208435297012329, local kl: 0.0 global kl: 1.110268127435532e-11 valid reconstr loss: -0.6364144682884216
Saving best model with reconstruction loss -0.63641447
it: 3100, train recon loss: -1.0167255401611328, local kl: 0.0 global kl: 2.2417137590657887e-11 valid reconstr loss: 0.00809488631784916
it: 3200, train recon loss: 0.09367246925830841, local kl: 0.0 global kl: 2.717999436629981e-12 valid reconstr loss: -0.5533945560455322
it: 3300, train recon loss: -1.0367918014526367, local kl: 0.0 global kl: 1.0787204463014177e-12 valid reconstr loss: 3398.04296875
it: 3400, train recon loss: 8.347277641296387, local kl: 0.0 global kl: 5.743905351351941e-11 valid reconstr loss: -0.32639259099960327
it: 3500, train recon loss: -0.8059602379798889, local kl: 0.0 global kl: 3.06810132855162e-12 valid reconstr loss: -0.6841045022010803
Saving best model with reconstruction loss -0.6841045
it: 3600, train recon loss: -0.09121432155370712, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.7658824324607849
Saving best model with reconstruction loss -0.76588243
it: 3700, train recon loss: -1.1838957071304321, local kl: 0.0 global kl: 8.511746862893688e-11 valid reconstr loss: -0.6192429065704346
it: 3800, train recon loss: -0.2957382798194885, local kl: 0.0 global kl: 1.6237011735142914e-13 valid reconstr loss: 2878.990966796875
it: 3900, train recon loss: -1.1102628707885742, local kl: 0.0 global kl: 2.1552204465535851e-13 valid reconstr loss: 1.3363566398620605
it: 4000, train recon loss: -0.8296719789505005, local kl: 0.0 global kl: 3.885780586188048e-13 valid reconstr loss: 462.29180908203125
it: 4100, train recon loss: -0.8319010138511658, local kl: 0.0 global kl: 3.019806626980426e-13 valid reconstr loss: -0.8472793102264404
Saving best model with reconstruction loss -0.8472793
it: 4200, train recon loss: 0.32784539461135864, local kl: 0.0 global kl: 3.818713401049223e-11 valid reconstr loss: -0.9455938339233398
Saving best model with reconstruction loss -0.94559383
it: 4300, train recon loss: -1.1044085025787354, local kl: 0.0 global kl: 0.0 valid reconstr loss: 4.292994022369385
it: 4400, train recon loss: -0.7271276116371155, local kl: 0.0 global kl: 2.1434590213864624e-12 valid reconstr loss: -0.5449877381324768
it: 4500, train recon loss: -0.895598828792572, local kl: 0.0 global kl: 2.731148640577885e-13 valid reconstr loss: 17.349138259887695
it: 4600, train recon loss: -1.210098385810852, local kl: 0.0 global kl: 4.79109044104753e-11 valid reconstr loss: 3531.934326171875
it: 4700, train recon loss: -1.026065468788147, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.9896373748779297
Saving best model with reconstruction loss -0.9896374
it: 4800, train recon loss: -1.3249064683914185, local kl: 0.0 global kl: 4.8232463162323214e-11 valid reconstr loss: 1.3072247505187988
it: 4900, train recon loss: -0.7916236519813538, local kl: 0.0 global kl: 1.072552723718756e-11 valid reconstr loss: 0.34842023253440857
it: 5000, train recon loss: 484.383056640625, local kl: 0.0 global kl: 0.0 valid reconstr loss: 414.64337158203125
it: 5100, train recon loss: -0.46823641657829285, local kl: 0.0 global kl: 1.0391687510491465e-12 valid reconstr loss: 2.875840663909912
it: 5200, train recon loss: -1.3014304637908936, local kl: 0.0 global kl: 0.0 valid reconstr loss: 6797.12744140625
it: 5300, train recon loss: -0.9156458973884583, local kl: 0.0 global kl: 1.9382966065517593e-10 valid reconstr loss: -0.9128713607788086
it: 5400, train recon loss: -1.2498290538787842, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.13478927314281464
it: 5500, train recon loss: 267.99566650390625, local kl: 0.0 global kl: 1.4751030258386422e-11 valid reconstr loss: -1.0755642652511597
Saving best model with reconstruction loss -1.0755643
it: 5600, train recon loss: 8.952546119689941, local kl: 0.0 global kl: 1.466245354297957e-11 valid reconstr loss: 11.317370414733887
it: 5700, train recon loss: 20.972980499267578, local kl: 0.0 global kl: 2.7629012966379207e-10 valid reconstr loss: 2.4846274852752686
it: 5800, train recon loss: -1.1499366760253906, local kl: 0.0 global kl: 3.4421909766990666e-11 valid reconstr loss: 0.6980741024017334
it: 5900, train recon loss: -1.2072296142578125, local kl: 0.0 global kl: 1.1560891133299833e-11 valid reconstr loss: -0.985122799873352
it: 6000, train recon loss: -1.3236662149429321, local kl: 0.0 global kl: 1.8497769288527977e-11 valid reconstr loss: 5708.95068359375
it: 6100, train recon loss: 6407.9013671875, local kl: 0.0 global kl: 1.0396197791529005e-12 valid reconstr loss: 39.61510467529297
it: 6200, train recon loss: -0.8620927333831787, local kl: 0.0 global kl: 2.3778539900987106e-12 valid reconstr loss: 123.2720947265625
it: 6300, train recon loss: -0.03240777552127838, local kl: 0.0 global kl: 1.3063335135843346e-13 valid reconstr loss: -1.1189160346984863
Saving best model with reconstruction loss -1.118916
it: 6400, train recon loss: 881.9768676757812, local kl: 0.0 global kl: 2.474260206442036e-11 valid reconstr loss: 60.98886489868164
it: 6500, train recon loss: -0.9745237827301025, local kl: 0.0 global kl: 4.533960012986782e-13 valid reconstr loss: 24.098142623901367
it: 6600, train recon loss: -0.778452455997467, local kl: 0.0 global kl: 1.5109658316192487e-13 valid reconstr loss: -1.2973017692565918
Saving best model with reconstruction loss -1.2973018
it: 6700, train recon loss: 2.988936424255371, local kl: 0.0 global kl: 6.87297441182011e-14 valid reconstr loss: 11.80400276184082
it: 6800, train recon loss: -1.014274001121521, local kl: 0.0 global kl: 7.764795750819786e-12 valid reconstr loss: -0.9818738698959351
it: 6900, train recon loss: 2.7538022994995117, local kl: 0.0 global kl: 4.801249328667545e-11 valid reconstr loss: -0.773591160774231
it: 7000, train recon loss: -0.9547179341316223, local kl: 0.0 global kl: 1.8493678810571623e-11 valid reconstr loss: -0.9434695243835449
it: 7100, train recon loss: -1.0984829664230347, local kl: 0.0 global kl: 3.736247422558847e-13 valid reconstr loss: 21.790817260742188
it: 7200, train recon loss: -1.0811128616333008, local kl: 0.0 global kl: 2.540687452090573e-11 valid reconstr loss: 1886.49853515625
it: 7300, train recon loss: 7.870398044586182, local kl: 0.0 global kl: 0.0 valid reconstr loss: 6477.42529296875
it: 7400, train recon loss: -0.9426507353782654, local kl: 0.0 global kl: 8.009218288584918e-13 valid reconstr loss: 110.17344665527344
it: 7500, train recon loss: -0.632754385471344, local kl: 0.0 global kl: 1.1203885041943806e-12 valid reconstr loss: 10436.654296875
it: 7600, train recon loss: -0.6682654619216919, local kl: 0.0 global kl: 2.587333125525504e-11 valid reconstr loss: 18.515077590942383
it: 7700, train recon loss: -1.2898993492126465, local kl: 0.0 global kl: 5.056400853575838e-10 valid reconstr loss: 12.38630485534668
it: 7800, train recon loss: 3.4170644283294678, local kl: 0.0 global kl: 1.8286788749932725e-10 valid reconstr loss: 2.9784188270568848
it: 7900, train recon loss: 3.0190675258636475, local kl: 0.0 global kl: 3.684574347018188e-13 valid reconstr loss: 684.3717651367188
it: 8000, train recon loss: -0.5221607089042664, local kl: 0.0 global kl: 2.2852414405249988e-11 valid reconstr loss: 103.31768798828125
it: 8100, train recon loss: 0.7232498526573181, local kl: 0.0 global kl: 5.703944261359339e-13 valid reconstr loss: 220.25485229492188
it: 8200, train recon loss: -1.288954496383667, local kl: 0.0 global kl: 2.467817616924606e-13 valid reconstr loss: 25.667402267456055
it: 8300, train recon loss: 2209.5224609375, local kl: 0.0 global kl: 4.784717760886181e-11 valid reconstr loss: 0.7621733546257019
it: 8400, train recon loss: 3607.61572265625, local kl: 0.0 global kl: 6.61866395024191e-13 valid reconstr loss: -0.8152633905410767
it: 8500, train recon loss: -1.2204761505126953, local kl: 0.0 global kl: 0.0 valid reconstr loss: 614.762451171875
it: 8600, train recon loss: -1.0587409734725952, local kl: 0.0 global kl: 7.688058523136476e-11 valid reconstr loss: -1.0152662992477417
it: 8700, train recon loss: 8.680449485778809, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.8278125524520874
it: 8800, train recon loss: -1.0037003755569458, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.0944037437438965
it: 8900, train recon loss: -1.2021671533584595, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.1627236604690552
it: 9000, train recon loss: 4138.59228515625, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.5335432887077332
it: 9100, train recon loss: 123.57260131835938, local kl: 0.0 global kl: 0.0 valid reconstr loss: 6677.36669921875
it: 9200, train recon loss: -1.127907633781433, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3218.219482421875
it: 9300, train recon loss: -0.606262743473053, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.366638422012329
it: 9400, train recon loss: -1.0396993160247803, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.7972379326820374
it: 9500, train recon loss: 1.1695140600204468, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.9876459836959839
it: 9600, train recon loss: 37.46001052856445, local kl: 0.0 global kl: 2.9090618802740664e-12 valid reconstr loss: 2104.513427734375
it: 9700, train recon loss: -1.323438286781311, local kl: 0.0 global kl: 9.83935155574045e-14 valid reconstr loss: 1071.0211181640625
it: 9800, train recon loss: -1.0990879535675049, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.0715227127075195
it: 9900, train recon loss: -1.3092702627182007, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.8402510285377502
beta 5.0 temperature 5.0
it: 0, train recon loss: 808.0764770507812, local kl: 0.0 global kl: 0.08431382477283478 valid reconstr loss: 237.4058837890625
Saving best model with reconstruction loss 237.40588
it: 100, train recon loss: 3.810129165649414, local kl: 0.0 global kl: 0.000948996574152261 valid reconstr loss: 4.130391597747803
Saving best model with reconstruction loss 4.1303916
it: 200, train recon loss: 3.9371793270111084, local kl: 0.0 global kl: 1.7813994190873927e-06 valid reconstr loss: 4.100443363189697
Saving best model with reconstruction loss 4.1004434
it: 300, train recon loss: 3.2100160121917725, local kl: 0.0 global kl: 0.00015875355165917426 valid reconstr loss: 3.479682683944702
Saving best model with reconstruction loss 3.4796827
it: 400, train recon loss: 2.610872745513916, local kl: 0.0 global kl: 0.00023579390835948288 valid reconstr loss: 2.7865371704101562
Saving best model with reconstruction loss 2.7865372
it: 500, train recon loss: 2.363839626312256, local kl: 0.0 global kl: 0.00041041840449906886 valid reconstr loss: 2.3095362186431885
Saving best model with reconstruction loss 2.3095362
it: 600, train recon loss: 2.013331413269043, local kl: 0.0 global kl: 7.658954018552322e-06 valid reconstr loss: 1.8839631080627441
Saving best model with reconstruction loss 1.8839631
it: 700, train recon loss: 6.427468776702881, local kl: 0.0 global kl: 1.7242318683940994e-11 valid reconstr loss: 2.5326170921325684
it: 800, train recon loss: 1.442582368850708, local kl: 0.0 global kl: 1.933906593892809e-12 valid reconstr loss: 1.3070979118347168
Saving best model with reconstruction loss 1.3070979
it: 900, train recon loss: 1.228488564491272, local kl: 0.0 global kl: 7.468470286653428e-12 valid reconstr loss: 2.1329872608184814
it: 1000, train recon loss: 0.9157993197441101, local kl: 0.0 global kl: 5.390826673945526e-12 valid reconstr loss: 82.42582702636719
it: 1100, train recon loss: 0.8414572477340698, local kl: 0.0 global kl: 4.330674707730964e-11 valid reconstr loss: 0.9408760666847229
Saving best model with reconstruction loss 0.94087607
it: 1200, train recon loss: 2.301609516143799, local kl: 0.0 global kl: 1.6020101911706774e-11 valid reconstr loss: 1.6496531963348389
it: 1300, train recon loss: 0.8274478316307068, local kl: 0.0 global kl: 1.668163003565315e-12 valid reconstr loss: 0.4953565299510956
Saving best model with reconstruction loss 0.49535653
it: 1400, train recon loss: 0.03862258419394493, local kl: 0.0 global kl: 4.879430193227563e-13 valid reconstr loss: 6.373316764831543
it: 1500, train recon loss: 2.5736820697784424, local kl: 0.0 global kl: 2.948585819950722e-11 valid reconstr loss: 1.6008391380310059
it: 1600, train recon loss: 4.206387042999268, local kl: 0.0 global kl: 1.940958505031176e-09 valid reconstr loss: 3.9656691551208496
it: 1700, train recon loss: 3.2723677158355713, local kl: 0.0 global kl: 1.559863349598345e-10 valid reconstr loss: 3.265625476837158
it: 1800, train recon loss: 2.556987762451172, local kl: 0.0 global kl: 1.4033219031261979e-12 valid reconstr loss: 3.6504127979278564
it: 1900, train recon loss: 1.605149507522583, local kl: 0.0 global kl: 2.2737367544323206e-12 valid reconstr loss: 2.1575961112976074
it: 2000, train recon loss: 1.551727056503296, local kl: 0.0 global kl: 2.440728175123752e-11 valid reconstr loss: 1.4750077724456787
it: 2100, train recon loss: 2.3904213905334473, local kl: 0.0 global kl: 1.5368609163068925e-12 valid reconstr loss: 1.349941611289978
it: 2200, train recon loss: 1.5057605504989624, local kl: 0.0 global kl: 3.925284403072382e-11 valid reconstr loss: 1.3217724561691284
it: 2300, train recon loss: 1.6158596277236938, local kl: 0.0 global kl: 1.4904025236184282e-10 valid reconstr loss: 1.3522566556930542
it: 2400, train recon loss: 1.4183858633041382, local kl: 0.0 global kl: 9.186103960834302e-11 valid reconstr loss: 1.4633511304855347
it: 2500, train recon loss: 1.3485759496688843, local kl: 0.0 global kl: 7.511616328947923e-11 valid reconstr loss: 1.0374195575714111
it: 2600, train recon loss: 0.742601752281189, local kl: 0.0 global kl: 2.4802831663506275e-11 valid reconstr loss: 1.1694377660751343
it: 2700, train recon loss: 1.1065515279769897, local kl: 0.0 global kl: 4.6767589800822407e-11 valid reconstr loss: 1.1884558200836182
it: 2800, train recon loss: 0.741425633430481, local kl: 0.0 global kl: 1.5529438629133252e-12 valid reconstr loss: 0.8576903343200684
it: 2900, train recon loss: 1.0040335655212402, local kl: 0.0 global kl: 3.3288649614604537e-12 valid reconstr loss: 0.8865399360656738
it: 3000, train recon loss: 2.120919942855835, local kl: 0.0 global kl: 1.1969814028844894e-10 valid reconstr loss: 2.218445062637329
it: 3100, train recon loss: 1.4773973226547241, local kl: 0.0 global kl: 4.436048750555699e-11 valid reconstr loss: 0.8779517412185669
it: 3200, train recon loss: 1.0602467060089111, local kl: 0.0 global kl: 7.322656370156722e-11 valid reconstr loss: 0.7568404674530029
it: 3300, train recon loss: 0.5801598429679871, local kl: 0.0 global kl: 1.2001510896197942e-12 valid reconstr loss: 0.9460351467132568
it: 3400, train recon loss: 0.6246869564056396, local kl: 0.0 global kl: 1.2763679002603112e-11 valid reconstr loss: 1.0630362033843994
it: 3500, train recon loss: 0.4362879991531372, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.1285197734832764
it: 3600, train recon loss: 0.6210910677909851, local kl: 0.0 global kl: 7.074396624062729e-11 valid reconstr loss: 0.6634728908538818
it: 3700, train recon loss: 0.7951162457466125, local kl: 0.0 global kl: 1.0550744999893169e-10 valid reconstr loss: 0.9024383425712585
it: 3800, train recon loss: 0.49749207496643066, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.8486936688423157
it: 3900, train recon loss: 0.18121983110904694, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.8863438367843628
it: 4000, train recon loss: 0.3301651179790497, local kl: 0.0 global kl: 1.5150380949791042e-12 valid reconstr loss: 0.6803058385848999
it: 4100, train recon loss: 0.28613877296447754, local kl: 0.0 global kl: 2.381783659188841e-10 valid reconstr loss: 0.36462247371673584
Saving best model with reconstruction loss 0.36462247
it: 4200, train recon loss: 0.8264535069465637, local kl: 0.0 global kl: 5.190431418000685e-12 valid reconstr loss: 2.4609122276306152
it: 4300, train recon loss: 0.8543993234634399, local kl: 0.0 global kl: 1.7693485565573042e-10 valid reconstr loss: 0.30376511812210083
Saving best model with reconstruction loss 0.30376512
it: 4400, train recon loss: 0.07254855334758759, local kl: 0.0 global kl: 1.2785406414139722e-11 valid reconstr loss: 0.43997329473495483
it: 4500, train recon loss: 0.26697298884391785, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.19905003905296326
Saving best model with reconstruction loss 0.19905004
it: 4600, train recon loss: -0.2220485806465149, local kl: 0.0 global kl: 1.657366188734244e-10 valid reconstr loss: 0.7914744019508362
it: 4700, train recon loss: 0.745566189289093, local kl: 0.0 global kl: 1.5020016480571385e-12 valid reconstr loss: -0.03924180939793587
Saving best model with reconstruction loss -0.03924181
it: 4800, train recon loss: -0.13650967180728912, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.230497121810913
it: 4900, train recon loss: -0.06424091756343842, local kl: 0.0 global kl: 4.913295464925582e-11 valid reconstr loss: 0.734050989151001
it: 5000, train recon loss: 0.26372000575065613, local kl: 0.0 global kl: 8.988643163121424e-13 valid reconstr loss: -0.15521655976772308
Saving best model with reconstruction loss -0.15521656
it: 5100, train recon loss: 1.9032671451568604, local kl: 0.0 global kl: 4.4926978803871975e-11 valid reconstr loss: 7.3301544189453125
it: 5200, train recon loss: -0.3227443993091583, local kl: 0.0 global kl: 7.982503547054876e-13 valid reconstr loss: -0.16984954476356506
Saving best model with reconstruction loss -0.16984954
it: 5300, train recon loss: -0.5265630483627319, local kl: 0.0 global kl: 9.664491429361988e-13 valid reconstr loss: -0.07736187428236008
it: 5400, train recon loss: 10.336138725280762, local kl: 0.0 global kl: 7.083372777216823e-11 valid reconstr loss: 0.26743176579475403
it: 5500, train recon loss: -0.3290247917175293, local kl: 0.0 global kl: 7.214454728066144e-13 valid reconstr loss: -0.4785858690738678
Saving best model with reconstruction loss -0.47858587
it: 5600, train recon loss: -0.6676117777824402, local kl: 0.0 global kl: 2.5951463200613034e-13 valid reconstr loss: 0.8433348536491394
it: 5700, train recon loss: 0.026425926014780998, local kl: 0.0 global kl: 1.8922038415303266e-14 valid reconstr loss: -0.15852215886116028
it: 5800, train recon loss: -0.4910757541656494, local kl: 0.0 global kl: 3.670674875166924e-13 valid reconstr loss: -0.2946271598339081
it: 5900, train recon loss: -0.5200664401054382, local kl: 0.0 global kl: 1.5350623550069997e-10 valid reconstr loss: -0.30365172028541565
it: 6000, train recon loss: -0.7074864506721497, local kl: 0.0 global kl: 1.6977238842183495e-12 valid reconstr loss: 462.94915771484375
it: 6100, train recon loss: -0.5829346179962158, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.28518134355545044
it: 6200, train recon loss: -0.4531922936439514, local kl: 0.0 global kl: 1.1405876243486546e-10 valid reconstr loss: -0.4317445158958435
it: 6300, train recon loss: 0.22262191772460938, local kl: 0.0 global kl: 2.761679773755077e-13 valid reconstr loss: -0.6137808561325073
Saving best model with reconstruction loss -0.61378086
it: 6400, train recon loss: 25303.3359375, local kl: 0.0 global kl: 2.655098363391062e-12 valid reconstr loss: -0.5018550753593445
it: 6500, train recon loss: 429.5285339355469, local kl: 0.0 global kl: 2.160771561676711e-13 valid reconstr loss: -0.4514773488044739
it: 6600, train recon loss: 121.24698638916016, local kl: 0.0 global kl: 3.1032522471857726e-13 valid reconstr loss: -0.3160794675350189
it: 6700, train recon loss: -0.3141325116157532, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.7669269442558289
it: 6800, train recon loss: -0.49107781052589417, local kl: 0.0 global kl: 2.577217952937083e-12 valid reconstr loss: -0.37086185812950134
it: 6900, train recon loss: -0.672532856464386, local kl: 0.0 global kl: 7.353000847087898e-11 valid reconstr loss: 162.69578552246094
it: 7000, train recon loss: -0.7322274446487427, local kl: 0.0 global kl: 1.571833080360463e-10 valid reconstr loss: -0.2522328495979309
it: 7100, train recon loss: -0.4127886891365051, local kl: 0.0 global kl: 3.110289803487376e-12 valid reconstr loss: 55.42866897583008
it: 7200, train recon loss: -0.5216495990753174, local kl: 0.0 global kl: 5.859895901849654e-13 valid reconstr loss: -0.4156155288219452
it: 7300, train recon loss: -0.4875272810459137, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.7375194430351257
Saving best model with reconstruction loss -0.73751944
it: 7400, train recon loss: -0.2813867926597595, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.3086823523044586
it: 7500, train recon loss: -0.8039935827255249, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.9199823141098022
it: 7600, train recon loss: -0.5555445551872253, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1376.602783203125
it: 7700, train recon loss: -1.1108484268188477, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.47787946462631226
it: 7800, train recon loss: -0.7809039354324341, local kl: 0.0 global kl: 0.0 valid reconstr loss: 107.27689361572266
it: 7900, train recon loss: -0.5003566741943359, local kl: 0.0 global kl: 4.615530180274163e-11 valid reconstr loss: -0.1714334785938263
it: 8000, train recon loss: -0.6623894572257996, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.7146875858306885
it: 8100, train recon loss: -0.10532072186470032, local kl: 0.0 global kl: 0.0 valid reconstr loss: 6.335262298583984
it: 8200, train recon loss: -0.7900740504264832, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.15352264046669006
it: 8300, train recon loss: -0.4858865439891815, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.43013709783554077
it: 8400, train recon loss: -0.7090650200843811, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.10059716552495956
it: 8500, train recon loss: 856852032.0, local kl: 0.0 global kl: 2.702449652947081e-10 valid reconstr loss: 59428552.0
it: 8600, train recon loss: 3.295849561691284, local kl: 0.0 global kl: 1.347016942432333e-10 valid reconstr loss: 3.6928327083587646
it: 8700, train recon loss: 3.0737428665161133, local kl: 0.0 global kl: 3.335665077486283e-12 valid reconstr loss: 2.902636766433716
it: 8800, train recon loss: 3.0676662921905518, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.416639804840088
it: 8900, train recon loss: 1.9557191133499146, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.21804141998291
it: 9000, train recon loss: 1.9018120765686035, local kl: 0.0 global kl: 5.204462555363776e-11 valid reconstr loss: 2.55389142036438
it: 9100, train recon loss: 2.0278944969177246, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.4629584550857544
it: 9200, train recon loss: 7.635969161987305, local kl: 0.0 global kl: 1.2551076844502518e-09 valid reconstr loss: 3.5964195728302
it: 9300, train recon loss: 1.5834382772445679, local kl: 0.0 global kl: 2.9665159217984183e-12 valid reconstr loss: 1.400102972984314
it: 9400, train recon loss: 1.5773686170578003, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.1982861757278442
it: 9500, train recon loss: 2.3317694664001465, local kl: 0.0 global kl: 7.616129948928574e-13 valid reconstr loss: 1.038569688796997
it: 9600, train recon loss: 0.6451359391212463, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.7144956588745117
it: 9700, train recon loss: 3.382817268371582, local kl: 0.0 global kl: 1.5165924072135795e-11 valid reconstr loss: 15.267556190490723
it: 9800, train recon loss: 0.16614319384098053, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.37842458486557007
it: 9900, train recon loss: 0.17719778418540955, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.40386322140693665
beta 5.0 temperature 1000.0
it: 0, train recon loss: 748.1515502929688, local kl: 0.0 global kl: 0.05257103964686394 valid reconstr loss: 431.4855651855469
Saving best model with reconstruction loss 431.48557
it: 100, train recon loss: 3.7586748600006104, local kl: 0.0 global kl: 0.001946666045114398 valid reconstr loss: 4.003453254699707
Saving best model with reconstruction loss 4.0034533
it: 200, train recon loss: 4.0630903244018555, local kl: 0.0 global kl: 3.0729779609828256e-08 valid reconstr loss: 4.207580089569092
it: 300, train recon loss: 3.0155954360961914, local kl: 0.0 global kl: 7.279850979102775e-05 valid reconstr loss: 3.1787145137786865
Saving best model with reconstruction loss 3.1787145
it: 400, train recon loss: 2.3359031677246094, local kl: 0.0 global kl: 2.102415463944851e-12 valid reconstr loss: 2.3872616291046143
Saving best model with reconstruction loss 2.3872616
it: 500, train recon loss: 2.2238833904266357, local kl: 0.0 global kl: 7.778652938786323e-13 valid reconstr loss: 2.23618483543396
Saving best model with reconstruction loss 2.2361848
it: 600, train recon loss: 2.2072036266326904, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.181762218475342
Saving best model with reconstruction loss 2.1817622
it: 700, train recon loss: 1.9633383750915527, local kl: 0.0 global kl: 2.1949336445614698e-11 valid reconstr loss: 1.703882098197937
Saving best model with reconstruction loss 1.7038821
it: 800, train recon loss: 1.1529027223587036, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.7387070655822754
it: 900, train recon loss: 1.2906051874160767, local kl: 0.0 global kl: 8.968936704434327e-12 valid reconstr loss: 1.5528866052627563
Saving best model with reconstruction loss 1.5528866
it: 1000, train recon loss: 1.6574972867965698, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.5892367362976074
it: 1100, train recon loss: 1.2033302783966064, local kl: 0.0 global kl: 9.80937553407557e-12 valid reconstr loss: 1.1590855121612549
Saving best model with reconstruction loss 1.1590855
it: 1200, train recon loss: 1.1717805862426758, local kl: 0.0 global kl: 5.422228638307658e-11 valid reconstr loss: 1.1005762815475464
Saving best model with reconstruction loss 1.1005763
it: 1300, train recon loss: 5.281080722808838, local kl: 0.0 global kl: 8.544970286905595e-12 valid reconstr loss: 1.5682872533798218
it: 1400, train recon loss: 0.8036726713180542, local kl: 0.0 global kl: 5.185851748024106e-12 valid reconstr loss: 2.2815823554992676
it: 1500, train recon loss: 0.9733395576477051, local kl: 0.0 global kl: 2.6697671851039217e-11 valid reconstr loss: 0.8961735963821411
Saving best model with reconstruction loss 0.8961736
it: 1600, train recon loss: 0.7976336479187012, local kl: 0.0 global kl: 1.279532035880493e-12 valid reconstr loss: 1.1843974590301514
it: 1700, train recon loss: 0.7395833134651184, local kl: 0.0 global kl: 8.68792052188816e-11 valid reconstr loss: 0.869539201259613
Saving best model with reconstruction loss 0.8695392
it: 1800, train recon loss: 0.7014865279197693, local kl: 0.0 global kl: 6.618386394485753e-11 valid reconstr loss: 0.7541548609733582
Saving best model with reconstruction loss 0.75415486
it: 1900, train recon loss: 105.52304077148438, local kl: 0.0 global kl: 2.0407616568851594e-11 valid reconstr loss: 1.258491039276123
it: 2000, train recon loss: 1.5334224700927734, local kl: 0.0 global kl: 5.995204332975845e-13 valid reconstr loss: 0.8661565184593201
it: 2100, train recon loss: 1.6326011419296265, local kl: 0.0 global kl: 9.188073912813621e-11 valid reconstr loss: 0.8483567833900452
it: 2200, train recon loss: 0.8456138968467712, local kl: 0.0 global kl: 3.931299730197679e-11 valid reconstr loss: 1.1922253370285034
it: 2300, train recon loss: 0.4570489227771759, local kl: 0.0 global kl: 1.517452830057664e-11 valid reconstr loss: 0.6052210330963135
Saving best model with reconstruction loss 0.60522103
it: 2400, train recon loss: 0.40425190329551697, local kl: 0.0 global kl: 2.7099189212065333e-11 valid reconstr loss: 0.5868588089942932
Saving best model with reconstruction loss 0.5868588
it: 2500, train recon loss: 1.8669445514678955, local kl: 0.0 global kl: 1.62623803312556e-10 valid reconstr loss: 0.6948249340057373
it: 2600, train recon loss: 0.39690306782722473, local kl: 0.0 global kl: 1.0458715143935038e-10 valid reconstr loss: 0.6234596371650696
it: 2700, train recon loss: 4.1233978271484375, local kl: 0.0 global kl: 7.00307867251837e-13 valid reconstr loss: 0.7133060693740845
it: 2800, train recon loss: 0.3279850482940674, local kl: 0.0 global kl: 1.0532704266055792e-12 valid reconstr loss: 0.23931284248828888
Saving best model with reconstruction loss 0.23931284
it: 2900, train recon loss: -0.010912487283349037, local kl: 0.0 global kl: 1.9154157426815033e-11 valid reconstr loss: 0.031802382320165634
Saving best model with reconstruction loss 0.031802382
it: 3000, train recon loss: 0.36013543605804443, local kl: 0.0 global kl: 3.4292013673109523e-13 valid reconstr loss: 3.1264662742614746
it: 3100, train recon loss: -0.08748985081911087, local kl: 0.0 global kl: 5.286396320691722e-13 valid reconstr loss: 0.0888310968875885
it: 3200, train recon loss: -0.09588807821273804, local kl: 0.0 global kl: 1.0735877464806975e-10 valid reconstr loss: 0.2984694242477417
it: 3300, train recon loss: 0.25485384464263916, local kl: 0.0 global kl: 3.079051180687742e-10 valid reconstr loss: 87.97505187988281
it: 3400, train recon loss: 17.32179832458496, local kl: 0.0 global kl: 3.455916108840995e-13 valid reconstr loss: 422.2956237792969
it: 3500, train recon loss: 0.15634164214134216, local kl: 0.0 global kl: 3.624878175401136e-13 valid reconstr loss: -0.040805377066135406
Saving best model with reconstruction loss -0.040805377
it: 3600, train recon loss: 0.16442087292671204, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.06674188375473022
Saving best model with reconstruction loss -0.066741884
it: 3700, train recon loss: 0.11310838162899017, local kl: 0.0 global kl: 1.9780135862568216e-10 valid reconstr loss: 0.07569928467273712
it: 3800, train recon loss: 0.39095038175582886, local kl: 0.0 global kl: 1.0437380126848694e-11 valid reconstr loss: -0.011321601457893848
it: 3900, train recon loss: -0.04866684228181839, local kl: 0.0 global kl: 5.359601651377943e-13 valid reconstr loss: 62.850345611572266
it: 4000, train recon loss: -0.5620202422142029, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.011737968772649765
it: 4100, train recon loss: -0.11699087172746658, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.2682516276836395
Saving best model with reconstruction loss -0.26825163
it: 4200, train recon loss: -0.2595036029815674, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.765408456325531
it: 4300, train recon loss: -0.2524624168872833, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.0050007933750748634
it: 4400, train recon loss: 3107.333984375, local kl: 0.0 global kl: 7.382983113757291e-14 valid reconstr loss: -0.27275550365448
Saving best model with reconstruction loss -0.2727555
it: 4500, train recon loss: 0.004282251466065645, local kl: 0.0 global kl: 1.3503087537003466e-13 valid reconstr loss: 42.468544006347656
it: 4600, train recon loss: 0.3393775522708893, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.14636631309986115
it: 4700, train recon loss: -0.32890087366104126, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.19764086604118347
it: 4800, train recon loss: -0.47873786091804504, local kl: 0.0 global kl: 3.4549064997779766e-11 valid reconstr loss: -0.15335077047348022
it: 4900, train recon loss: 0.3352711796760559, local kl: 0.0 global kl: 3.6453833007765724e-11 valid reconstr loss: -0.2038460373878479
it: 5000, train recon loss: 0.08763082325458527, local kl: 0.0 global kl: 2.637999410928793e-13 valid reconstr loss: -0.39649415016174316
Saving best model with reconstruction loss -0.39649415
it: 5100, train recon loss: -0.6265829801559448, local kl: 0.0 global kl: 6.85203629946507e-11 valid reconstr loss: 0.03401634097099304
it: 5200, train recon loss: 1.081085443496704, local kl: 0.0 global kl: 7.928380174604399e-13 valid reconstr loss: 1.633315086364746
it: 5300, train recon loss: -0.376598596572876, local kl: 0.0 global kl: 5.125067037425879e-13 valid reconstr loss: -0.013872239738702774
it: 5400, train recon loss: 0.13525567948818207, local kl: 0.0 global kl: 4.8396842089459824e-11 valid reconstr loss: 0.8812750577926636
it: 5500, train recon loss: 9.701554298400879, local kl: 0.0 global kl: 1.9565876696603368e-11 valid reconstr loss: 0.07967871427536011
it: 5600, train recon loss: 17.887340545654297, local kl: 0.0 global kl: 5.056458723950996e-13 valid reconstr loss: 0.11343441158533096
it: 5700, train recon loss: 0.2951878607273102, local kl: 0.0 global kl: 6.862566070964249e-12 valid reconstr loss: -0.315959095954895
it: 5800, train recon loss: -0.425724595785141, local kl: 0.0 global kl: 1.0325074129013956e-13 valid reconstr loss: -0.5310346484184265
Saving best model with reconstruction loss -0.53103465
it: 5900, train recon loss: -0.6837438941001892, local kl: 0.0 global kl: 6.40099084847634e-11 valid reconstr loss: -0.020631317049264908
it: 6000, train recon loss: -0.6485440135002136, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.43460091948509216
it: 6100, train recon loss: -0.2964552938938141, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.07903039455413818
it: 6200, train recon loss: -0.7071398496627808, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.5088502764701843
it: 6300, train recon loss: -0.49407345056533813, local kl: 0.0 global kl: 5.0066287921035624e-14 valid reconstr loss: 0.6921007037162781
it: 6400, train recon loss: -0.5299766063690186, local kl: 0.0 global kl: 2.201289150960406e-10 valid reconstr loss: -0.17723184823989868
it: 6500, train recon loss: -0.04633826017379761, local kl: 0.0 global kl: 2.8582136657462343e-11 valid reconstr loss: -0.46343085169792175
it: 6600, train recon loss: 5820.13818359375, local kl: 0.0 global kl: 6.512151928816934e-14 valid reconstr loss: 583.3056030273438
it: 6700, train recon loss: 0.005861266981810331, local kl: 0.0 global kl: 2.8171909249863347e-14 valid reconstr loss: -0.3958592116832733
it: 6800, train recon loss: -0.48275449872016907, local kl: 0.0 global kl: 4.1514101184469965e-14 valid reconstr loss: -0.5794335603713989
Saving best model with reconstruction loss -0.57943356
it: 6900, train recon loss: -0.5048570036888123, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.13143181800842285
it: 7000, train recon loss: -0.43963342905044556, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.4185125231742859
it: 7100, train recon loss: -0.5262599587440491, local kl: 0.0 global kl: 5.083959512056091e-13 valid reconstr loss: -0.8179536461830139
Saving best model with reconstruction loss -0.81795365
it: 7200, train recon loss: -0.46187058091163635, local kl: 0.0 global kl: 3.6329446395644283e-13 valid reconstr loss: 2.966677188873291
it: 7300, train recon loss: -0.41443008184432983, local kl: 0.0 global kl: 2.0926767263507173e-11 valid reconstr loss: -0.16597460210323334
it: 7400, train recon loss: -0.4516221284866333, local kl: 0.0 global kl: 1.1172187480701368e-10 valid reconstr loss: -0.6400021314620972
it: 7500, train recon loss: -0.4455327093601227, local kl: 0.0 global kl: 4.349667327707696e-13 valid reconstr loss: -0.42526334524154663
it: 7600, train recon loss: -0.43989554047584534, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.43495649099349976
it: 7700, train recon loss: -0.8347169756889343, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.27296990156173706
it: 7800, train recon loss: -0.012515339069068432, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.43937358260154724
it: 7900, train recon loss: -0.5700202584266663, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.6262841820716858
it: 8000, train recon loss: -0.636611819267273, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.5535962581634521
it: 8100, train recon loss: 0.3696351945400238, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.7609346508979797
it: 8200, train recon loss: -0.5802475810050964, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.4034234285354614
it: 8300, train recon loss: -0.39236488938331604, local kl: 0.0 global kl: 0.0 valid reconstr loss: 16.449928283691406
it: 8400, train recon loss: 2.2482545375823975, local kl: 0.0 global kl: 0.0 valid reconstr loss: 59.80018615722656
it: 8500, train recon loss: -0.05790024623274803, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.5962142944335938
it: 8600, train recon loss: -0.5684893727302551, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.6310994029045105
it: 8700, train recon loss: -0.45444735884666443, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.7837128639221191
it: 8800, train recon loss: -0.7708426117897034, local kl: 0.0 global kl: 3.3605063176622707e-12 valid reconstr loss: 92.42278289794922
it: 8900, train recon loss: 52.49882888793945, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.06991791725158691
it: 9000, train recon loss: -0.8755432367324829, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.6633850336074829
it: 9100, train recon loss: 145.62384033203125, local kl: 0.0 global kl: 5.775111291961288e-14 valid reconstr loss: 76.48561096191406
it: 9200, train recon loss: -0.6824593544006348, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.06536601483821869
it: 9300, train recon loss: -0.6500487923622131, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.5129042863845825
it: 9400, train recon loss: -0.2594629228115082, local kl: 0.0 global kl: 0.0 valid reconstr loss: 7.69345235824585
it: 9500, train recon loss: -0.6558599472045898, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.3459506928920746
it: 9600, train recon loss: -0.08481516689062119, local kl: 0.0 global kl: 0.0 valid reconstr loss: 192.8319091796875
it: 9700, train recon loss: 0.13692180812358856, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.2190942019224167
it: 9800, train recon loss: -0.8357651829719543, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.746902585029602
it: 9900, train recon loss: 6.974874496459961, local kl: 0.0 global kl: 1.2542744620702706e-12 valid reconstr loss: -0.4711521565914154
beta 1000.0 temperature 0.001
it: 0, train recon loss: 1386866.25, local kl: 0.0 global kl: 8.648841857910156 valid reconstr loss: 13067.8310546875
Saving best model with reconstruction loss 13067.831
it: 100, train recon loss: 4.534503936767578, local kl: 0.0 global kl: 0.0 valid reconstr loss: 4.802481174468994
Saving best model with reconstruction loss 4.802481
it: 200, train recon loss: 2.9682559967041016, local kl: 0.0 global kl: 0.00032507855212315917 valid reconstr loss: 2.813178539276123
Saving best model with reconstruction loss 2.8131785
it: 300, train recon loss: 2.3840208053588867, local kl: 0.0 global kl: 1.7200934962602332e-05 valid reconstr loss: 3.088855028152466
it: 400, train recon loss: 2.450338840484619, local kl: 0.0 global kl: 4.600094325724058e-05 valid reconstr loss: 2.2920517921447754
Saving best model with reconstruction loss 2.2920518
it: 500, train recon loss: 2.0510520935058594, local kl: 0.0 global kl: 0.00017161040159408003 valid reconstr loss: 1.977281093597412
Saving best model with reconstruction loss 1.9772811
it: 600, train recon loss: 6.867739677429199, local kl: 0.0 global kl: 3.2003892556531355e-05 valid reconstr loss: 4.998849391937256
it: 700, train recon loss: 1.6462973356246948, local kl: 0.0 global kl: 0.004530379548668861 valid reconstr loss: 1.9442098140716553
Saving best model with reconstruction loss 1.9442098
it: 800, train recon loss: 1.4736521244049072, local kl: 0.0 global kl: 0.00036161011666990817 valid reconstr loss: 1.4129955768585205
Saving best model with reconstruction loss 1.4129956
it: 900, train recon loss: 1.267403483390808, local kl: 0.0 global kl: 6.504719203803688e-05 valid reconstr loss: 1.3969744443893433
Saving best model with reconstruction loss 1.3969744
it: 1000, train recon loss: 1.5173457860946655, local kl: 0.0 global kl: 0.00023353114374913275 valid reconstr loss: 1.3501920700073242
Saving best model with reconstruction loss 1.3501921
it: 1100, train recon loss: 1.781531810760498, local kl: 0.0 global kl: 0.0005246338550932705 valid reconstr loss: 1.0666358470916748
Saving best model with reconstruction loss 1.0666358
it: 1200, train recon loss: 9.602669715881348, local kl: 0.0 global kl: 0.00010139423829969019 valid reconstr loss: 0.9929043054580688
Saving best model with reconstruction loss 0.9929043
it: 1300, train recon loss: 8.744902610778809, local kl: 0.0 global kl: 0.0010055501479655504 valid reconstr loss: 2.903961420059204
it: 1400, train recon loss: 3.914138078689575, local kl: 0.0 global kl: 0.02096492424607277 valid reconstr loss: 2.8022212982177734
it: 1500, train recon loss: 0.35044997930526733, local kl: 0.0 global kl: 0.00039692819700576365 valid reconstr loss: 5.45297384262085
it: 1600, train recon loss: 0.6241451501846313, local kl: 0.0 global kl: 0.0013543571112677455 valid reconstr loss: 0.964941143989563
Saving best model with reconstruction loss 0.96494114
it: 1700, train recon loss: 0.4931979775428772, local kl: 0.0 global kl: 0.00017124638543464243 valid reconstr loss: 0.9627894759178162
Saving best model with reconstruction loss 0.9627895
it: 1800, train recon loss: 1.2139185667037964, local kl: 0.0 global kl: 0.0004811263643205166 valid reconstr loss: 0.9428209066390991
Saving best model with reconstruction loss 0.9428209
it: 1900, train recon loss: 1.2140682935714722, local kl: 0.0 global kl: 6.456291885115206e-05 valid reconstr loss: 213.8919219970703
it: 2000, train recon loss: 1.614318609237671, local kl: 0.0 global kl: 7.817377627361566e-05 valid reconstr loss: 0.743439257144928
Saving best model with reconstruction loss 0.74343926
it: 2100, train recon loss: 0.23965618014335632, local kl: 0.0 global kl: 0.002204833086580038 valid reconstr loss: 0.8572022318840027
it: 2200, train recon loss: 1.4353243112564087, local kl: 0.0 global kl: 5.8697827626019716e-05 valid reconstr loss: 0.3097130358219147
Saving best model with reconstruction loss 0.30971304
it: 2300, train recon loss: 0.5329698920249939, local kl: 0.0 global kl: 0.00016779839643277228 valid reconstr loss: 0.6011945009231567
it: 2400, train recon loss: 45.47803497314453, local kl: 0.0 global kl: 1.4900380847393535e-05 valid reconstr loss: 0.18256251513957977
Saving best model with reconstruction loss 0.18256252
it: 2500, train recon loss: 1.154317855834961, local kl: 0.0 global kl: 8.078600512817502e-05 valid reconstr loss: 0.5007094144821167
it: 2600, train recon loss: 0.8227810263633728, local kl: 0.0 global kl: 0.000543139991350472 valid reconstr loss: 0.08919510990381241
Saving best model with reconstruction loss 0.08919511
it: 2700, train recon loss: 0.488406777381897, local kl: 0.0 global kl: 0.007811438292264938 valid reconstr loss: 0.6277697086334229
it: 2800, train recon loss: 0.16351260244846344, local kl: 0.0 global kl: 2.880343163269572e-05 valid reconstr loss: 0.4187238812446594
it: 2900, train recon loss: 3.373713970184326, local kl: 0.0 global kl: 0.00032176493550650775 valid reconstr loss: 3.7841858863830566
it: 3000, train recon loss: 0.12531358003616333, local kl: 0.0 global kl: 0.002204085933044553 valid reconstr loss: 1.3347970247268677
it: 3100, train recon loss: 1611.144287109375, local kl: 0.0 global kl: 0.01021568477153778 valid reconstr loss: 2.6502811908721924
it: 3200, train recon loss: 0.46469366550445557, local kl: 0.0 global kl: 0.0033555559348315 valid reconstr loss: 4796.83984375
it: 3300, train recon loss: -0.05223432183265686, local kl: 0.0 global kl: 9.3825892690802e-06 valid reconstr loss: 10.60641860961914
it: 3400, train recon loss: -0.11546049267053604, local kl: 0.0 global kl: 0.00014961560373194516 valid reconstr loss: 0.1820148229598999
it: 3500, train recon loss: 0.45639941096305847, local kl: 0.0 global kl: 1.1571686627576128e-05 valid reconstr loss: 0.06335799396038055
Saving best model with reconstruction loss 0.063357994
it: 3600, train recon loss: 0.1705309897661209, local kl: 0.0 global kl: 0.0006885218317620456 valid reconstr loss: 1428.260986328125
it: 3700, train recon loss: -0.10355561226606369, local kl: 0.0 global kl: 5.497729216585867e-05 valid reconstr loss: 0.2010834962129593
it: 3800, train recon loss: -0.18166489899158478, local kl: 0.0 global kl: 1.5810364857316017e-06 valid reconstr loss: -0.19746513664722443
Saving best model with reconstruction loss -0.19746514
it: 3900, train recon loss: -0.0262365210801363, local kl: 0.0 global kl: 1.6036003671615617e-06 valid reconstr loss: -0.44506293535232544
Saving best model with reconstruction loss -0.44506294
it: 4000, train recon loss: -0.7846373319625854, local kl: 0.0 global kl: 4.0582503970654216e-06 valid reconstr loss: -0.7078155875205994
Saving best model with reconstruction loss -0.7078156
it: 4100, train recon loss: -0.9163408279418945, local kl: 0.0 global kl: 1.7256432329304516e-05 valid reconstr loss: -0.4991103410720825
it: 4200, train recon loss: -0.6969265937805176, local kl: 0.0 global kl: 8.354335534477286e-08 valid reconstr loss: -0.38366177678108215
it: 4300, train recon loss: 1.8644793033599854, local kl: 0.0 global kl: 2.1811958504258655e-05 valid reconstr loss: 2.182846784591675
it: 4400, train recon loss: 0.23978963494300842, local kl: 0.0 global kl: 4.1598901589168236e-05 valid reconstr loss: -0.20555715262889862
it: 4500, train recon loss: 684.7499389648438, local kl: 0.0 global kl: 2.1716487026424147e-06 valid reconstr loss: 1630.6937255859375
it: 4600, train recon loss: 2350.52685546875, local kl: 0.0 global kl: 2.9149016427254537e-06 valid reconstr loss: 2419.816162109375
it: 4700, train recon loss: -0.26610928773880005, local kl: 0.0 global kl: 3.247037238907069e-05 valid reconstr loss: -0.18462485074996948
it: 4800, train recon loss: -1.176965594291687, local kl: 0.0 global kl: 8.390381844947115e-05 valid reconstr loss: -0.711121678352356
Saving best model with reconstruction loss -0.7111217
it: 4900, train recon loss: -0.8780008554458618, local kl: 0.0 global kl: 7.048906809359323e-06 valid reconstr loss: -0.7170150279998779
Saving best model with reconstruction loss -0.717015
it: 5000, train recon loss: -0.7243523597717285, local kl: 0.0 global kl: 5.116812462802045e-05 valid reconstr loss: -0.8998262882232666
Saving best model with reconstruction loss -0.8998263
it: 5100, train recon loss: -0.7557437419891357, local kl: 0.0 global kl: 1.4454542451858288e-06 valid reconstr loss: -0.65633624792099
it: 5200, train recon loss: -0.3766532838344574, local kl: 0.0 global kl: 5.478317461893312e-07 valid reconstr loss: 43.70163345336914
it: 5300, train recon loss: -1.0054141283035278, local kl: 0.0 global kl: 1.4746645149443793e-07 valid reconstr loss: -0.6709219813346863
it: 5400, train recon loss: 3.1207973957061768, local kl: 0.0 global kl: 3.6611504583561327e-07 valid reconstr loss: 778.5327758789062
it: 5500, train recon loss: -1.101106882095337, local kl: 0.0 global kl: 2.619976658024825e-05 valid reconstr loss: -1.071085810661316
Saving best model with reconstruction loss -1.0710858
it: 5600, train recon loss: -1.2547019720077515, local kl: 0.0 global kl: 6.580008005130367e-08 valid reconstr loss: -0.9901995658874512
it: 5700, train recon loss: -0.9905582070350647, local kl: 0.0 global kl: 1.5619740224792622e-05 valid reconstr loss: -0.8269635438919067
it: 5800, train recon loss: 1.769919753074646, local kl: 0.0 global kl: 3.318475137348287e-05 valid reconstr loss: 1.8395779132843018
it: 5900, train recon loss: 0.17121881246566772, local kl: 0.0 global kl: 9.541145118419081e-06 valid reconstr loss: 3.1379172801971436
it: 6000, train recon loss: -0.38782963156700134, local kl: 0.0 global kl: 1.0603925737484587e-08 valid reconstr loss: -0.3482166528701782
it: 6100, train recon loss: 0.07340601831674576, local kl: 0.0 global kl: 1.128073563450016e-05 valid reconstr loss: -0.18655909597873688
it: 6200, train recon loss: -0.5738797783851624, local kl: 0.0 global kl: 4.783870735991513e-06 valid reconstr loss: -0.9230464696884155
it: 6300, train recon loss: -0.21002578735351562, local kl: 0.0 global kl: 5.375477485358715e-06 valid reconstr loss: -0.6690490245819092
it: 6400, train recon loss: -0.6851038336753845, local kl: 0.0 global kl: 1.171456551674055e-05 valid reconstr loss: -0.836295485496521
it: 6500, train recon loss: 4.783352851867676, local kl: 0.0 global kl: 4.7288063797168434e-05 valid reconstr loss: 5.144687652587891
it: 6600, train recon loss: 2.9016263484954834, local kl: 0.0 global kl: 4.575639866999381e-09 valid reconstr loss: 3.297590732574463
it: 6700, train recon loss: 3.081008195877075, local kl: 0.0 global kl: 3.198108444735226e-09 valid reconstr loss: 3.0831503868103027
it: 6800, train recon loss: 3.0698232650756836, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.123002529144287
it: 6900, train recon loss: 3.0389370918273926, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.136267900466919
it: 7000, train recon loss: 3.078207492828369, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.0921406745910645
it: 7100, train recon loss: 2.9718501567840576, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.0555474758148193
it: 7200, train recon loss: 2.997591733932495, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.0471298694610596
it: 7300, train recon loss: 3.2052574157714844, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.0554893016815186
it: 7400, train recon loss: 3.233161449432373, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.028717041015625
it: 7500, train recon loss: 3.0298349857330322, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.036344289779663
it: 7600, train recon loss: 3.113800287246704, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.018014907836914
it: 7700, train recon loss: 2.8256723880767822, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.9984772205352783
it: 7800, train recon loss: 2.9150609970092773, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.0032553672790527
it: 7900, train recon loss: 2.7848339080810547, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.0117201805114746
it: 8000, train recon loss: 2.8857996463775635, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.9248526096343994
it: 8100, train recon loss: 2.9849836826324463, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.97621488571167
it: 8200, train recon loss: 2.9182028770446777, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.981281042098999
it: 8300, train recon loss: 2.961634635925293, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.9355762004852295
it: 8400, train recon loss: 2.7859573364257812, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.9983789920806885
it: 8500, train recon loss: 3.0485434532165527, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.952174425125122
it: 8600, train recon loss: 3.142740249633789, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.9416744709014893
it: 8700, train recon loss: 2.960045337677002, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.936274766921997
it: 8800, train recon loss: 2.785625696182251, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.933330535888672
it: 8900, train recon loss: 2.7759807109832764, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.9309520721435547
it: 9000, train recon loss: 2.737314462661743, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.935082197189331
it: 9100, train recon loss: 3.1176328659057617, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.9659414291381836
it: 9200, train recon loss: 2.87721586227417, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.9350745677948
it: 9300, train recon loss: 2.9181113243103027, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.898247480392456
it: 9400, train recon loss: 3.008823871612549, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.931082010269165
it: 9500, train recon loss: 2.9998295307159424, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.962134838104248
it: 9600, train recon loss: 2.8709678649902344, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.936554431915283
it: 9700, train recon loss: 2.9353315830230713, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.966620683670044
it: 9800, train recon loss: 3.1049563884735107, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.9237351417541504
it: 9900, train recon loss: 2.6896347999572754, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.9306535720825195
beta 1000.0 temperature 0.5
it: 0, train recon loss: 1004.7537841796875, local kl: 0.0 global kl: 12.124734878540039 valid reconstr loss: 453.8767395019531
Saving best model with reconstruction loss 453.87674
it: 100, train recon loss: 3.8898541927337646, local kl: 0.0 global kl: 0.0003853798261843622 valid reconstr loss: 4.059910774230957
Saving best model with reconstruction loss 4.059911
it: 200, train recon loss: 432.8486022949219, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.7830560207366943
Saving best model with reconstruction loss 3.783056
it: 300, train recon loss: 3.1900222301483154, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.096367597579956
Saving best model with reconstruction loss 3.0963676
it: 400, train recon loss: 2.5902347564697266, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.884993314743042
Saving best model with reconstruction loss 2.8849933
it: 500, train recon loss: 1.965448021888733, local kl: 0.0 global kl: 4.862960122409277e-05 valid reconstr loss: 2.1704792976379395
Saving best model with reconstruction loss 2.1704793
it: 600, train recon loss: 1.6610889434814453, local kl: 0.0 global kl: 7.2795519372448325e-06 valid reconstr loss: 1.6731846332550049
Saving best model with reconstruction loss 1.6731846
it: 700, train recon loss: 1.2915630340576172, local kl: 0.0 global kl: 5.8900750445900485e-05 valid reconstr loss: 2.078399419784546
it: 800, train recon loss: 1.8038908243179321, local kl: 0.0 global kl: 6.028321877238341e-05 valid reconstr loss: 1.1517382860183716
Saving best model with reconstruction loss 1.1517383
it: 900, train recon loss: 0.8260079026222229, local kl: 0.0 global kl: 1.7370332443533698e-06 valid reconstr loss: 1.560331106185913
it: 1000, train recon loss: 3.0803349018096924, local kl: 0.0 global kl: 0.00012934797268826514 valid reconstr loss: 2.237260580062866
it: 1100, train recon loss: 794.2515258789062, local kl: 0.0 global kl: 0.0002987803891301155 valid reconstr loss: 3.252950668334961
it: 1200, train recon loss: 0.9902176260948181, local kl: 0.0 global kl: 2.0803234292543493e-05 valid reconstr loss: 0.6549525260925293
Saving best model with reconstruction loss 0.6549525
it: 1300, train recon loss: 4.027535438537598, local kl: 0.0 global kl: 6.30449503660202e-05 valid reconstr loss: 6.766127109527588
it: 1400, train recon loss: 0.3644908368587494, local kl: 0.0 global kl: 4.4603228161577135e-05 valid reconstr loss: 3.9143857955932617
it: 1500, train recon loss: 0.09599295258522034, local kl: 0.0 global kl: 1.4944799175964363e-08 valid reconstr loss: 0.14579856395721436
Saving best model with reconstruction loss 0.14579856
it: 1600, train recon loss: 0.24685882031917572, local kl: 0.0 global kl: 0.00016124898684211075 valid reconstr loss: -0.03873306140303612
Saving best model with reconstruction loss -0.03873306
it: 1700, train recon loss: -0.18763381242752075, local kl: 0.0 global kl: 4.654944802950922e-07 valid reconstr loss: 1.1314311027526855
it: 1800, train recon loss: 0.17082929611206055, local kl: 0.0 global kl: 0.0001362679759040475 valid reconstr loss: 944.6622314453125
it: 1900, train recon loss: -0.0992204025387764, local kl: 0.0 global kl: 5.399833753472194e-05 valid reconstr loss: 1.1363531351089478
it: 2000, train recon loss: 0.01952081359922886, local kl: 0.0 global kl: 0.005132799036800861 valid reconstr loss: -0.08331792056560516
Saving best model with reconstruction loss -0.08331792
it: 2100, train recon loss: -0.4527890086174011, local kl: 0.0 global kl: 0.0013376803835853934 valid reconstr loss: 0.5490128993988037
it: 2200, train recon loss: -0.24883675575256348, local kl: 0.0 global kl: 0.0001495366741437465 valid reconstr loss: -0.2663513720035553
Saving best model with reconstruction loss -0.26635137
it: 2300, train recon loss: -0.4291420876979828, local kl: 0.0 global kl: 2.0929881429765373e-05 valid reconstr loss: -0.2933100163936615
Saving best model with reconstruction loss -0.29331002
it: 2400, train recon loss: 0.0998760387301445, local kl: 0.0 global kl: 2.619062979647424e-05 valid reconstr loss: 1423.899658203125
it: 2500, train recon loss: 281.13287353515625, local kl: 0.0 global kl: 0.0009447872871533036 valid reconstr loss: 1060.593017578125
it: 2600, train recon loss: -0.8216462135314941, local kl: 0.0 global kl: 7.35514986445196e-05 valid reconstr loss: -0.34645822644233704
Saving best model with reconstruction loss -0.34645823
it: 2700, train recon loss: -0.4341384470462799, local kl: 0.0 global kl: 0.0004952301387675107 valid reconstr loss: 0.13144522905349731
it: 2800, train recon loss: -0.4734242260456085, local kl: 0.0 global kl: 0.000264419533777982 valid reconstr loss: 980.5551147460938
it: 2900, train recon loss: -0.5245329141616821, local kl: 0.0 global kl: 6.240852235350758e-05 valid reconstr loss: -0.6335572600364685
Saving best model with reconstruction loss -0.63355726
it: 3000, train recon loss: -0.4821270704269409, local kl: 0.0 global kl: 0.00014713112614117563 valid reconstr loss: -0.6659966111183167
Saving best model with reconstruction loss -0.6659966
it: 3100, train recon loss: -0.6060856580734253, local kl: 0.0 global kl: 2.3527352823293768e-05 valid reconstr loss: -0.6461771130561829
it: 3200, train recon loss: 134.18824768066406, local kl: 0.0 global kl: 3.4487068205635296e-06 valid reconstr loss: 12.159577369689941
it: 3300, train recon loss: -0.6354978680610657, local kl: 0.0 global kl: 0.007038419134914875 valid reconstr loss: -0.7826096415519714
Saving best model with reconstruction loss -0.78260964
it: 3400, train recon loss: 3.225497245788574, local kl: 0.0 global kl: 0.000695478287525475 valid reconstr loss: 7.16570520401001
it: 3500, train recon loss: -0.1504872739315033, local kl: 0.0 global kl: 0.000258596264757216 valid reconstr loss: -0.6382002234458923
it: 3600, train recon loss: 445.4111328125, local kl: 0.0 global kl: 9.065270205610432e-07 valid reconstr loss: -0.6647671461105347
it: 3700, train recon loss: -1.0033177137374878, local kl: 0.0 global kl: 0.0003840770514216274 valid reconstr loss: -0.016299225389957428
it: 3800, train recon loss: 6.01602029800415, local kl: 0.0 global kl: 0.0003066958161070943 valid reconstr loss: -0.7868599891662598
Saving best model with reconstruction loss -0.78686
it: 3900, train recon loss: -0.8642079830169678, local kl: 0.0 global kl: 0.001619972288608551 valid reconstr loss: -0.8814547061920166
Saving best model with reconstruction loss -0.8814547
it: 4000, train recon loss: -1.0665603876113892, local kl: 0.0 global kl: 4.069351234647911e-06 valid reconstr loss: -0.588691771030426
it: 4100, train recon loss: 2302.635009765625, local kl: 0.0 global kl: 0.0002231552207376808 valid reconstr loss: 13.956125259399414
it: 4200, train recon loss: -0.5747277140617371, local kl: 0.0 global kl: 2.5789380742935464e-05 valid reconstr loss: -0.8122577667236328
it: 4300, train recon loss: -0.7621843218803406, local kl: 0.0 global kl: 0.0013077588519081473 valid reconstr loss: -0.8320088386535645
it: 4400, train recon loss: -0.9169511795043945, local kl: 0.0 global kl: 4.0703318518353626e-05 valid reconstr loss: 0.025051524862647057
it: 4500, train recon loss: -1.1188404560089111, local kl: 0.0 global kl: 0.0018495892873033881 valid reconstr loss: -1.0697492361068726
Saving best model with reconstruction loss -1.0697492
it: 4600, train recon loss: -1.1854008436203003, local kl: 0.0 global kl: 0.00012916181003674865 valid reconstr loss: -1.0138577222824097
it: 4700, train recon loss: -0.9714909195899963, local kl: 0.0 global kl: 8.90522642293945e-05 valid reconstr loss: -0.8356085419654846
it: 4800, train recon loss: -1.3963459730148315, local kl: 0.0 global kl: 0.0003726293216459453 valid reconstr loss: 855.3652954101562
it: 4900, train recon loss: -1.0504809617996216, local kl: 0.0 global kl: 0.00024510695948265493 valid reconstr loss: -1.0513261556625366
it: 5000, train recon loss: 2.0021042823791504, local kl: 0.0 global kl: 4.5575266995001584e-05 valid reconstr loss: -1.113426685333252
Saving best model with reconstruction loss -1.1134267
it: 5100, train recon loss: -0.6744396686553955, local kl: 0.0 global kl: 0.00314299832098186 valid reconstr loss: 0.12468969821929932
it: 5200, train recon loss: -0.9082925915718079, local kl: 0.0 global kl: 0.0005466962466016412 valid reconstr loss: -0.9943230152130127
it: 5300, train recon loss: -0.8661925792694092, local kl: 0.0 global kl: 0.0007121275411918759 valid reconstr loss: -1.0966215133666992
it: 5400, train recon loss: -1.195019245147705, local kl: 0.0 global kl: 0.0 valid reconstr loss: 6.752893447875977
it: 5500, train recon loss: -1.2253077030181885, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.1175094842910767
Saving best model with reconstruction loss -1.1175095
it: 5600, train recon loss: -1.2305784225463867, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.051941990852356
it: 5700, train recon loss: 3.686217784881592, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.734928607940674
it: 5800, train recon loss: -1.1644930839538574, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.7874506711959839
it: 5900, train recon loss: 0.716804563999176, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.2114274799823761
it: 6000, train recon loss: -1.3127480745315552, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.2449508905410767
Saving best model with reconstruction loss -1.2449509
it: 6100, train recon loss: -0.23223769664764404, local kl: 0.0 global kl: 0.0009738301741890609 valid reconstr loss: 849.316650390625
it: 6200, train recon loss: 1.8362127542495728, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.9090828895568848
it: 6300, train recon loss: 24.180618286132812, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.9694565534591675
it: 6400, train recon loss: -1.2580775022506714, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.7671046257019043
it: 6500, train recon loss: -1.0319675207138062, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.0361860990524292
it: 6600, train recon loss: -0.8021591901779175, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.3405030965805054
Saving best model with reconstruction loss -1.3405031
it: 6700, train recon loss: -1.017316222190857, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.9925261735916138
it: 6800, train recon loss: -0.5886531472206116, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.3707003891468048
it: 6900, train recon loss: -0.8644788861274719, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.0250598192214966
it: 7000, train recon loss: -0.9718557596206665, local kl: 0.0 global kl: 0.0 valid reconstr loss: 5462.080078125
it: 7100, train recon loss: -1.0265969038009644, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.8756073117256165
it: 7200, train recon loss: -0.8538376688957214, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.0715436935424805
it: 7300, train recon loss: -0.9770018458366394, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.0498508214950562
it: 7400, train recon loss: 4.988245487213135, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.3166821002960205
it: 7500, train recon loss: -1.3740510940551758, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2918.99072265625
it: 7600, train recon loss: -0.7849476337432861, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.114715814590454
it: 7700, train recon loss: -1.4683846235275269, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.8428264856338501
it: 7800, train recon loss: 1.9592883586883545, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.0939823389053345
it: 7900, train recon loss: -0.6145185828208923, local kl: 0.0 global kl: 0.022991973906755447 valid reconstr loss: -1.0946124792099
it: 8000, train recon loss: -1.2340149879455566, local kl: 0.0 global kl: 5.014820203541603e-07 valid reconstr loss: -1.262823462486267
it: 8100, train recon loss: -1.3149783611297607, local kl: 0.0 global kl: 1.351618124090237e-07 valid reconstr loss: -1.1717289686203003
it: 8200, train recon loss: -1.5173745155334473, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.079784870147705
it: 8300, train recon loss: 3.19608736038208, local kl: 0.0 global kl: 0.00016605913697276264 valid reconstr loss: 7855.53076171875
it: 8400, train recon loss: -0.9205703735351562, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.6172406077384949
it: 8500, train recon loss: 18.139419555664062, local kl: 0.0 global kl: 0.0 valid reconstr loss: 8.57971477508545
it: 8600, train recon loss: -0.948206901550293, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.8211671710014343
it: 8700, train recon loss: 16.196006774902344, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.0112322568893433
it: 8800, train recon loss: -1.3435640335083008, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.8925895094871521
it: 8900, train recon loss: -0.9810194373130798, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.103884220123291
it: 9000, train recon loss: -1.3841079473495483, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.2052277326583862
it: 9100, train recon loss: -0.9463165998458862, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.258120059967041
it: 9200, train recon loss: -1.3015375137329102, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.2126877307891846
it: 9300, train recon loss: -0.9956040978431702, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.1020312309265137
it: 9400, train recon loss: -0.6889468431472778, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.9868723750114441
it: 9500, train recon loss: -0.7559865117073059, local kl: 0.0 global kl: 0.0 valid reconstr loss: 902.251953125
it: 9600, train recon loss: -1.0693089962005615, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.2116948366165161
it: 9700, train recon loss: 10946.5458984375, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.225135087966919
it: 9800, train recon loss: 160.27310180664062, local kl: 0.0 global kl: 0.0 valid reconstr loss: 12405.9697265625
it: 9900, train recon loss: -1.1555907726287842, local kl: 0.0 global kl: 0.0 valid reconstr loss: 4410.361328125
beta 1000.0 temperature 1.0
it: 0, train recon loss: 171.95339965820312, local kl: 0.0 global kl: 13.613787651062012 valid reconstr loss: 32.9027099609375
Saving best model with reconstruction loss 32.90271
it: 100, train recon loss: 3.7722623348236084, local kl: 0.0 global kl: 0.0 valid reconstr loss: 4.14141845703125
Saving best model with reconstruction loss 4.1414185
it: 200, train recon loss: 3.9048590660095215, local kl: 0.0 global kl: 0.0005021111574023962 valid reconstr loss: 3.8661413192749023
Saving best model with reconstruction loss 3.8661413
it: 300, train recon loss: 3.1439027786254883, local kl: 0.0 global kl: 0.0050761280581355095 valid reconstr loss: 10.054361343383789
it: 400, train recon loss: 2.469630718231201, local kl: 0.0 global kl: 6.014130121911876e-05 valid reconstr loss: 5.20400857925415
it: 500, train recon loss: 2.4683756828308105, local kl: 0.0 global kl: 0.005205621477216482 valid reconstr loss: 6.908956527709961
it: 600, train recon loss: 1.6040712594985962, local kl: 0.0 global kl: 0.005533711519092321 valid reconstr loss: 1.6384902000427246
Saving best model with reconstruction loss 1.6384902
it: 700, train recon loss: 1.4147205352783203, local kl: 0.0 global kl: 3.2060419471235946e-05 valid reconstr loss: 1.329764723777771
Saving best model with reconstruction loss 1.3297647
it: 800, train recon loss: 1.1123820543289185, local kl: 0.0 global kl: 9.20609018066898e-05 valid reconstr loss: 1.0345708131790161
Saving best model with reconstruction loss 1.0345708
it: 900, train recon loss: 6.522067546844482, local kl: 0.0 global kl: 7.308863860089332e-05 valid reconstr loss: 2.2710211277008057
it: 1000, train recon loss: 0.6777616739273071, local kl: 0.0 global kl: 0.001755769713781774 valid reconstr loss: 1.9712486267089844
it: 1100, train recon loss: 0.1448737382888794, local kl: 0.0 global kl: 7.115841526683653e-06 valid reconstr loss: 0.2828434109687805
Saving best model with reconstruction loss 0.2828434
it: 1200, train recon loss: 0.45936504006385803, local kl: 0.0 global kl: 0.0011108068283647299 valid reconstr loss: 0.28362932801246643
it: 1300, train recon loss: 0.180547833442688, local kl: 0.0 global kl: 0.0008669455419294536 valid reconstr loss: 0.4752729833126068
it: 1400, train recon loss: 92.63143920898438, local kl: 0.0 global kl: 6.529319944093004e-05 valid reconstr loss: 1.236232876777649
it: 1500, train recon loss: 0.4162521958351135, local kl: 0.0 global kl: 0.019252313300967216 valid reconstr loss: 109.28917694091797
it: 1600, train recon loss: -0.18693721294403076, local kl: 0.0 global kl: 0.0001652158098295331 valid reconstr loss: 0.025847190991044044
Saving best model with reconstruction loss 0.025847191
it: 1700, train recon loss: 0.10605596750974655, local kl: 0.0 global kl: 2.9705464839935303e-05 valid reconstr loss: -0.4474470019340515
Saving best model with reconstruction loss -0.447447
it: 1800, train recon loss: -0.37010955810546875, local kl: 0.0 global kl: 7.052799628581852e-05 valid reconstr loss: 2782.434814453125
it: 1900, train recon loss: -0.5651789903640747, local kl: 0.0 global kl: 0.028076106682419777 valid reconstr loss: 1.2613003253936768
it: 2000, train recon loss: -0.05826496705412865, local kl: 0.0 global kl: 0.013697432354092598 valid reconstr loss: -0.5348532795906067
Saving best model with reconstruction loss -0.5348533
it: 2100, train recon loss: -0.24791522324085236, local kl: 0.0 global kl: 0.01719491556286812 valid reconstr loss: -0.0874696597456932
it: 2200, train recon loss: 0.5094929337501526, local kl: 0.0 global kl: 0.026511210948228836 valid reconstr loss: 206.32737731933594
it: 2300, train recon loss: 3.9824023246765137, local kl: 0.0 global kl: 0.0002400205994490534 valid reconstr loss: 4.0016069412231445
it: 2400, train recon loss: 3.482726573944092, local kl: 0.0 global kl: 2.7351575226930436e-06 valid reconstr loss: 4.193971157073975
it: 2500, train recon loss: 3.51224422454834, local kl: 0.0 global kl: 4.036110681227001e-07 valid reconstr loss: 4.31326150894165
it: 2600, train recon loss: 3.2126240730285645, local kl: 0.0 global kl: 4.3460679499673915e-10 valid reconstr loss: 3.368372917175293
it: 2700, train recon loss: 3.2831225395202637, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.6045076847076416
it: 2800, train recon loss: 3.6441280841827393, local kl: 0.0 global kl: 1.5691788146643404e-10 valid reconstr loss: 3.2289717197418213
it: 2900, train recon loss: 3.1424343585968018, local kl: 0.0 global kl: 1.6977548256136288e-09 valid reconstr loss: 3.2228691577911377
it: 3000, train recon loss: 3.099174737930298, local kl: 0.0 global kl: 5.868264207897766e-10 valid reconstr loss: 4.306501865386963
it: 3100, train recon loss: 3.103731393814087, local kl: 0.0 global kl: 4.935635233849212e-10 valid reconstr loss: 3.2057077884674072
it: 3200, train recon loss: 3.267963409423828, local kl: 0.0 global kl: 7.581105743170724e-10 valid reconstr loss: 3.4565248489379883
it: 3300, train recon loss: 2.866044282913208, local kl: 0.0 global kl: 1.6433052696385175e-09 valid reconstr loss: 3.4926252365112305
it: 3400, train recon loss: 3.434757947921753, local kl: 0.0 global kl: 1.1581152703499242e-09 valid reconstr loss: 3.0332376956939697
it: 3500, train recon loss: 2.907733917236328, local kl: 0.0 global kl: 3.497646061667581e-10 valid reconstr loss: 3.3733248710632324
it: 3600, train recon loss: 3.3032078742980957, local kl: 0.0 global kl: 1.301012741983243e-09 valid reconstr loss: 3.0251147747039795
it: 3700, train recon loss: 2.8683106899261475, local kl: 0.0 global kl: 5.162506533373801e-10 valid reconstr loss: 3.0823774337768555
it: 3800, train recon loss: 3.1300177574157715, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.059708833694458
it: 3900, train recon loss: 2.607323408126831, local kl: 0.0 global kl: 7.549516567451064e-12 valid reconstr loss: 2.8770625591278076
it: 4000, train recon loss: 2.3653533458709717, local kl: 0.0 global kl: 2.4869359904755584e-09 valid reconstr loss: 2.7714672088623047
it: 4100, train recon loss: 2.5789666175842285, local kl: 0.0 global kl: 1.5155376953401856e-09 valid reconstr loss: 2.660797119140625
it: 4200, train recon loss: 2.4354517459869385, local kl: 0.0 global kl: 7.759208386914906e-09 valid reconstr loss: 2.46879506111145
it: 4300, train recon loss: 2.120589017868042, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.326597213745117
it: 4400, train recon loss: 2.6011998653411865, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.2218739986419678
it: 4500, train recon loss: 2.2093257904052734, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.1440601348876953
it: 4600, train recon loss: 1.0097625255584717, local kl: 0.0 global kl: 9.736655925962623e-11 valid reconstr loss: 1.259341835975647
it: 4700, train recon loss: 1.1057209968566895, local kl: 0.0 global kl: 7.016609515630989e-11 valid reconstr loss: 1.080528974533081
it: 4800, train recon loss: 1.8023744821548462, local kl: 0.0 global kl: 8.380396931961798e-10 valid reconstr loss: 0.9435340762138367
it: 4900, train recon loss: 0.5346744656562805, local kl: 0.0 global kl: 4.1775088943651895e-10 valid reconstr loss: 1.0438064336776733
it: 5000, train recon loss: 0.7982664704322815, local kl: 0.0 global kl: 1.2583727115877963e-10 valid reconstr loss: 1.003641963005066
it: 5100, train recon loss: 0.7550307512283325, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.461265802383423
it: 5200, train recon loss: 0.3344407081604004, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.8498514890670776
it: 5300, train recon loss: 0.45948299765586853, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.6838691234588623
it: 5400, train recon loss: 0.6501589417457581, local kl: 0.0 global kl: 2.5646151868841116e-10 valid reconstr loss: 0.3277999460697174
it: 5500, train recon loss: 62223.62109375, local kl: 0.0 global kl: 1.0919900539363425e-09 valid reconstr loss: 2.114922523498535
it: 5600, train recon loss: 3.170085906982422, local kl: 0.0 global kl: 1.743905286843983e-08 valid reconstr loss: 3.247122287750244
it: 5700, train recon loss: 0.7551975250244141, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.7445909380912781
it: 5800, train recon loss: 0.776127278804779, local kl: 0.0 global kl: 1.7674750552032492e-10 valid reconstr loss: 0.8630071878433228
it: 5900, train recon loss: 0.25205087661743164, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.34933823347091675
it: 6000, train recon loss: 0.13431128859519958, local kl: 0.0 global kl: 1.4753403831946343e-08 valid reconstr loss: 0.3044179677963257
it: 6100, train recon loss: 0.1636059582233429, local kl: 0.0 global kl: 0.0 valid reconstr loss: 83.42259979248047
it: 6200, train recon loss: -0.06783153116703033, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.37373560667037964
it: 6300, train recon loss: 0.2615624666213989, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.028112106025218964
it: 6400, train recon loss: -0.19959427416324615, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.5070878267288208
it: 6500, train recon loss: -0.06146476790308952, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.12382923811674118
it: 6600, train recon loss: -0.14381487667560577, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.2314511388540268
it: 6700, train recon loss: 1.7651441097259521, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.3339971899986267
it: 6800, train recon loss: -0.043751221150159836, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.11135611683130264
it: 6900, train recon loss: 0.1260432004928589, local kl: 0.0 global kl: 2.3005799931752335e-08 valid reconstr loss: 3.77744197845459
it: 7000, train recon loss: 0.6895469427108765, local kl: 0.0 global kl: 0.0 valid reconstr loss: 6.627508163452148
it: 7100, train recon loss: -0.16779941320419312, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.05823947861790657
it: 7200, train recon loss: -0.4081752896308899, local kl: 0.0 global kl: 2.6954131371326184e-08 valid reconstr loss: -0.11651923507452011
it: 7300, train recon loss: 0.24678511917591095, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.04040776193141937
it: 7400, train recon loss: 0.24625912308692932, local kl: 0.0 global kl: 4.6161163780311654e-10 valid reconstr loss: -0.17319968342781067
it: 7500, train recon loss: -0.43761372566223145, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.21712850034236908
it: 7600, train recon loss: 0.2172260284423828, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.0371805876493454
it: 7700, train recon loss: -0.14082200825214386, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.09076805412769318
it: 7800, train recon loss: -0.41960442066192627, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.37912097573280334
it: 7900, train recon loss: -0.15702956914901733, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.10910150408744812
it: 8000, train recon loss: -0.21978935599327087, local kl: 0.0 global kl: 8.303097653872271e-10 valid reconstr loss: -0.13816934823989868
it: 8100, train recon loss: -0.5207036137580872, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.2815873622894287
it: 8200, train recon loss: -0.4146774411201477, local kl: 0.0 global kl: 2.160771561676711e-10 valid reconstr loss: 0.0761852115392685
it: 8300, train recon loss: 0.13239306211471558, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.42716047167778015
it: 8400, train recon loss: -0.48121708631515503, local kl: 0.0 global kl: 2.2262747201295952e-10 valid reconstr loss: 1.5777812004089355
it: 8500, train recon loss: 0.11697667092084885, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.03629874065518379
it: 8600, train recon loss: 38.450599670410156, local kl: 0.0 global kl: 4.2080228190854996e-10 valid reconstr loss: 53.62388610839844
it: 8700, train recon loss: -0.3277952969074249, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.4704158306121826
it: 8800, train recon loss: -0.6772310733795166, local kl: 0.0 global kl: 5.1962340680278274e-11 valid reconstr loss: -0.09211674332618713
it: 8900, train recon loss: 0.025091616436839104, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.26938316226005554
it: 9000, train recon loss: 3.4158129692077637, local kl: 0.0 global kl: 4.4633741147492856e-10 valid reconstr loss: 5.188159465789795
it: 9100, train recon loss: -0.15251587331295013, local kl: 0.0 global kl: 2.622565706267288e-11 valid reconstr loss: -0.3069547414779663
it: 9200, train recon loss: 1.2624918222427368, local kl: 0.0 global kl: 8.509748461449362e-09 valid reconstr loss: 14.996219635009766
it: 9300, train recon loss: -0.12473133206367493, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.09598895162343979
it: 9400, train recon loss: -0.09782442450523376, local kl: 0.0 global kl: 2.697495005143935e-11 valid reconstr loss: -0.24282722175121307
it: 9500, train recon loss: 5.671321392059326, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.0335235521197319
it: 9600, train recon loss: -0.2173653095960617, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.4147922098636627
it: 9700, train recon loss: -0.19005998969078064, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.30816030502319336
it: 9800, train recon loss: -0.07105842232704163, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.0002596845733933151
it: 9900, train recon loss: 0.20285697281360626, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.3514544665813446
beta 1000.0 temperature 2.0
it: 0, train recon loss: 2830.59912109375, local kl: 0.0 global kl: 22.64004898071289 valid reconstr loss: 374.8570556640625
Saving best model with reconstruction loss 374.85706
it: 100, train recon loss: 3.6864914894104004, local kl: 0.0 global kl: 7.28012528270483e-05 valid reconstr loss: 3.8947410583496094
Saving best model with reconstruction loss 3.894741
it: 200, train recon loss: 3.870795249938965, local kl: 0.0 global kl: 0.016738569363951683 valid reconstr loss: 3.8615193367004395
Saving best model with reconstruction loss 3.8615193
it: 300, train recon loss: 3.5348589420318604, local kl: 0.0 global kl: 0.001974090002477169 valid reconstr loss: 8.370223045349121
it: 400, train recon loss: 2.8313851356506348, local kl: 0.0 global kl: 0.0002603168832138181 valid reconstr loss: 3.138598680496216
Saving best model with reconstruction loss 3.1385987
it: 500, train recon loss: 2.426149368286133, local kl: 0.0 global kl: 0.0015716779744252563 valid reconstr loss: 2.530210018157959
Saving best model with reconstruction loss 2.53021
it: 600, train recon loss: 4.272588729858398, local kl: 0.0 global kl: 1.9947607142967172e-05 valid reconstr loss: 3.155198574066162
it: 700, train recon loss: 1.7631969451904297, local kl: 0.0 global kl: 7.401301793663606e-10 valid reconstr loss: 1.918485164642334
Saving best model with reconstruction loss 1.9184852
it: 800, train recon loss: 1.461991310119629, local kl: 0.0 global kl: 1.2212453270876722e-11 valid reconstr loss: 2.012526035308838
it: 900, train recon loss: 1.132763385772705, local kl: 0.0 global kl: 2.8532731732866523e-11 valid reconstr loss: 2.255603551864624
it: 1000, train recon loss: 1.587320327758789, local kl: 0.0 global kl: 2.3314683517128287e-12 valid reconstr loss: 1.3569506406784058
Saving best model with reconstruction loss 1.3569506
it: 1100, train recon loss: 0.7077456712722778, local kl: 0.0 global kl: 1.382893799473095e-09 valid reconstr loss: 1.3101352453231812
Saving best model with reconstruction loss 1.3101352
it: 1200, train recon loss: 1.3059192895889282, local kl: 0.0 global kl: 1.7071456748229963e-10 valid reconstr loss: 1.0214046239852905
Saving best model with reconstruction loss 1.0214046
it: 1300, train recon loss: 0.5460055470466614, local kl: 0.0 global kl: 9.880984919163893e-12 valid reconstr loss: 0.8622444272041321
Saving best model with reconstruction loss 0.8622444
it: 1400, train recon loss: 0.6116318106651306, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.6914056539535522
Saving best model with reconstruction loss 0.69140565
it: 1500, train recon loss: 0.5029886960983276, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.2435383796691895
it: 1600, train recon loss: 0.4861750900745392, local kl: 0.0 global kl: 5.861686247499165e-09 valid reconstr loss: 0.753268837928772
it: 1700, train recon loss: 0.3267308473587036, local kl: 0.0 global kl: 1.4211548604592394e-10 valid reconstr loss: 411.1529235839844
it: 1800, train recon loss: 1.1313860416412354, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.3267960548400879
Saving best model with reconstruction loss 0.32679605
it: 1900, train recon loss: 0.4520074725151062, local kl: 0.0 global kl: 5.0530690742789375e-09 valid reconstr loss: 1.4009572267532349
it: 2000, train recon loss: 0.26463693380355835, local kl: 0.0 global kl: 2.1250528003946556e-09 valid reconstr loss: 0.09850317984819412
Saving best model with reconstruction loss 0.09850318
it: 2100, train recon loss: 0.11804132908582687, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.36298424005508423
it: 2200, train recon loss: 0.12830834090709686, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.0510951466858387
Saving best model with reconstruction loss -0.051095147
it: 2300, train recon loss: 0.04558488726615906, local kl: 0.0 global kl: 3.5062230896443225e-10 valid reconstr loss: 0.010082683525979519
it: 2400, train recon loss: 0.25559720396995544, local kl: 0.0 global kl: 8.495870673641548e-09 valid reconstr loss: -0.036474574357271194
it: 2500, train recon loss: 0.14396926760673523, local kl: 0.0 global kl: 5.694333893302428e-10 valid reconstr loss: 0.02726726606488228
it: 2600, train recon loss: -0.1391419917345047, local kl: 0.0 global kl: 8.663958439569797e-09 valid reconstr loss: -0.0777050331234932
Saving best model with reconstruction loss -0.07770503
it: 2700, train recon loss: -0.05377872288227081, local kl: 0.0 global kl: 4.7003094749697993e-08 valid reconstr loss: -0.12460716813802719
Saving best model with reconstruction loss -0.12460717
it: 2800, train recon loss: -0.40148526430130005, local kl: 0.0 global kl: 2.8936961715686493e-09 valid reconstr loss: 3.6074695587158203
it: 2900, train recon loss: -0.25278645753860474, local kl: 0.0 global kl: 1.1154660306544883e-08 valid reconstr loss: 8.72559928894043
it: 3000, train recon loss: -0.23837022483348846, local kl: 0.0 global kl: 3.0734583766900414e-09 valid reconstr loss: -0.11705505102872849
it: 3100, train recon loss: -0.28612077236175537, local kl: 0.0 global kl: 0.0 valid reconstr loss: 24.564571380615234
it: 3200, train recon loss: -0.5390169620513916, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.22412890195846558
it: 3300, train recon loss: -0.4898863732814789, local kl: 0.0 global kl: 2.4082419614046557e-08 valid reconstr loss: 20.668556213378906
it: 3400, train recon loss: -0.6231452822685242, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.49751898646354675
Saving best model with reconstruction loss -0.497519
it: 3500, train recon loss: -0.3735903799533844, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.6182355880737305
Saving best model with reconstruction loss -0.6182356
it: 3600, train recon loss: -0.2661825120449066, local kl: 0.0 global kl: 4.6800142428082836e-08 valid reconstr loss: -0.5340527296066284
it: 3700, train recon loss: -0.5281475186347961, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.6198433041572571
Saving best model with reconstruction loss -0.6198433
it: 3800, train recon loss: 16.799503326416016, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2457.916748046875
it: 3900, train recon loss: -0.5258933305740356, local kl: 0.0 global kl: 5.520862655927772e-10 valid reconstr loss: -0.35400012135505676
it: 4000, train recon loss: -0.7749700546264648, local kl: 0.0 global kl: 5.123440560694803e-10 valid reconstr loss: 10.637295722961426
it: 4100, train recon loss: -0.3735467493534088, local kl: 0.0 global kl: 5.900835375882707e-10 valid reconstr loss: -0.8392845988273621
Saving best model with reconstruction loss -0.8392846
it: 4200, train recon loss: 0.6952241659164429, local kl: 0.0 global kl: 6.353322312691034e-09 valid reconstr loss: 158.92166137695312
it: 4300, train recon loss: -0.7950069308280945, local kl: 0.0 global kl: 8.711615429035646e-09 valid reconstr loss: -0.4970564544200897
it: 4400, train recon loss: -0.46011650562286377, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.619753360748291
it: 4500, train recon loss: -0.7769359946250916, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.1411489099264145
it: 4600, train recon loss: -0.7384469509124756, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.6914627552032471
it: 4700, train recon loss: -0.783388614654541, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.45884057879447937
it: 4800, train recon loss: -0.9506803750991821, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.713187038898468
it: 4900, train recon loss: -0.9713264107704163, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.7100698351860046
it: 5000, train recon loss: -0.6381437182426453, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.060412779450416565
it: 5100, train recon loss: -0.8740845322608948, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.5801623463630676
it: 5200, train recon loss: -0.884093701839447, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.27989599108695984
it: 5300, train recon loss: -0.5659895539283752, local kl: 0.0 global kl: 4.871610226331313e-09 valid reconstr loss: -0.21612930297851562
it: 5400, train recon loss: 0.33819881081581116, local kl: 0.0 global kl: 1.4844306228667392e-08 valid reconstr loss: 2.2199878692626953
it: 5500, train recon loss: 284.3802795410156, local kl: 0.0 global kl: 1.1896039708858552e-10 valid reconstr loss: 804.3907470703125
it: 5600, train recon loss: -1.2414350509643555, local kl: 0.0 global kl: 3.629746156419422e-11 valid reconstr loss: 50.58626174926758
it: 5700, train recon loss: -0.6058624982833862, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.8105905652046204
it: 5800, train recon loss: -0.9103280901908875, local kl: 0.0 global kl: 1.9365065107024293e-10 valid reconstr loss: -0.857565701007843
Saving best model with reconstruction loss -0.8575657
it: 5900, train recon loss: -0.7139927744865417, local kl: 0.0 global kl: 4.643300410833717e-09 valid reconstr loss: -0.7275523543357849
it: 6000, train recon loss: -0.9558155536651611, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.0353659391403198
Saving best model with reconstruction loss -1.0353659
it: 6100, train recon loss: -0.14441579580307007, local kl: 0.0 global kl: 2.552680289369391e-10 valid reconstr loss: -0.6894665360450745
it: 6200, train recon loss: -1.039861798286438, local kl: 0.0 global kl: 0.0 valid reconstr loss: 93.3154525756836
it: 6300, train recon loss: 41.70265579223633, local kl: 0.0 global kl: 2.0703778247366245e-08 valid reconstr loss: -0.6473962664604187
it: 6400, train recon loss: 100.67060852050781, local kl: 0.0 global kl: 1.2567339169322622e-08 valid reconstr loss: 0.4938279390335083
it: 6500, train recon loss: -0.8159927129745483, local kl: 0.0 global kl: 0.0 valid reconstr loss: 6.534568786621094
it: 6600, train recon loss: -0.9410176873207092, local kl: 0.0 global kl: 8.134118378855248e-12 valid reconstr loss: 0.6451979875564575
it: 6700, train recon loss: -0.2856519818305969, local kl: 0.0 global kl: 2.4350747113155435e-11 valid reconstr loss: -0.0954553410410881
it: 6800, train recon loss: -0.9819750189781189, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.1131856441497803
Saving best model with reconstruction loss -1.1131856
it: 6900, train recon loss: -0.8529981374740601, local kl: 0.0 global kl: 5.843702854946287e-09 valid reconstr loss: 1.593384861946106
it: 7000, train recon loss: -0.7608273029327393, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.0731557607650757
it: 7100, train recon loss: -0.96822190284729, local kl: 0.0 global kl: 1.8020307468447072e-11 valid reconstr loss: -0.8754223585128784
it: 7200, train recon loss: 9.964974403381348, local kl: 0.0 global kl: 6.761258219967203e-11 valid reconstr loss: -0.6767333149909973
it: 7300, train recon loss: 1277.9097900390625, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.07495217770338058
it: 7400, train recon loss: -0.8375864624977112, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.8676220774650574
it: 7500, train recon loss: -1.3434027433395386, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.44781631231307983
it: 7600, train recon loss: -0.9949004054069519, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.20589753985404968
it: 7700, train recon loss: -0.9453920722007751, local kl: 0.0 global kl: 0.0 valid reconstr loss: 6526.95361328125
it: 7800, train recon loss: -0.9789835214614868, local kl: 0.0 global kl: 0.0 valid reconstr loss: 8.845609664916992
it: 7900, train recon loss: -1.0009170770645142, local kl: 0.0 global kl: 0.0 valid reconstr loss: 38.8994255065918
it: 8000, train recon loss: -0.40163975954055786, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.1861755847930908
Saving best model with reconstruction loss -1.1861756
it: 8100, train recon loss: -0.1262546330690384, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.29884228110313416
it: 8200, train recon loss: 2.8736209869384766, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.473918080329895
it: 8300, train recon loss: -1.018197774887085, local kl: 0.0 global kl: 6.442069100387471e-11 valid reconstr loss: 12.801867485046387
it: 8400, train recon loss: -1.0880961418151855, local kl: 0.0 global kl: 0.0 valid reconstr loss: 189.53187561035156
it: 8500, train recon loss: 19.572418212890625, local kl: 0.0 global kl: 6.0285110237146e-11 valid reconstr loss: 8.735387802124023
it: 8600, train recon loss: 487.99896240234375, local kl: 0.0 global kl: 8.243475346780826e-10 valid reconstr loss: 109.85768127441406
it: 8700, train recon loss: 21.149425506591797, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.6014083623886108
it: 8800, train recon loss: -0.7999836206436157, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.9659964442253113
it: 8900, train recon loss: -1.140520453453064, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.1838780641555786
it: 9000, train recon loss: 5.054104328155518, local kl: 0.0 global kl: 0.0 valid reconstr loss: 4.911561012268066
it: 9100, train recon loss: -1.0351406335830688, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.943419337272644
it: 9200, train recon loss: -1.1504242420196533, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3219.60546875
it: 9300, train recon loss: -0.7742071151733398, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.0549429655075073
it: 9400, train recon loss: -1.0337202548980713, local kl: 0.0 global kl: 0.0 valid reconstr loss: 9.118828773498535
it: 9500, train recon loss: 916.4674072265625, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.7076783180236816
it: 9600, train recon loss: -1.0426603555679321, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.226477861404419
Saving best model with reconstruction loss -1.2264779
it: 9700, train recon loss: -1.337982416152954, local kl: 0.0 global kl: 0.0 valid reconstr loss: 937.7999267578125
it: 9800, train recon loss: -1.023287057876587, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.2640912532806396
Saving best model with reconstruction loss -1.2640913
it: 9900, train recon loss: 27.961076736450195, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.7412933707237244
beta 1000.0 temperature 5.0
it: 0, train recon loss: 481.9523010253906, local kl: 0.0 global kl: 11.392507553100586 valid reconstr loss: 234.53468322753906
Saving best model with reconstruction loss 234.53468
it: 100, train recon loss: 4.7559709548950195, local kl: 0.0 global kl: 0.23184195160865784 valid reconstr loss: 4.0961079597473145
Saving best model with reconstruction loss 4.096108
it: 200, train recon loss: 3.990844488143921, local kl: 0.0 global kl: 0.0 valid reconstr loss: 4.197142124176025
it: 300, train recon loss: 3.5173943042755127, local kl: 0.0 global kl: 0.004144585691392422 valid reconstr loss: 3.4538707733154297
Saving best model with reconstruction loss 3.4538708
it: 400, train recon loss: 2.324218273162842, local kl: 0.0 global kl: 0.013160687871277332 valid reconstr loss: 2.899407386779785
Saving best model with reconstruction loss 2.8994074
it: 500, train recon loss: 1.7182046175003052, local kl: 0.0 global kl: 9.356424925499596e-06 valid reconstr loss: 1.6205971240997314
Saving best model with reconstruction loss 1.6205971
it: 600, train recon loss: 1.2281949520111084, local kl: 0.0 global kl: 8.992806499463768e-12 valid reconstr loss: 1.6405506134033203
it: 700, train recon loss: 1.1275184154510498, local kl: 0.0 global kl: 6.955425679855409e-10 valid reconstr loss: 1.554532766342163
Saving best model with reconstruction loss 1.5545328
it: 800, train recon loss: 0.9825175404548645, local kl: 0.0 global kl: 7.105427357601002e-12 valid reconstr loss: 1.4377185106277466
Saving best model with reconstruction loss 1.4377185
it: 900, train recon loss: 355.3116760253906, local kl: 0.0 global kl: 8.797962358642053e-09 valid reconstr loss: 1.1519123315811157
Saving best model with reconstruction loss 1.1519123
it: 1000, train recon loss: 1.9663387537002563, local kl: 0.0 global kl: 1.9372559112440513e-09 valid reconstr loss: 0.6496678590774536
Saving best model with reconstruction loss 0.64966786
it: 1100, train recon loss: 0.7952372431755066, local kl: 0.0 global kl: 8.486017222253395e-09 valid reconstr loss: 5.1332573890686035
it: 1200, train recon loss: 0.33243268728256226, local kl: 0.0 global kl: 4.074796056130481e-10 valid reconstr loss: 0.5681175589561462
Saving best model with reconstruction loss 0.56811756
it: 1300, train recon loss: 66.7418441772461, local kl: 0.0 global kl: 4.914984863546579e-09 valid reconstr loss: 421.68621826171875
it: 1400, train recon loss: -0.18669472634792328, local kl: 0.0 global kl: 6.901062388209311e-09 valid reconstr loss: 14.291086196899414
it: 1500, train recon loss: 0.14280083775520325, local kl: 0.0 global kl: 1.4613310561628623e-10 valid reconstr loss: 0.40447282791137695
Saving best model with reconstruction loss 0.40447283
it: 1600, train recon loss: 0.3448224663734436, local kl: 0.0 global kl: 2.9774099852275526e-10 valid reconstr loss: 0.29188376665115356
Saving best model with reconstruction loss 0.29188377
it: 1700, train recon loss: 26.86540412902832, local kl: 0.0 global kl: 5.963675386855272e-11 valid reconstr loss: 30.495792388916016
it: 1800, train recon loss: 0.27398157119750977, local kl: 0.0 global kl: 2.244768593229196e-09 valid reconstr loss: -0.16402682662010193
Saving best model with reconstruction loss -0.16402683
it: 1900, train recon loss: -0.2829984128475189, local kl: 0.0 global kl: 4.759803662324202e-10 valid reconstr loss: 0.4682922065258026
it: 2000, train recon loss: -0.05069595202803612, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.20538835227489471
Saving best model with reconstruction loss -0.20538835
it: 2100, train recon loss: -0.4712778329849243, local kl: 0.0 global kl: 2.2977987512717846e-08 valid reconstr loss: 0.06177531182765961
it: 2200, train recon loss: 0.21313756704330444, local kl: 0.0 global kl: 9.825473767932635e-11 valid reconstr loss: -0.10562440752983093
it: 2300, train recon loss: -0.2605723738670349, local kl: 0.0 global kl: 8.484879465697759e-11 valid reconstr loss: 37.90656280517578
it: 2400, train recon loss: -0.36259111762046814, local kl: 0.0 global kl: 4.0446812565875234e-11 valid reconstr loss: 0.300076961517334
it: 2500, train recon loss: 0.5586583018302917, local kl: 0.0 global kl: 1.795177340113696e-08 valid reconstr loss: -0.36220166087150574
Saving best model with reconstruction loss -0.36220166
it: 2600, train recon loss: -0.6246392726898193, local kl: 0.0 global kl: 1.1561204438237382e-08 valid reconstr loss: -0.2013852447271347
it: 2700, train recon loss: 21.16312026977539, local kl: 0.0 global kl: 4.468647674116255e-11 valid reconstr loss: -0.29075169563293457
it: 2800, train recon loss: -0.4106796383857727, local kl: 0.0 global kl: 5.995204332975845e-12 valid reconstr loss: -0.25930455327033997
it: 2900, train recon loss: -0.701556921005249, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.9727115631103516
it: 3000, train recon loss: -0.49517905712127686, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.2756502032279968
it: 3100, train recon loss: -0.8494720458984375, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.49304670095443726
Saving best model with reconstruction loss -0.4930467
it: 3200, train recon loss: -0.6681375503540039, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.49454793334007263
it: 3300, train recon loss: 36.53773498535156, local kl: 0.0 global kl: 0.0 valid reconstr loss: 760.1763916015625
it: 3400, train recon loss: -0.8651407957077026, local kl: 0.0 global kl: 4.807189313282834e-09 valid reconstr loss: 0.24632887542247772
it: 3500, train recon loss: -0.3956962525844574, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.7686842083930969
Saving best model with reconstruction loss -0.7686842
it: 3600, train recon loss: -0.7073336243629456, local kl: 0.0 global kl: 3.574918139293004e-11 valid reconstr loss: -0.876846194267273
Saving best model with reconstruction loss -0.8768462
it: 3700, train recon loss: -1.0557907819747925, local kl: 0.0 global kl: 9.811929047032208e-09 valid reconstr loss: 3.2818548679351807
it: 3800, train recon loss: -0.790839672088623, local kl: 0.0 global kl: 5.384581669432009e-11 valid reconstr loss: 0.8143050074577332
it: 3900, train recon loss: -0.7777918577194214, local kl: 0.0 global kl: 3.930286998632404e-11 valid reconstr loss: 97.71871948242188
it: 4000, train recon loss: -0.814186692237854, local kl: 0.0 global kl: 1.4536067460824142e-08 valid reconstr loss: -0.14574328064918518
it: 4100, train recon loss: -0.8712564706802368, local kl: 0.0 global kl: 4.999322911203308e-08 valid reconstr loss: 25.58209228515625
it: 4200, train recon loss: -0.0746040940284729, local kl: 0.0 global kl: 0.0 valid reconstr loss: 18.153467178344727
it: 4300, train recon loss: 4174.9326171875, local kl: 0.0 global kl: 1.4886009092052177e-10 valid reconstr loss: -0.526561975479126
it: 4400, train recon loss: 1.942061185836792, local kl: 0.0 global kl: 0.0 valid reconstr loss: 240.01126098632812
it: 4500, train recon loss: -0.187649667263031, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.6865897178649902
it: 4600, train recon loss: -1.1695542335510254, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.953014850616455
it: 4700, train recon loss: -0.8598236441612244, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.04438600316643715
it: 4800, train recon loss: -1.106508493423462, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.260302782058716
it: 4900, train recon loss: -1.0967929363250732, local kl: 0.0 global kl: 7.438543114801632e-09 valid reconstr loss: -0.5474348068237305
it: 5000, train recon loss: 1.1644777059555054, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.675676167011261
it: 5100, train recon loss: -0.9945460557937622, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.559056282043457
it: 5200, train recon loss: -0.323664128780365, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.8762868046760559
it: 5300, train recon loss: -0.6296323537826538, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.9266201853752136
Saving best model with reconstruction loss -0.9266202
it: 5400, train recon loss: 3024.176513671875, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.3679686784744263
it: 5500, train recon loss: -0.771611213684082, local kl: 0.0 global kl: 3.6647853640658923e-09 valid reconstr loss: 1.4399479627609253
it: 5600, train recon loss: -0.9082351922988892, local kl: 0.0 global kl: 1.0733081090563701e-10 valid reconstr loss: -0.9597377777099609
Saving best model with reconstruction loss -0.9597378
it: 5700, train recon loss: -0.7724238038063049, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.7110834717750549
it: 5800, train recon loss: 6.8489251136779785, local kl: 0.0 global kl: 1.572353358625378e-10 valid reconstr loss: -0.7106177806854248
it: 5900, train recon loss: -0.9270647764205933, local kl: 0.0 global kl: 2.40766608039511e-11 valid reconstr loss: -0.7003067135810852
it: 6000, train recon loss: 3.867945909500122, local kl: 0.0 global kl: 4.81906181626357e-11 valid reconstr loss: 8805.9677734375
it: 6100, train recon loss: -0.8736175894737244, local kl: 0.0 global kl: 0.0 valid reconstr loss: 118.25334930419922
it: 6200, train recon loss: -0.8411880135536194, local kl: 0.0 global kl: 1.301736496372996e-11 valid reconstr loss: 48.02980041503906
it: 6300, train recon loss: -0.680457592010498, local kl: 0.0 global kl: 2.2977696190196184e-08 valid reconstr loss: -0.9299380779266357
it: 6400, train recon loss: 1974.8758544921875, local kl: 0.0 global kl: 6.116273709722009e-09 valid reconstr loss: 15.451729774475098
it: 6500, train recon loss: -1.1954952478408813, local kl: 0.0 global kl: 4.618628146602077e-09 valid reconstr loss: -0.7139122486114502
it: 6600, train recon loss: 0.9358693957328796, local kl: 0.0 global kl: 1.6076716846669115e-08 valid reconstr loss: -0.9320867657661438
it: 6700, train recon loss: -1.1909730434417725, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.2591273784637451
Saving best model with reconstruction loss -1.2591274
it: 6800, train recon loss: -1.1959775686264038, local kl: 0.0 global kl: 0.0 valid reconstr loss: 287.1063537597656
it: 6900, train recon loss: 190.42596435546875, local kl: 0.0 global kl: 0.0 valid reconstr loss: 5.64970064163208
it: 7000, train recon loss: -1.0646460056304932, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.012980580329895
it: 7100, train recon loss: -0.7814540863037109, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.9297018647193909
it: 7200, train recon loss: -1.256139874458313, local kl: 0.0 global kl: 1.1236927832669608e-08 valid reconstr loss: -0.5820314884185791
it: 7300, train recon loss: -1.2020190954208374, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.6322218179702759
it: 7400, train recon loss: -0.7309173345565796, local kl: 0.0 global kl: 1.0586965970560414e-08 valid reconstr loss: -0.9334519505500793
it: 7500, train recon loss: -1.4184587001800537, local kl: 0.0 global kl: 5.5712379154471137e-11 valid reconstr loss: 4.138049602508545
it: 7600, train recon loss: -1.012431025505066, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.8327440023422241
it: 7700, train recon loss: -1.0606862306594849, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.0704079866409302
it: 7800, train recon loss: -0.8463085889816284, local kl: 0.0 global kl: 0.0 valid reconstr loss: 269.3778381347656
it: 7900, train recon loss: 791.4749755859375, local kl: 0.0 global kl: 4.262215580475015e-12 valid reconstr loss: 9.253677368164062
it: 8000, train recon loss: -1.2217313051223755, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.930561363697052
it: 8100, train recon loss: 47.48126983642578, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.0123920440673828
it: 8200, train recon loss: -1.2794079780578613, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.7989022731781006
it: 8300, train recon loss: -1.1634984016418457, local kl: 0.0 global kl: 0.0 valid reconstr loss: 21.824687957763672
it: 8400, train recon loss: -1.2905417680740356, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.6629566550254822
it: 8500, train recon loss: -1.0570045709609985, local kl: 0.0 global kl: 0.0 valid reconstr loss: 307.5519104003906
it: 8600, train recon loss: 0.046395979821681976, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2226.463134765625
it: 8700, train recon loss: 1663.32666015625, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2381.68212890625
it: 8800, train recon loss: -1.236755132675171, local kl: 0.0 global kl: 0.0 valid reconstr loss: -0.8229465484619141
it: 8900, train recon loss: 8.89980411529541, local kl: 0.0 global kl: 0.0 valid reconstr loss: 3.403883218765259
it: 9000, train recon loss: 5.27682638168335, local kl: 0.0 global kl: 0.0 valid reconstr loss: 27.519794464111328
it: 9100, train recon loss: 1.7778254747390747, local kl: 0.0 global kl: 0.0 valid reconstr loss: 23195.720703125
it: 9200, train recon loss: 0.8421308398246765, local kl: 0.0 global kl: 0.0 valid reconstr loss: 23.222558975219727
it: 9300, train recon loss: -0.2537040710449219, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.4701478481292725
it: 9400, train recon loss: 0.5114028453826904, local kl: 0.0 global kl: 0.0 valid reconstr loss: 41447.51953125
it: 9500, train recon loss: -0.45569854974746704, local kl: 0.0 global kl: 0.0 valid reconstr loss: -1.1658815145492554
it: 9600, train recon loss: -0.8958063125610352, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.5952391624450684
it: 9700, train recon loss: 88.8175048828125, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.234168529510498
it: 9800, train recon loss: -1.1524317264556885, local kl: 0.0 global kl: 8.113648641838722e-11 valid reconstr loss: 7.737255096435547
it: 9900, train recon loss: -1.234090805053711, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.037634048610925674
beta 1000.0 temperature 1000.0
it: 0, train recon loss: 6012.85107421875, local kl: 0.0 global kl: 47.132041931152344 valid reconstr loss: 882.06787109375
Saving best model with reconstruction loss 882.0679
it: 100, train recon loss: 4.965723991394043, local kl: 0.0 global kl: 0.0 valid reconstr loss: 4.2424468994140625
Saving best model with reconstruction loss 4.242447
it: 200, train recon loss: 4.12451171875, local kl: 0.0 global kl: 0.0 valid reconstr loss: 4.0316572189331055
Saving best model with reconstruction loss 4.031657
it: 300, train recon loss: 3.496910810470581, local kl: 0.0 global kl: 1.5540535969194025e-06 valid reconstr loss: 3.580368995666504
Saving best model with reconstruction loss 3.580369
it: 400, train recon loss: 3.555624008178711, local kl: 0.0 global kl: 0.00011776429892051965 valid reconstr loss: 3.0766284465789795
Saving best model with reconstruction loss 3.0766284
it: 500, train recon loss: 1.9641910791397095, local kl: 0.0 global kl: 2.5270983314840123e-05 valid reconstr loss: 1.946200966835022
Saving best model with reconstruction loss 1.946201
it: 600, train recon loss: 0.9956468939781189, local kl: 0.0 global kl: 6.426251388802484e-07 valid reconstr loss: 1.1645094156265259
Saving best model with reconstruction loss 1.1645094
it: 700, train recon loss: 0.9225211143493652, local kl: 0.0 global kl: 1.9581281041070042e-09 valid reconstr loss: 0.9805411696434021
Saving best model with reconstruction loss 0.98054117
it: 800, train recon loss: 1.0649861097335815, local kl: 0.0 global kl: 2.6876556535881946e-09 valid reconstr loss: 1.203018069267273
it: 900, train recon loss: 1.1289268732070923, local kl: 0.0 global kl: 1.9557688801796758e-08 valid reconstr loss: 0.9025200605392456
Saving best model with reconstruction loss 0.90252006
it: 1000, train recon loss: 0.5354803800582886, local kl: 0.0 global kl: 3.228239719987869e-08 valid reconstr loss: 0.8960138559341431
Saving best model with reconstruction loss 0.89601386
it: 1100, train recon loss: 0.10436263680458069, local kl: 0.0 global kl: 9.715450666192282e-09 valid reconstr loss: 0.7204921245574951
Saving best model with reconstruction loss 0.7204921
it: 1200, train recon loss: 0.032730765640735626, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.5442034006118774
Saving best model with reconstruction loss 0.5442034
it: 1300, train recon loss: 166.3771209716797, local kl: 0.0 global kl: 6.677991493120317e-11 valid reconstr loss: 0.43271222710609436
Saving best model with reconstruction loss 0.43271223
it: 1400, train recon loss: -0.06069738790392876, local kl: 0.0 global kl: 1.1678241484958107e-08 valid reconstr loss: 0.023313242942094803
Saving best model with reconstruction loss 0.023313243
it: 1500, train recon loss: 6021.74609375, local kl: 0.0 global kl: 8.427495368223958e-10 valid reconstr loss: 0.012302931398153305
Saving best model with reconstruction loss 0.012302931
it: 1600, train recon loss: -0.12021052092313766, local kl: 0.0 global kl: 3.727206043802056e-11 valid reconstr loss: -0.15050038695335388
Saving best model with reconstruction loss -0.15050039
it: 1700, train recon loss: 0.3696502447128296, local kl: 0.0 global kl: 4.216793580980038e-09 valid reconstr loss: 0.057763345539569855
it: 1800, train recon loss: -0.11094698309898376, local kl: 0.0 global kl: 7.535819079862449e-09 valid reconstr loss: 0.010428882203996181
it: 1900, train recon loss: -0.44218528270721436, local kl: 0.0 global kl: 4.0164160192146525e-11 valid reconstr loss: -0.1410408318042755
it: 2000, train recon loss: 0.11463086307048798, local kl: 0.0 global kl: 2.201135052004588e-09 valid reconstr loss: -0.020742403343319893
it: 2100, train recon loss: 23.811304092407227, local kl: 0.0 global kl: 3.3529601317638935e-10 valid reconstr loss: 0.08851252496242523
it: 2200, train recon loss: -0.44188398122787476, local kl: 0.0 global kl: 1.733888055355237e-08 valid reconstr loss: 0.23422153294086456
it: 2300, train recon loss: 22.462682723999023, local kl: 0.0 global kl: 1.7550322084503023e-08 valid reconstr loss: 261.33587646484375
it: 2400, train recon loss: 3.4621927738189697, local kl: 0.0 global kl: 7.478129226967667e-09 valid reconstr loss: 7.568845748901367
it: 2500, train recon loss: 2.501568555831909, local kl: 0.0 global kl: 5.846017891997235e-09 valid reconstr loss: 2.3280582427978516
it: 2600, train recon loss: 2.3899834156036377, local kl: 0.0 global kl: 1.7279676356452e-08 valid reconstr loss: 2.1975107192993164
it: 2700, train recon loss: 1.5815850496292114, local kl: 0.0 global kl: 4.440892098500626e-13 valid reconstr loss: 1.8736701011657715
it: 2800, train recon loss: 1.2071647644042969, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.2886989116668701
it: 2900, train recon loss: 0.8705083131790161, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.6790781021118164
it: 3000, train recon loss: 0.9654611945152283, local kl: 0.0 global kl: 3.19300141882195e-10 valid reconstr loss: 1.1084351539611816
it: 3100, train recon loss: 0.4370701014995575, local kl: 0.0 global kl: 4.206635040304718e-10 valid reconstr loss: 2.6475324630737305
it: 3200, train recon loss: 5223.74365234375, local kl: 0.0 global kl: 8.316646260553284e-10 valid reconstr loss: 1.3041332960128784
it: 3300, train recon loss: 0.6115128993988037, local kl: 0.0 global kl: 0.0 valid reconstr loss: 1.7753088474273682
it: 3400, train recon loss: 2563.668701171875, local kl: 0.0 global kl: 4.6046388035847485e-08 valid reconstr loss: 0.8234246969223022
it: 3500, train recon loss: 0.6270681023597717, local kl: 0.0 global kl: 2.184086689283049e-08 valid reconstr loss: 0.4399414360523224
it: 3600, train recon loss: 0.6315621733665466, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.381866455078125
it: 3700, train recon loss: 0.21069054305553436, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.5372843146324158
it: 3800, train recon loss: 1.578153133392334, local kl: 0.0 global kl: 4.817285237379565e-09 valid reconstr loss: 0.6867123246192932
it: 3900, train recon loss: -0.0005696983425877988, local kl: 0.0 global kl: 2.9595577544228036e-08 valid reconstr loss: 0.8945991396903992
it: 4000, train recon loss: 0.41231802105903625, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.11880061775445938
it: 4100, train recon loss: -0.07331137359142303, local kl: 0.0 global kl: 2.826649803111536e-09 valid reconstr loss: 0.2691344618797302
it: 4200, train recon loss: 0.18317832052707672, local kl: 0.0 global kl: 1.709370645031072e-09 valid reconstr loss: 0.6437427401542664
it: 4300, train recon loss: -0.11709568649530411, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.5070971250534058
it: 4400, train recon loss: 0.2808212339878082, local kl: 0.0 global kl: 5.4414805994440485e-11 valid reconstr loss: 0.20037008821964264
it: 4500, train recon loss: -0.17523236572742462, local kl: 0.0 global kl: 3.042233132077854e-09 valid reconstr loss: 1.5961618423461914
it: 4600, train recon loss: 0.06915826350450516, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.0019061965867877007
it: 4700, train recon loss: 0.04732425510883331, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.030762238427996635
it: 4800, train recon loss: -0.0863649919629097, local kl: 0.0 global kl: 2.0067169259618822e-08 valid reconstr loss: 0.6922110319137573
it: 4900, train recon loss: 0.36429038643836975, local kl: 0.0 global kl: 8.166540221843377e-10 valid reconstr loss: -0.06021200492978096
it: 5000, train recon loss: 0.021914856508374214, local kl: 0.0 global kl: 4.458239333260394e-11 valid reconstr loss: 0.38920578360557556
it: 5100, train recon loss: 0.10464578866958618, local kl: 0.0 global kl: 3.878147580849145e-09 valid reconstr loss: 25.24039077758789
it: 5200, train recon loss: -0.19864073395729065, local kl: 0.0 global kl: 0.0 valid reconstr loss: 0.4231807589530945
it: 5300, train recon loss: 197.46482849121094, local kl: 0.0 global kl: 1.6123064483508642e-07 valid reconstr loss: 186.01133728027344
it: 5400, train recon loss: 71.67805480957031, local kl: 0.0 global kl: 5.645309286705924e-08 valid reconstr loss: 62.14430618286133
it: 5500, train recon loss: 22.845563888549805, local kl: 0.0 global kl: 1.0078160528337321e-08 valid reconstr loss: 28.244752883911133
it: 5600, train recon loss: 13.469521522521973, local kl: 0.0 global kl: 4.518252438856507e-08 valid reconstr loss: 16.15766143798828
it: 5700, train recon loss: 9.798039436340332, local kl: 0.0 global kl: 2.4937383713563577e-08 valid reconstr loss: 10.604793548583984
it: 5800, train recon loss: 6.414186954498291, local kl: 0.0 global kl: 2.298117252053089e-08 valid reconstr loss: 7.671436786651611
it: 5900, train recon loss: 5.583086013793945, local kl: 0.0 global kl: 2.298430956670927e-08 valid reconstr loss: 6.032748222351074
it: 6000, train recon loss: 4.518730640411377, local kl: 0.0 global kl: 3.448182184229154e-08 valid reconstr loss: 4.992420673370361
it: 6100, train recon loss: 4.133905410766602, local kl: 0.0 global kl: 4.0033100390246545e-08 valid reconstr loss: 4.334167003631592
it: 6200, train recon loss: 3.8905043601989746, local kl: 0.0 global kl: 1.0786399329276719e-08 valid reconstr loss: 3.9259872436523438
it: 6300, train recon loss: 3.73455548286438, local kl: 0.0 global kl: 2.622381423122988e-09 valid reconstr loss: 3.613975763320923
it: 6400, train recon loss: 3.167792320251465, local kl: 0.0 global kl: 8.439970500262461e-09 valid reconstr loss: 3.426414966583252
it: 6500, train recon loss: 3.112016201019287, local kl: 0.0 global kl: 2.0001472478270443e-08 valid reconstr loss: 3.2645182609558105
it: 6600, train recon loss: 2.8932738304138184, local kl: 0.0 global kl: 2.5254575319877404e-08 valid reconstr loss: 3.1316945552825928
it: 6700, train recon loss: 2.9708662033081055, local kl: 0.0 global kl: 9.303454007181244e-09 valid reconstr loss: 3.1035258769989014
it: 6800, train recon loss: 2.903681993484497, local kl: 0.0 global kl: 7.651164146693645e-09 valid reconstr loss: 2.9996001720428467
it: 6900, train recon loss: 3.010990619659424, local kl: 0.0 global kl: 6.733954283077992e-09 valid reconstr loss: 2.9814436435699463
it: 7000, train recon loss: 2.943591356277466, local kl: 0.0 global kl: 1.4140307591503642e-08 valid reconstr loss: 2.9324426651000977
it: 7100, train recon loss: 2.7969295978546143, local kl: 0.0 global kl: 9.030554082301023e-10 valid reconstr loss: 2.927797317504883
it: 7200, train recon loss: 2.9171366691589355, local kl: 0.0 global kl: 2.7782498523976074e-09 valid reconstr loss: 2.9106311798095703
it: 7300, train recon loss: 3.143118381500244, local kl: 0.0 global kl: 3.900491041264331e-10 valid reconstr loss: 2.9305996894836426
it: 7400, train recon loss: 3.147919178009033, local kl: 0.0 global kl: 1.2805589921782712e-09 valid reconstr loss: 2.9047086238861084
it: 7500, train recon loss: 2.962339401245117, local kl: 0.0 global kl: 8.26434032319412e-09 valid reconstr loss: 2.9080851078033447
it: 7600, train recon loss: 3.0528037548065186, local kl: 0.0 global kl: 6.727951529228449e-10 valid reconstr loss: 2.892259359359741
it: 7700, train recon loss: 2.7797598838806152, local kl: 0.0 global kl: 6.480371794737039e-10 valid reconstr loss: 2.9113404750823975
it: 7800, train recon loss: 2.840885639190674, local kl: 0.0 global kl: 4.554204235951431e-10 valid reconstr loss: 2.9040286540985107
it: 7900, train recon loss: 2.7074458599090576, local kl: 0.0 global kl: 5.691280779984709e-10 valid reconstr loss: 2.907912015914917
it: 8000, train recon loss: 2.971446990966797, local kl: 0.0 global kl: 1.6279269043906197e-09 valid reconstr loss: 2.9047720432281494
it: 8100, train recon loss: 2.846205711364746, local kl: 0.0 global kl: 2.4523716390945083e-09 valid reconstr loss: 2.8873393535614014
it: 8200, train recon loss: 2.8623735904693604, local kl: 0.0 global kl: 1.3622714067906827e-09 valid reconstr loss: 2.929685592651367
it: 8300, train recon loss: 2.941462516784668, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.889328718185425
it: 8400, train recon loss: 2.731945753097534, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.9077365398406982
it: 8500, train recon loss: 3.0027177333831787, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.911207675933838
it: 8600, train recon loss: 3.0973546504974365, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.888925790786743
it: 8700, train recon loss: 2.8473474979400635, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.91672945022583
it: 8800, train recon loss: 2.6577515602111816, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.966163158416748
it: 8900, train recon loss: 2.711578607559204, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.959385633468628
it: 9000, train recon loss: 2.644774913787842, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.887988328933716
it: 9100, train recon loss: 3.0662519931793213, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.9218533039093018
it: 9200, train recon loss: 2.859793186187744, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.9028143882751465
it: 9300, train recon loss: 2.8743648529052734, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.8964269161224365
it: 9400, train recon loss: 2.987020492553711, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.895077705383301
it: 9500, train recon loss: 3.0018293857574463, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.924391984939575
it: 9600, train recon loss: 2.871422290802002, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.904257297515869
it: 9700, train recon loss: 2.9390804767608643, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.898618459701538
it: 9800, train recon loss: 2.9917805194854736, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.9141845703125
it: 9900, train recon loss: 2.6633143424987793, local kl: 0.0 global kl: 0.0 valid reconstr loss: 2.921175956726074

In [0]:

    

model_type = 'cnp'
x_y_encoder_net_sizes = [HIDDEN_SIZE]*4
global_latent_net_sizes = None
local_latent_net_sizes = None

model_hparams = tf.contrib.training.HParams(activation=tf.nn.relu,
                                            output_activation=tf.nn.relu,
                                            x_encoder_net_sizes=x_encoder_net_sizes,
                                            x_y_encoder_net_sizes=x_y_encoder_net_sizes,
                                            global_latent_net_sizes=global_latent_net_sizes,
                                            local_latent_net_sizes=local_latent_net_sizes,
                                            decoder_net_sizes=decoder_net_sizes, 
                                            heteroskedastic_net_sizes=heteroskedastic_net_sizes,
                                            att_type=att_type,
                                            att_heads=att_heads,
                                            model_type=model_type,
                                            data_uncertainty=data_uncertainty)
save_path = os.path.join(savedir, 'gnp_' + model_type + '.ckpt')
training_hparams = tf.contrib.training.HParams(lr=0.01,
                                               optimizer=tf.train.RMSPropOptimizer,
                                               num_iterations=10000,
                                               batch_size=10,
                                               num_context=num_context,
                                               num_target=num_target, 
                                               print_every=50,
                                               save_path=save_path,
                                               max_grad_norm=1000.0)

train(data_hparams,
      model_hparams,
      training_hparams)


    

    




it: 0, train mse: 66.05705261230469, local kl: 0.0 global kl: 0.0 valid mse: 84.91262817382812, local kl: 0.0 global kl: 0.0
Saving best model with MSE 84.91263
it: 50, train mse: 65.36094665527344, local kl: 0.0 global kl: 0.0 valid mse: 78.23589324951172, local kl: 0.0 global kl: 0.0
Saving best model with MSE 78.23589
it: 100, train mse: 26.443212509155273, local kl: 0.0 global kl: 0.0 valid mse: 28.662900924682617, local kl: 0.0 global kl: 0.0
Saving best model with MSE 28.6629
it: 150, train mse: 16.119831085205078, local kl: 0.0 global kl: 0.0 valid mse: 19.205217361450195, local kl: 0.0 global kl: 0.0
Saving best model with MSE 19.205217
it: 200, train mse: 14.47561264038086, local kl: 0.0 global kl: 0.0 valid mse: 17.22771453857422, local kl: 0.0 global kl: 0.0
Saving best model with MSE 17.227715
it: 250, train mse: 15.408416748046875, local kl: 0.0 global kl: 0.0 valid mse: 15.189467430114746, local kl: 0.0 global kl: 0.0
Saving best model with MSE 15.189467
it: 300, train mse: 17.508974075317383, local kl: 0.0 global kl: 0.0 valid mse: 19.6955509185791, local kl: 0.0 global kl: 0.0
it: 350, train mse: 17.769800186157227, local kl: 0.0 global kl: 0.0 valid mse: 21.61673927307129, local kl: 0.0 global kl: 0.0
it: 400, train mse: 10.128580093383789, local kl: 0.0 global kl: 0.0 valid mse: 15.104981422424316, local kl: 0.0 global kl: 0.0
Saving best model with MSE 15.104981
it: 450, train mse: 12.804296493530273, local kl: 0.0 global kl: 0.0 valid mse: 13.651142120361328, local kl: 0.0 global kl: 0.0
Saving best model with MSE 13.651142
it: 500, train mse: 7.45737886428833, local kl: 0.0 global kl: 0.0 valid mse: 12.355969429016113, local kl: 0.0 global kl: 0.0
Saving best model with MSE 12.355969
it: 550, train mse: 14.835765838623047, local kl: 0.0 global kl: 0.0 valid mse: 18.019624710083008, local kl: 0.0 global kl: 0.0
it: 600, train mse: 10.593198776245117, local kl: 0.0 global kl: 0.0 valid mse: 12.482098579406738, local kl: 0.0 global kl: 0.0
it: 650, train mse: 10.514266014099121, local kl: 0.0 global kl: 0.0 valid mse: 12.650646209716797, local kl: 0.0 global kl: 0.0
it: 700, train mse: 8.624704360961914, local kl: 0.0 global kl: 0.0 valid mse: 11.19351577758789, local kl: 0.0 global kl: 0.0
Saving best model with MSE 11.193516
it: 750, train mse: 17.021738052368164, local kl: 0.0 global kl: 0.0 valid mse: 16.983949661254883, local kl: 0.0 global kl: 0.0
it: 800, train mse: 6.449836254119873, local kl: 0.0 global kl: 0.0 valid mse: 10.915543556213379, local kl: 0.0 global kl: 0.0
Saving best model with MSE 10.915544
it: 850, train mse: 12.018802642822266, local kl: 0.0 global kl: 0.0 valid mse: 10.594204902648926, local kl: 0.0 global kl: 0.0
Saving best model with MSE 10.594205
it: 900, train mse: 9.907962799072266, local kl: 0.0 global kl: 0.0 valid mse: 21.90072250366211, local kl: 0.0 global kl: 0.0
it: 950, train mse: 8.6156587600708, local kl: 0.0 global kl: 0.0 valid mse: 13.209843635559082, local kl: 0.0 global kl: 0.0
it: 1000, train mse: 8.021354675292969, local kl: 0.0 global kl: 0.0 valid mse: 12.763694763183594, local kl: 0.0 global kl: 0.0
it: 1050, train mse: 7.635591983795166, local kl: 0.0 global kl: 0.0 valid mse: 14.34134578704834, local kl: 0.0 global kl: 0.0
it: 1100, train mse: 7.75965690612793, local kl: 0.0 global kl: 0.0 valid mse: 17.00267791748047, local kl: 0.0 global kl: 0.0
it: 1150, train mse: 8.402264595031738, local kl: 0.0 global kl: 0.0 valid mse: 6.810908794403076, local kl: 0.0 global kl: 0.0
Saving best model with MSE 6.810909
it: 1200, train mse: 9.235950469970703, local kl: 0.0 global kl: 0.0 valid mse: 11.007472038269043, local kl: 0.0 global kl: 0.0
it: 1250, train mse: 6.8675360679626465, local kl: 0.0 global kl: 0.0 valid mse: 13.332806587219238, local kl: 0.0 global kl: 0.0
it: 1300, train mse: 8.83800983428955, local kl: 0.0 global kl: 0.0 valid mse: 8.181134223937988, local kl: 0.0 global kl: 0.0
it: 1350, train mse: 5.059090614318848, local kl: 0.0 global kl: 0.0 valid mse: 7.5062103271484375, local kl: 0.0 global kl: 0.0
it: 1400, train mse: 8.695988655090332, local kl: 0.0 global kl: 0.0 valid mse: 8.836488723754883, local kl: 0.0 global kl: 0.0
it: 1450, train mse: 10.769319534301758, local kl: 0.0 global kl: 0.0 valid mse: 10.88158130645752, local kl: 0.0 global kl: 0.0
it: 1500, train mse: 5.2554850578308105, local kl: 0.0 global kl: 0.0 valid mse: 9.523558616638184, local kl: 0.0 global kl: 0.0
it: 1550, train mse: 8.347359657287598, local kl: 0.0 global kl: 0.0 valid mse: 8.056888580322266, local kl: 0.0 global kl: 0.0
it: 1600, train mse: 3.9624195098876953, local kl: 0.0 global kl: 0.0 valid mse: 7.780689239501953, local kl: 0.0 global kl: 0.0
it: 1650, train mse: 5.071132183074951, local kl: 0.0 global kl: 0.0 valid mse: 12.765825271606445, local kl: 0.0 global kl: 0.0
it: 1700, train mse: 10.678670883178711, local kl: 0.0 global kl: 0.0 valid mse: 6.819207191467285, local kl: 0.0 global kl: 0.0
it: 1750, train mse: 8.499160766601562, local kl: 0.0 global kl: 0.0 valid mse: 8.202756881713867, local kl: 0.0 global kl: 0.0
it: 1800, train mse: 14.32585334777832, local kl: 0.0 global kl: 0.0 valid mse: 12.907402038574219, local kl: 0.0 global kl: 0.0
it: 1850, train mse: 6.925199031829834, local kl: 0.0 global kl: 0.0 valid mse: 17.481109619140625, local kl: 0.0 global kl: 0.0
it: 1900, train mse: 8.09041976928711, local kl: 0.0 global kl: 0.0 valid mse: 7.704461097717285, local kl: 0.0 global kl: 0.0
it: 1950, train mse: 6.341521739959717, local kl: 0.0 global kl: 0.0 valid mse: 10.365133285522461, local kl: 0.0 global kl: 0.0
it: 2000, train mse: 4.221902847290039, local kl: 0.0 global kl: 0.0 valid mse: 11.237725257873535, local kl: 0.0 global kl: 0.0
it: 2050, train mse: 6.100094795227051, local kl: 0.0 global kl: 0.0 valid mse: 6.6067705154418945, local kl: 0.0 global kl: 0.0
Saving best model with MSE 6.6067705
it: 2100, train mse: 5.2984795570373535, local kl: 0.0 global kl: 0.0 valid mse: 8.96905517578125, local kl: 0.0 global kl: 0.0
it: 2150, train mse: 16.065439224243164, local kl: 0.0 global kl: 0.0 valid mse: 6.344666957855225, local kl: 0.0 global kl: 0.0
Saving best model with MSE 6.344667
it: 2200, train mse: 6.621694564819336, local kl: 0.0 global kl: 0.0 valid mse: 8.849959373474121, local kl: 0.0 global kl: 0.0
it: 2250, train mse: 7.690698623657227, local kl: 0.0 global kl: 0.0 valid mse: 15.568334579467773, local kl: 0.0 global kl: 0.0
it: 2300, train mse: 3.872899293899536, local kl: 0.0 global kl: 0.0 valid mse: 7.443742752075195, local kl: 0.0 global kl: 0.0
it: 2350, train mse: 4.448376655578613, local kl: 0.0 global kl: 0.0 valid mse: 12.500809669494629, local kl: 0.0 global kl: 0.0
it: 2400, train mse: 6.562459945678711, local kl: 0.0 global kl: 0.0 valid mse: 12.618840217590332, local kl: 0.0 global kl: 0.0
it: 2450, train mse: 5.121094703674316, local kl: 0.0 global kl: 0.0 valid mse: 10.753654479980469, local kl: 0.0 global kl: 0.0
it: 2500, train mse: 9.483473777770996, local kl: 0.0 global kl: 0.0 valid mse: 7.137670993804932, local kl: 0.0 global kl: 0.0
it: 2550, train mse: 8.184005737304688, local kl: 0.0 global kl: 0.0 valid mse: 11.371467590332031, local kl: 0.0 global kl: 0.0
it: 2600, train mse: 4.751481056213379, local kl: 0.0 global kl: 0.0 valid mse: 6.807375907897949, local kl: 0.0 global kl: 0.0
it: 2650, train mse: 6.784178733825684, local kl: 0.0 global kl: 0.0 valid mse: 9.12680435180664, local kl: 0.0 global kl: 0.0
it: 2700, train mse: 7.962423324584961, local kl: 0.0 global kl: 0.0 valid mse: 8.132933616638184, local kl: 0.0 global kl: 0.0
it: 2750, train mse: 4.453100204467773, local kl: 0.0 global kl: 0.0 valid mse: 5.682377815246582, local kl: 0.0 global kl: 0.0
Saving best model with MSE 5.682378
it: 2800, train mse: 5.877711296081543, local kl: 0.0 global kl: 0.0 valid mse: 13.420318603515625, local kl: 0.0 global kl: 0.0
it: 2850, train mse: 7.095983028411865, local kl: 0.0 global kl: 0.0 valid mse: 10.662031173706055, local kl: 0.0 global kl: 0.0
it: 2900, train mse: 7.062972545623779, local kl: 0.0 global kl: 0.0 valid mse: 8.28528881072998, local kl: 0.0 global kl: 0.0
it: 2950, train mse: 5.280788898468018, local kl: 0.0 global kl: 0.0 valid mse: 9.568038940429688, local kl: 0.0 global kl: 0.0
it: 3000, train mse: 11.717610359191895, local kl: 0.0 global kl: 0.0 valid mse: 9.247783660888672, local kl: 0.0 global kl: 0.0
it: 3050, train mse: 7.611526012420654, local kl: 0.0 global kl: 0.0 valid mse: 8.77609920501709, local kl: 0.0 global kl: 0.0
it: 3100, train mse: 7.473079681396484, local kl: 0.0 global kl: 0.0 valid mse: 8.066476821899414, local kl: 0.0 global kl: 0.0
it: 3150, train mse: 5.418025493621826, local kl: 0.0 global kl: 0.0 valid mse: 10.203829765319824, local kl: 0.0 global kl: 0.0
it: 3200, train mse: 6.09105920791626, local kl: 0.0 global kl: 0.0 valid mse: 8.80147647857666, local kl: 0.0 global kl: 0.0
it: 3250, train mse: 5.963371753692627, local kl: 0.0 global kl: 0.0 valid mse: 11.1729154586792, local kl: 0.0 global kl: 0.0
it: 3300, train mse: 4.167073726654053, local kl: 0.0 global kl: 0.0 valid mse: 6.948697566986084, local kl: 0.0 global kl: 0.0
it: 3350, train mse: 3.6569266319274902, local kl: 0.0 global kl: 0.0 valid mse: 7.593774318695068, local kl: 0.0 global kl: 0.0
it: 3400, train mse: 6.3019914627075195, local kl: 0.0 global kl: 0.0 valid mse: 10.38212776184082, local kl: 0.0 global kl: 0.0
it: 3450, train mse: 4.435804843902588, local kl: 0.0 global kl: 0.0 valid mse: 10.263124465942383, local kl: 0.0 global kl: 0.0
it: 3500, train mse: 9.841634750366211, local kl: 0.0 global kl: 0.0 valid mse: 15.20864200592041, local kl: 0.0 global kl: 0.0
it: 3550, train mse: 7.070883750915527, local kl: 0.0 global kl: 0.0 valid mse: 7.4892258644104, local kl: 0.0 global kl: 0.0
it: 3600, train mse: 4.469335556030273, local kl: 0.0 global kl: 0.0 valid mse: 12.393798828125, local kl: 0.0 global kl: 0.0
it: 3650, train mse: 10.646888732910156, local kl: 0.0 global kl: 0.0 valid mse: 6.9256110191345215, local kl: 0.0 global kl: 0.0
it: 3700, train mse: 4.706447601318359, local kl: 0.0 global kl: 0.0 valid mse: 6.638940811157227, local kl: 0.0 global kl: 0.0
it: 3750, train mse: 13.738762855529785, local kl: 0.0 global kl: 0.0 valid mse: 11.308430671691895, local kl: 0.0 global kl: 0.0
it: 3800, train mse: 9.059133529663086, local kl: 0.0 global kl: 0.0 valid mse: 10.65356159210205, local kl: 0.0 global kl: 0.0
it: 3850, train mse: 2.853177070617676, local kl: 0.0 global kl: 0.0 valid mse: 6.859042167663574, local kl: 0.0 global kl: 0.0
it: 3900, train mse: 6.039935111999512, local kl: 0.0 global kl: 0.0 valid mse: 8.484132766723633, local kl: 0.0 global kl: 0.0
it: 3950, train mse: 8.857287406921387, local kl: 0.0 global kl: 0.0 valid mse: 9.446547508239746, local kl: 0.0 global kl: 0.0
it: 4000, train mse: 6.8550615310668945, local kl: 0.0 global kl: 0.0 valid mse: 9.192556381225586, local kl: 0.0 global kl: 0.0
it: 4050, train mse: 5.614525318145752, local kl: 0.0 global kl: 0.0 valid mse: 8.8878173828125, local kl: 0.0 global kl: 0.0
it: 4100, train mse: 10.093817710876465, local kl: 0.0 global kl: 0.0 valid mse: 5.756629467010498, local kl: 0.0 global kl: 0.0
it: 4150, train mse: 4.7568583488464355, local kl: 0.0 global kl: 0.0 valid mse: 6.217566967010498, local kl: 0.0 global kl: 0.0
it: 4200, train mse: 5.016478061676025, local kl: 0.0 global kl: 0.0 valid mse: 8.869423866271973, local kl: 0.0 global kl: 0.0
it: 4250, train mse: 3.8003528118133545, local kl: 0.0 global kl: 0.0 valid mse: 6.133528709411621, local kl: 0.0 global kl: 0.0
it: 4300, train mse: 4.452188491821289, local kl: 0.0 global kl: 0.0 valid mse: 8.486845016479492, local kl: 0.0 global kl: 0.0
it: 4350, train mse: 6.077689170837402, local kl: 0.0 global kl: 0.0 valid mse: 8.379814147949219, local kl: 0.0 global kl: 0.0
it: 4400, train mse: 8.86538314819336, local kl: 0.0 global kl: 0.0 valid mse: 9.517661094665527, local kl: 0.0 global kl: 0.0
it: 4450, train mse: 16.850255966186523, local kl: 0.0 global kl: 0.0 valid mse: 13.927631378173828, local kl: 0.0 global kl: 0.0
it: 4500, train mse: 6.211893558502197, local kl: 0.0 global kl: 0.0 valid mse: 8.548407554626465, local kl: 0.0 global kl: 0.0
it: 4550, train mse: 6.294714450836182, local kl: 0.0 global kl: 0.0 valid mse: 6.277410507202148, local kl: 0.0 global kl: 0.0
it: 4600, train mse: 2.657541036605835, local kl: 0.0 global kl: 0.0 valid mse: 5.3463335037231445, local kl: 0.0 global kl: 0.0
Saving best model with MSE 5.3463335
it: 4650, train mse: 4.077348709106445, local kl: 0.0 global kl: 0.0 valid mse: 7.436420440673828, local kl: 0.0 global kl: 0.0
it: 4700, train mse: 6.188862323760986, local kl: 0.0 global kl: 0.0 valid mse: 9.27371883392334, local kl: 0.0 global kl: 0.0
it: 4750, train mse: 7.037395477294922, local kl: 0.0 global kl: 0.0 valid mse: 9.923429489135742, local kl: 0.0 global kl: 0.0
it: 4800, train mse: 11.100876808166504, local kl: 0.0 global kl: 0.0 valid mse: 10.00271987915039, local kl: 0.0 global kl: 0.0
it: 4850, train mse: 5.139604568481445, local kl: 0.0 global kl: 0.0 valid mse: 8.183032989501953, local kl: 0.0 global kl: 0.0
it: 4900, train mse: 5.712069988250732, local kl: 0.0 global kl: 0.0 valid mse: 5.6780548095703125, local kl: 0.0 global kl: 0.0
it: 4950, train mse: 5.638814926147461, local kl: 0.0 global kl: 0.0 valid mse: 8.428606986999512, local kl: 0.0 global kl: 0.0
it: 5000, train mse: 5.230318546295166, local kl: 0.0 global kl: 0.0 valid mse: 7.840363502502441, local kl: 0.0 global kl: 0.0
it: 5050, train mse: 8.328577041625977, local kl: 0.0 global kl: 0.0 valid mse: 19.42652702331543, local kl: 0.0 global kl: 0.0
it: 5100, train mse: 5.747538089752197, local kl: 0.0 global kl: 0.0 valid mse: 5.705214977264404, local kl: 0.0 global kl: 0.0
it: 5150, train mse: 5.638016223907471, local kl: 0.0 global kl: 0.0 valid mse: 7.469056129455566, local kl: 0.0 global kl: 0.0
it: 5200, train mse: 4.500269412994385, local kl: 0.0 global kl: 0.0 valid mse: 10.899223327636719, local kl: 0.0 global kl: 0.0
it: 5250, train mse: 8.304844856262207, local kl: 0.0 global kl: 0.0 valid mse: 8.736930847167969, local kl: 0.0 global kl: 0.0
it: 5300, train mse: 2.914673089981079, local kl: 0.0 global kl: 0.0 valid mse: 8.554244995117188, local kl: 0.0 global kl: 0.0
it: 5350, train mse: 4.273291110992432, local kl: 0.0 global kl: 0.0 valid mse: 7.6545820236206055, local kl: 0.0 global kl: 0.0
it: 5400, train mse: 4.834151744842529, local kl: 0.0 global kl: 0.0 valid mse: 7.461828708648682, local kl: 0.0 global kl: 0.0
it: 5450, train mse: 4.509783744812012, local kl: 0.0 global kl: 0.0 valid mse: 6.526172637939453, local kl: 0.0 global kl: 0.0
it: 5500, train mse: 1.17620849609375, local kl: 0.0 global kl: 0.0 valid mse: 5.592796325683594, local kl: 0.0 global kl: 0.0
it: 5550, train mse: 5.51279354095459, local kl: 0.0 global kl: 0.0 valid mse: 6.207424163818359, local kl: 0.0 global kl: 0.0
it: 5600, train mse: 5.563343048095703, local kl: 0.0 global kl: 0.0 valid mse: 9.516047477722168, local kl: 0.0 global kl: 0.0
it: 5650, train mse: 4.281150817871094, local kl: 0.0 global kl: 0.0 valid mse: 7.429060935974121, local kl: 0.0 global kl: 0.0
it: 5700, train mse: 4.822322845458984, local kl: 0.0 global kl: 0.0 valid mse: 5.168940544128418, local kl: 0.0 global kl: 0.0
Saving best model with MSE 5.1689405
it: 5750, train mse: 8.368767738342285, local kl: 0.0 global kl: 0.0 valid mse: 8.77314567565918, local kl: 0.0 global kl: 0.0
it: 5800, train mse: 2.7109594345092773, local kl: 0.0 global kl: 0.0 valid mse: 7.703765392303467, local kl: 0.0 global kl: 0.0
it: 5850, train mse: 3.839125394821167, local kl: 0.0 global kl: 0.0 valid mse: 7.144743919372559, local kl: 0.0 global kl: 0.0
it: 5900, train mse: 8.911205291748047, local kl: 0.0 global kl: 0.0 valid mse: 4.8888068199157715, local kl: 0.0 global kl: 0.0
Saving best model with MSE 4.888807
it: 5950, train mse: 4.644564151763916, local kl: 0.0 global kl: 0.0 valid mse: 5.957622051239014, local kl: 0.0 global kl: 0.0
it: 6000, train mse: 5.133630275726318, local kl: 0.0 global kl: 0.0 valid mse: 7.986464023590088, local kl: 0.0 global kl: 0.0
it: 6050, train mse: 7.044681072235107, local kl: 0.0 global kl: 0.0 valid mse: 5.603495121002197, local kl: 0.0 global kl: 0.0
it: 6100, train mse: 2.600158452987671, local kl: 0.0 global kl: 0.0 valid mse: 5.6196818351745605, local kl: 0.0 global kl: 0.0
it: 6150, train mse: 8.682184219360352, local kl: 0.0 global kl: 0.0 valid mse: 8.219542503356934, local kl: 0.0 global kl: 0.0
it: 6200, train mse: 3.0910840034484863, local kl: 0.0 global kl: 0.0 valid mse: 6.591758728027344, local kl: 0.0 global kl: 0.0
it: 6250, train mse: 3.802211046218872, local kl: 0.0 global kl: 0.0 valid mse: 5.996861934661865, local kl: 0.0 global kl: 0.0
it: 6300, train mse: 5.6643242835998535, local kl: 0.0 global kl: 0.0 valid mse: 7.120818138122559, local kl: 0.0 global kl: 0.0
it: 6350, train mse: 4.1019158363342285, local kl: 0.0 global kl: 0.0 valid mse: 7.229722499847412, local kl: 0.0 global kl: 0.0
it: 6400, train mse: 7.545364856719971, local kl: 0.0 global kl: 0.0 valid mse: 7.970773220062256, local kl: 0.0 global kl: 0.0
it: 6450, train mse: 5.5231781005859375, local kl: 0.0 global kl: 0.0 valid mse: 6.338512897491455, local kl: 0.0 global kl: 0.0
it: 6500, train mse: 8.598823547363281, local kl: 0.0 global kl: 0.0 valid mse: 7.869621753692627, local kl: 0.0 global kl: 0.0
it: 6550, train mse: 11.145733833312988, local kl: 0.0 global kl: 0.0 valid mse: 7.388317108154297, local kl: 0.0 global kl: 0.0
it: 6600, train mse: 6.534750461578369, local kl: 0.0 global kl: 0.0 valid mse: 7.284392833709717, local kl: 0.0 global kl: 0.0
it: 6650, train mse: 6.301437854766846, local kl: 0.0 global kl: 0.0 valid mse: 5.9103498458862305, local kl: 0.0 global kl: 0.0
it: 6700, train mse: 8.867037773132324, local kl: 0.0 global kl: 0.0 valid mse: 8.993256568908691, local kl: 0.0 global kl: 0.0
it: 6750, train mse: 4.926150798797607, local kl: 0.0 global kl: 0.0 valid mse: 6.496001720428467, local kl: 0.0 global kl: 0.0
it: 6800, train mse: 5.570132255554199, local kl: 0.0 global kl: 0.0 valid mse: 6.720869541168213, local kl: 0.0 global kl: 0.0
it: 6850, train mse: 6.935928821563721, local kl: 0.0 global kl: 0.0 valid mse: 6.013933181762695, local kl: 0.0 global kl: 0.0
it: 6900, train mse: 7.504579067230225, local kl: 0.0 global kl: 0.0 valid mse: 7.31773042678833, local kl: 0.0 global kl: 0.0
it: 6950, train mse: 11.206644058227539, local kl: 0.0 global kl: 0.0 valid mse: 8.619681358337402, local kl: 0.0 global kl: 0.0
it: 7000, train mse: 5.93508243560791, local kl: 0.0 global kl: 0.0 valid mse: 7.344940662384033, local kl: 0.0 global kl: 0.0
it: 7050, train mse: 4.431743621826172, local kl: 0.0 global kl: 0.0 valid mse: 6.980828285217285, local kl: 0.0 global kl: 0.0
it: 7100, train mse: 12.748468399047852, local kl: 0.0 global kl: 0.0 valid mse: 7.040874004364014, local kl: 0.0 global kl: 0.0
it: 7150, train mse: 15.17672348022461, local kl: 0.0 global kl: 0.0 valid mse: 6.0524797439575195, local kl: 0.0 global kl: 0.0
it: 7200, train mse: 9.212101936340332, local kl: 0.0 global kl: 0.0 valid mse: 8.515445709228516, local kl: 0.0 global kl: 0.0
it: 7250, train mse: 4.3776726722717285, local kl: 0.0 global kl: 0.0 valid mse: 7.5849609375, local kl: 0.0 global kl: 0.0
it: 7300, train mse: 8.42762565612793, local kl: 0.0 global kl: 0.0 valid mse: 6.555792808532715, local kl: 0.0 global kl: 0.0
it: 7350, train mse: 7.231533050537109, local kl: 0.0 global kl: 0.0 valid mse: 5.976547718048096, local kl: 0.0 global kl: 0.0
it: 7400, train mse: 8.192923545837402, local kl: 0.0 global kl: 0.0 valid mse: 6.502047538757324, local kl: 0.0 global kl: 0.0
it: 7450, train mse: 5.598387718200684, local kl: 0.0 global kl: 0.0 valid mse: 5.519688606262207, local kl: 0.0 global kl: 0.0
it: 7500, train mse: 8.168920516967773, local kl: 0.0 global kl: 0.0 valid mse: 15.593903541564941, local kl: 0.0 global kl: 0.0
it: 7550, train mse: 7.351479530334473, local kl: 0.0 global kl: 0.0 valid mse: 7.531446933746338, local kl: 0.0 global kl: 0.0
it: 7600, train mse: 4.687422275543213, local kl: 0.0 global kl: 0.0 valid mse: 12.130295753479004, local kl: 0.0 global kl: 0.0
it: 7650, train mse: 6.395683288574219, local kl: 0.0 global kl: 0.0 valid mse: 6.213014125823975, local kl: 0.0 global kl: 0.0
it: 7700, train mse: 8.766763687133789, local kl: 0.0 global kl: 0.0 valid mse: 7.024875164031982, local kl: 0.0 global kl: 0.0
it: 7750, train mse: 5.83670711517334, local kl: 0.0 global kl: 0.0 valid mse: 11.541559219360352, local kl: 0.0 global kl: 0.0
it: 7800, train mse: 6.727958679199219, local kl: 0.0 global kl: 0.0 valid mse: 4.693368911743164, local kl: 0.0 global kl: 0.0
Saving best model with MSE 4.693369
it: 7850, train mse: 3.8043148517608643, local kl: 0.0 global kl: 0.0 valid mse: 6.003916263580322, local kl: 0.0 global kl: 0.0
it: 7900, train mse: 3.848611831665039, local kl: 0.0 global kl: 0.0 valid mse: 6.415050983428955, local kl: 0.0 global kl: 0.0
it: 7950, train mse: 5.1251726150512695, local kl: 0.0 global kl: 0.0 valid mse: 6.48779821395874, local kl: 0.0 global kl: 0.0
it: 8000, train mse: 4.398504257202148, local kl: 0.0 global kl: 0.0 valid mse: 6.367981910705566, local kl: 0.0 global kl: 0.0
it: 8050, train mse: 6.673210144042969, local kl: 0.0 global kl: 0.0 valid mse: 5.452404499053955, local kl: 0.0 global kl: 0.0
it: 8100, train mse: 7.504153251647949, local kl: 0.0 global kl: 0.0 valid mse: 5.227799415588379, local kl: 0.0 global kl: 0.0
it: 8150, train mse: 5.018385410308838, local kl: 0.0 global kl: 0.0 valid mse: 9.269495010375977, local kl: 0.0 global kl: 0.0
it: 8200, train mse: 10.261159896850586, local kl: 0.0 global kl: 0.0 valid mse: 7.880271911621094, local kl: 0.0 global kl: 0.0
it: 8250, train mse: 7.370147705078125, local kl: 0.0 global kl: 0.0 valid mse: 7.462297439575195, local kl: 0.0 global kl: 0.0
it: 8300, train mse: 5.905923843383789, local kl: 0.0 global kl: 0.0 valid mse: 6.35527229309082, local kl: 0.0 global kl: 0.0
it: 8350, train mse: 7.2466959953308105, local kl: 0.0 global kl: 0.0 valid mse: 6.281361103057861, local kl: 0.0 global kl: 0.0
it: 8400, train mse: 7.050768852233887, local kl: 0.0 global kl: 0.0 valid mse: 8.928950309753418, local kl: 0.0 global kl: 0.0
it: 8450, train mse: 7.618067264556885, local kl: 0.0 global kl: 0.0 valid mse: 5.885317802429199, local kl: 0.0 global kl: 0.0
it: 8500, train mse: 10.129331588745117, local kl: 0.0 global kl: 0.0 valid mse: 6.705228328704834, local kl: 0.0 global kl: 0.0
it: 8550, train mse: 4.612322807312012, local kl: 0.0 global kl: 0.0 valid mse: 6.639267921447754, local kl: 0.0 global kl: 0.0
it: 8600, train mse: 6.2995285987854, local kl: 0.0 global kl: 0.0 valid mse: 6.010883808135986, local kl: 0.0 global kl: 0.0
it: 8650, train mse: 4.843682289123535, local kl: 0.0 global kl: 0.0 valid mse: 5.587864875793457, local kl: 0.0 global kl: 0.0
it: 8700, train mse: 18.112197875976562, local kl: 0.0 global kl: 0.0 valid mse: 6.70115327835083, local kl: 0.0 global kl: 0.0
it: 8750, train mse: 9.142450332641602, local kl: 0.0 global kl: 0.0 valid mse: 6.175296306610107, local kl: 0.0 global kl: 0.0
it: 8800, train mse: 7.6715216636657715, local kl: 0.0 global kl: 0.0 valid mse: 7.495497703552246, local kl: 0.0 global kl: 0.0
it: 8850, train mse: 3.8423688411712646, local kl: 0.0 global kl: 0.0 valid mse: 6.344202518463135, local kl: 0.0 global kl: 0.0
it: 8900, train mse: 8.98909854888916, local kl: 0.0 global kl: 0.0 valid mse: 6.959056377410889, local kl: 0.0 global kl: 0.0
it: 8950, train mse: 8.005345344543457, local kl: 0.0 global kl: 0.0 valid mse: 5.80885648727417, local kl: 0.0 global kl: 0.0
it: 9000, train mse: 4.39376163482666, local kl: 0.0 global kl: 0.0 valid mse: 7.23040771484375, local kl: 0.0 global kl: 0.0
it: 9050, train mse: 4.421983242034912, local kl: 0.0 global kl: 0.0 valid mse: 4.801156044006348, local kl: 0.0 global kl: 0.0
it: 9100, train mse: 5.656516075134277, local kl: 0.0 global kl: 0.0 valid mse: 7.484582901000977, local kl: 0.0 global kl: 0.0
it: 9150, train mse: 7.371536731719971, local kl: 0.0 global kl: 0.0 valid mse: 4.972528457641602, local kl: 0.0 global kl: 0.0
it: 9200, train mse: 7.817051410675049, local kl: 0.0 global kl: 0.0 valid mse: 6.90757417678833, local kl: 0.0 global kl: 0.0
it: 9250, train mse: 5.171108722686768, local kl: 0.0 global kl: 0.0 valid mse: 9.470877647399902, local kl: 0.0 global kl: 0.0
it: 9300, train mse: 7.45332670211792, local kl: 0.0 global kl: 0.0 valid mse: 10.106136322021484, local kl: 0.0 global kl: 0.0
it: 9350, train mse: 5.193520545959473, local kl: 0.0 global kl: 0.0 valid mse: 6.4881367683410645, local kl: 0.0 global kl: 0.0
it: 9400, train mse: 4.577541828155518, local kl: 0.0 global kl: 0.0 valid mse: 7.396413326263428, local kl: 0.0 global kl: 0.0
it: 9450, train mse: 5.638520240783691, local kl: 0.0 global kl: 0.0 valid mse: 4.871909141540527, local kl: 0.0 global kl: 0.0
it: 9500, train mse: 12.87208366394043, local kl: 0.0 global kl: 0.0 valid mse: 16.074718475341797, local kl: 0.0 global kl: 0.0
it: 9550, train mse: 12.80510139465332, local kl: 0.0 global kl: 0.0 valid mse: 20.35961151123047, local kl: 0.0 global kl: 0.0
it: 9600, train mse: 13.101369857788086, local kl: 0.0 global kl: 0.0 valid mse: 15.559097290039062, local kl: 0.0 global kl: 0.0
it: 9650, train mse: 12.020488739013672, local kl: 0.0 global kl: 0.0 valid mse: 14.863801956176758, local kl: 0.0 global kl: 0.0
it: 9700, train mse: 18.83301544189453, local kl: 0.0 global kl: 0.0 valid mse: 14.714988708496094, local kl: 0.0 global kl: 0.0
it: 9750, train mse: 16.534849166870117, local kl: 0.0 global kl: 0.0 valid mse: 14.749890327453613, local kl: 0.0 global kl: 0.0
it: 9800, train mse: 15.274186134338379, local kl: 0.0 global kl: 0.0 valid mse: 16.404050827026367, local kl: 0.0 global kl: 0.0
it: 9850, train mse: 17.628807067871094, local kl: 0.0 global kl: 0.0 valid mse: 15.966422080993652, local kl: 0.0 global kl: 0.0
it: 9900, train mse: 21.721044540405273, local kl: 0.0 global kl: 0.0 valid mse: 17.288433074951172, local kl: 0.0 global kl: 0.0
it: 9950, train mse: 7.468124866485596, local kl: 0.0 global kl: 0.0 valid mse: 16.686586380004883, local kl: 0.0 global kl: 0.0

In [0]:

    

model_type = 'cnp'
x_y_encoder_net_sizes = [HIDDEN_SIZE]*4
global_latent_net_sizes = None
local_latent_net_sizes = None

model_hparams = tf.contrib.training.HParams(activation=tf.nn.relu,
                                            output_activation=tf.nn.relu,
                                            x_encoder_net_sizes=x_encoder_net_sizes,
                                            x_y_encoder_net_sizes=x_y_encoder_net_sizes,
                                            global_latent_net_sizes=global_latent_net_sizes,
                                            local_latent_net_sizes=local_latent_net_sizes,
                                            decoder_net_sizes=decoder_net_sizes, 
                                            heteroskedastic_net_sizes=heteroskedastic_net_sizes,
                                            att_type=att_type,
                                            att_heads=att_heads,
                                            model_type=model_type,
                                            data_uncertainty=data_uncertainty)
save_path = os.path.join(savedir, 'gnp_nll_' + model_type + '.ckpt')
training_hparams = tf.contrib.training.HParams(lr=0.01,
                                               optimizer=tf.train.RMSPropOptimizer,
                                               num_iterations=10000,
                                               batch_size=10,
                                               num_context=num_context,
                                               num_target=num_target, 
                                               print_every=50,
                                               save_path=save_path,
                                               max_grad_norm=1000.0,
                                               is_nll=True)

train(data_hparams,
      model_hparams,
      training_hparams)


    

In [0]:

    

model_type = 'np'
x_y_encoder_net_sizes = [HIDDEN_SIZE]*2
global_latent_net_sizes = [HIDDEN_SIZE]*2
local_latent_net_sizes = None

model_hparams = tf.contrib.training.HParams(activation=tf.nn.relu,
                                            output_activation=tf.nn.relu,
                                            x_encoder_net_sizes=x_encoder_net_sizes,
                                            x_y_encoder_net_sizes=x_y_encoder_net_sizes,
                                            global_latent_net_sizes=global_latent_net_sizes,
                                            local_latent_net_sizes=local_latent_net_sizes,
                                            decoder_net_sizes=decoder_net_sizes, 
                                            heteroskedastic_net_sizes=heteroskedastic_net_sizes,
                                            att_type=att_type,
                                            att_heads=att_heads,
                                            model_type=model_type,
                                            data_uncertainty=data_uncertainty)
save_path = os.path.join(savedir, 'gnp_' + model_type + '.ckpt')
training_hparams = tf.contrib.training.HParams(lr=0.01,
                                               optimizer=tf.train.RMSPropOptimizer,
                                               num_iterations=10000,
                                               batch_size=10,
                                               num_context=num_context,
                                               num_target=num_target, 
                                               print_every=50,
                                               save_path=save_path,
                                               max_grad_norm=1000.0)

train(data_hparams,
      model_hparams,
      training_hparams)


    

    




it: 0, train mse: 65.76097869873047, local kl: 0.0 global kl: 0.054639000445604324 valid mse: 83.96866607666016, local kl: 0.0 global kl: 0.07167519629001617
Saving best model with MSE 83.968666
it: 50, train mse: 57.198280334472656, local kl: 0.0 global kl: 0.017055261880159378 valid mse: 67.92539978027344, local kl: 0.0 global kl: 0.03785083070397377
Saving best model with MSE 67.9254
it: 100, train mse: 30.777847290039062, local kl: 0.0 global kl: 0.03610922396183014 valid mse: 33.16203689575195, local kl: 0.0 global kl: 0.11787480115890503
Saving best model with MSE 33.162037
it: 150, train mse: 24.821523666381836, local kl: 0.0 global kl: 0.04832617938518524 valid mse: 24.692289352416992, local kl: 0.0 global kl: 0.058483920991420746
Saving best model with MSE 24.69229
it: 200, train mse: 22.564250946044922, local kl: 0.0 global kl: 0.017527341842651367 valid mse: 20.14684295654297, local kl: 0.0 global kl: 0.09147223830223083
Saving best model with MSE 20.146843
it: 250, train mse: 13.489233016967773, local kl: 0.0 global kl: 0.1045757308602333 valid mse: 15.312090873718262, local kl: 0.0 global kl: 0.21254530549049377
Saving best model with MSE 15.312091
it: 300, train mse: 13.116039276123047, local kl: 0.0 global kl: 0.10509645938873291 valid mse: 20.458494186401367, local kl: 0.0 global kl: 0.12832245230674744
it: 350, train mse: 15.904216766357422, local kl: 0.0 global kl: 0.15157441794872284 valid mse: 17.16631317138672, local kl: 0.0 global kl: 0.17886143922805786
it: 400, train mse: 11.340956687927246, local kl: 0.0 global kl: 0.1478172242641449 valid mse: 12.591387748718262, local kl: 0.0 global kl: 0.1302706003189087
Saving best model with MSE 12.591388
it: 450, train mse: 15.042787551879883, local kl: 0.0 global kl: 0.17714284360408783 valid mse: 14.652362823486328, local kl: 0.0 global kl: 0.2707001566886902
it: 500, train mse: 14.881776809692383, local kl: 0.0 global kl: 0.09715428948402405 valid mse: 15.167205810546875, local kl: 0.0 global kl: 0.17649833858013153
it: 550, train mse: 9.534775733947754, local kl: 0.0 global kl: 0.125936359167099 valid mse: 14.04762077331543, local kl: 0.0 global kl: 0.39081618189811707
it: 600, train mse: 10.67753791809082, local kl: 0.0 global kl: 0.1544128805398941 valid mse: 14.358002662658691, local kl: 0.0 global kl: 0.494093120098114
it: 650, train mse: 7.732235908508301, local kl: 0.0 global kl: 0.116935133934021 valid mse: 14.264815330505371, local kl: 0.0 global kl: 0.31675615906715393
it: 700, train mse: 9.20595645904541, local kl: 0.0 global kl: 0.23000077903270721 valid mse: 10.581260681152344, local kl: 0.0 global kl: 0.2160193920135498
Saving best model with MSE 10.581261
it: 750, train mse: 16.144533157348633, local kl: 0.0 global kl: 0.0911886915564537 valid mse: 16.204015731811523, local kl: 0.0 global kl: 0.2451643943786621
it: 800, train mse: 8.892898559570312, local kl: 0.0 global kl: 0.13421142101287842 valid mse: 16.510469436645508, local kl: 0.0 global kl: 0.21452109515666962
it: 850, train mse: 7.690763473510742, local kl: 0.0 global kl: 0.10199811309576035 valid mse: 12.654316902160645, local kl: 0.0 global kl: 0.4183596670627594
it: 900, train mse: 18.442398071289062, local kl: 0.0 global kl: 0.12928704917430878 valid mse: 17.352256774902344, local kl: 0.0 global kl: 0.28427863121032715
it: 950, train mse: 6.338474750518799, local kl: 0.0 global kl: 0.1682315468788147 valid mse: 8.376662254333496, local kl: 0.0 global kl: 0.2932228744029999
Saving best model with MSE 8.376662
it: 1000, train mse: 7.555363178253174, local kl: 0.0 global kl: 0.12827828526496887 valid mse: 12.467473983764648, local kl: 0.0 global kl: 0.21405544877052307
it: 1050, train mse: 7.382735729217529, local kl: 0.0 global kl: 0.24700161814689636 valid mse: 10.812040328979492, local kl: 0.0 global kl: 0.3351515531539917
it: 1100, train mse: 11.826491355895996, local kl: 0.0 global kl: 0.07987679541110992 valid mse: 17.062219619750977, local kl: 0.0 global kl: 0.41916710138320923
it: 1150, train mse: 8.427934646606445, local kl: 0.0 global kl: 0.19679410755634308 valid mse: 8.912471771240234, local kl: 0.0 global kl: 0.3277907073497772
it: 1200, train mse: 8.379342079162598, local kl: 0.0 global kl: 0.1369568556547165 valid mse: 10.922608375549316, local kl: 0.0 global kl: 0.3203553557395935
it: 1250, train mse: 6.792413234710693, local kl: 0.0 global kl: 0.18436762690544128 valid mse: 10.786442756652832, local kl: 0.0 global kl: 0.3038572072982788
it: 1300, train mse: 7.740433692932129, local kl: 0.0 global kl: 0.20666618645191193 valid mse: 12.046319961547852, local kl: 0.0 global kl: 0.21469278633594513
it: 1350, train mse: 5.972925186157227, local kl: 0.0 global kl: 0.39754271507263184 valid mse: 7.023183822631836, local kl: 0.0 global kl: 0.38116636872291565
Saving best model with MSE 7.023184
it: 1400, train mse: 10.60970401763916, local kl: 0.0 global kl: 0.14856190979480743 valid mse: 10.271946907043457, local kl: 0.0 global kl: 0.38779565691947937
it: 1450, train mse: 9.16211986541748, local kl: 0.0 global kl: 0.11888305097818375 valid mse: 12.562023162841797, local kl: 0.0 global kl: 0.3161516487598419
it: 1500, train mse: 3.134488582611084, local kl: 0.0 global kl: 0.3573802411556244 valid mse: 8.63361644744873, local kl: 0.0 global kl: 0.24431192874908447
it: 1550, train mse: 5.610084056854248, local kl: 0.0 global kl: 0.26678380370140076 valid mse: 9.303025245666504, local kl: 0.0 global kl: 0.3231288492679596
it: 1600, train mse: 5.826364040374756, local kl: 0.0 global kl: 0.08549083769321442 valid mse: 6.212775707244873, local kl: 0.0 global kl: 0.284353643655777
Saving best model with MSE 6.2127757
it: 1650, train mse: 4.611496925354004, local kl: 0.0 global kl: 0.07007263600826263 valid mse: 11.719185829162598, local kl: 0.0 global kl: 0.19688056409358978
it: 1700, train mse: 11.371675491333008, local kl: 0.0 global kl: 0.1802794635295868 valid mse: 8.27175235748291, local kl: 0.0 global kl: 0.4620773196220398
it: 1750, train mse: 11.198724746704102, local kl: 0.0 global kl: 0.23512299358844757 valid mse: 10.494462013244629, local kl: 0.0 global kl: 0.44514065980911255
it: 1800, train mse: 12.170236587524414, local kl: 0.0 global kl: 0.07062403857707977 valid mse: 16.192943572998047, local kl: 0.0 global kl: 0.24359658360481262
it: 1850, train mse: 6.926184177398682, local kl: 0.0 global kl: 0.1169949397444725 valid mse: 13.89795970916748, local kl: 0.0 global kl: 0.3372594118118286
it: 1900, train mse: 6.7664289474487305, local kl: 0.0 global kl: 0.2340356558561325 valid mse: 7.677259922027588, local kl: 0.0 global kl: 0.22004374861717224
it: 1950, train mse: 8.522906303405762, local kl: 0.0 global kl: 0.24651682376861572 valid mse: 14.568679809570312, local kl: 0.0 global kl: 0.2974093556404114
it: 2000, train mse: 4.516035556793213, local kl: 0.0 global kl: 0.0867505818605423 valid mse: 6.60709285736084, local kl: 0.0 global kl: 0.5500221848487854
it: 2050, train mse: 4.406867027282715, local kl: 0.0 global kl: 0.11712487041950226 valid mse: 7.31290864944458, local kl: 0.0 global kl: 0.2846735119819641
it: 2100, train mse: 3.4695730209350586, local kl: 0.0 global kl: 0.2806495130062103 valid mse: 10.390429496765137, local kl: 0.0 global kl: 0.42837533354759216
it: 2150, train mse: 12.719401359558105, local kl: 0.0 global kl: 0.16507330536842346 valid mse: 5.458312511444092, local kl: 0.0 global kl: 0.28171461820602417
Saving best model with MSE 5.4583125
it: 2200, train mse: 7.4011735916137695, local kl: 0.0 global kl: 0.1250155121088028 valid mse: 7.14968729019165, local kl: 0.0 global kl: 0.24911625683307648
it: 2250, train mse: 8.659893035888672, local kl: 0.0 global kl: 0.029294883832335472 valid mse: 9.940103530883789, local kl: 0.0 global kl: 0.27238336205482483
it: 2300, train mse: 7.2967400550842285, local kl: 0.0 global kl: 0.07395368069410324 valid mse: 13.058205604553223, local kl: 0.0 global kl: 0.3927754759788513
it: 2350, train mse: 9.403611183166504, local kl: 0.0 global kl: 0.21801309287548065 valid mse: 7.716958522796631, local kl: 0.0 global kl: 0.37828224897384644
it: 2400, train mse: 3.911987781524658, local kl: 0.0 global kl: 0.03749764710664749 valid mse: 9.386279106140137, local kl: 0.0 global kl: 0.43720942735671997
it: 2450, train mse: 4.820832252502441, local kl: 0.0 global kl: 0.2392536699771881 valid mse: 7.872775077819824, local kl: 0.0 global kl: 0.30039241909980774
it: 2500, train mse: 6.899110794067383, local kl: 0.0 global kl: 0.2700323462486267 valid mse: 11.59496784210205, local kl: 0.0 global kl: 0.20490789413452148
it: 2550, train mse: 12.186851501464844, local kl: 0.0 global kl: 0.1546345204114914 valid mse: 9.826447486877441, local kl: 0.0 global kl: 0.28802409768104553
it: 2600, train mse: 3.9052484035491943, local kl: 0.0 global kl: 0.2517492473125458 valid mse: 6.484438896179199, local kl: 0.0 global kl: 0.2721247673034668
it: 2650, train mse: 5.212562561035156, local kl: 0.0 global kl: 0.1802128106355667 valid mse: 12.421480178833008, local kl: 0.0 global kl: 0.44573554396629333
it: 2700, train mse: 4.720589637756348, local kl: 0.0 global kl: 0.43730202317237854 valid mse: 10.604140281677246, local kl: 0.0 global kl: 0.2764524817466736
it: 2750, train mse: 6.041147708892822, local kl: 0.0 global kl: 0.20272274315357208 valid mse: 6.152613639831543, local kl: 0.0 global kl: 0.3266988694667816
it: 2800, train mse: 3.965606689453125, local kl: 0.0 global kl: 0.140569806098938 valid mse: 7.868770122528076, local kl: 0.0 global kl: 0.34088048338890076
it: 2850, train mse: 6.297147750854492, local kl: 0.0 global kl: 0.45554599165916443 valid mse: 6.681821823120117, local kl: 0.0 global kl: 0.4116120934486389
it: 2900, train mse: 5.309905052185059, local kl: 0.0 global kl: 0.22068099677562714 valid mse: 8.920299530029297, local kl: 0.0 global kl: 0.4095873236656189
it: 2950, train mse: 3.0267210006713867, local kl: 0.0 global kl: 0.23946984112262726 valid mse: 7.751859188079834, local kl: 0.0 global kl: 0.32985490560531616
it: 3000, train mse: 9.758288383483887, local kl: 0.0 global kl: 0.2895983159542084 valid mse: 10.41353702545166, local kl: 0.0 global kl: 0.314820796251297
it: 3050, train mse: 8.473219871520996, local kl: 0.0 global kl: 0.28835493326187134 valid mse: 6.7067742347717285, local kl: 0.0 global kl: 0.5905652046203613
it: 3100, train mse: 5.053966045379639, local kl: 0.0 global kl: 0.3140164315700531 valid mse: 6.0604472160339355, local kl: 0.0 global kl: 0.28155016899108887
it: 3150, train mse: 3.0540456771850586, local kl: 0.0 global kl: 0.061497438699007034 valid mse: 7.312976360321045, local kl: 0.0 global kl: 0.3602420389652252
it: 3200, train mse: 4.761858940124512, local kl: 0.0 global kl: 0.09011410176753998 valid mse: 9.813344955444336, local kl: 0.0 global kl: 0.2712126672267914
it: 3250, train mse: 9.5326509475708, local kl: 0.0 global kl: 0.0663316398859024 valid mse: 10.091053009033203, local kl: 0.0 global kl: 0.33996230363845825
it: 3300, train mse: 5.365706920623779, local kl: 0.0 global kl: 0.10492024570703506 valid mse: 5.37766695022583, local kl: 0.0 global kl: 0.4146920144557953
Saving best model with MSE 5.377667
it: 3350, train mse: 1.899269938468933, local kl: 0.0 global kl: 0.14059288799762726 valid mse: 9.084376335144043, local kl: 0.0 global kl: 0.4039068818092346
it: 3400, train mse: 3.986853837966919, local kl: 0.0 global kl: 0.06741016358137131 valid mse: 7.601688385009766, local kl: 0.0 global kl: 0.376309871673584
it: 3450, train mse: 4.048999786376953, local kl: 0.0 global kl: 0.18143513798713684 valid mse: 5.811102390289307, local kl: 0.0 global kl: 0.41353273391723633
it: 3500, train mse: 6.87155294418335, local kl: 0.0 global kl: 0.11555725336074829 valid mse: 13.586838722229004, local kl: 0.0 global kl: 0.3295601010322571
it: 3550, train mse: 7.851707935333252, local kl: 0.0 global kl: 0.21300263702869415 valid mse: 5.36508321762085, local kl: 0.0 global kl: 0.45388931035995483
Saving best model with MSE 5.365083
it: 3600, train mse: 5.925521373748779, local kl: 0.0 global kl: 0.13601389527320862 valid mse: 8.39080810546875, local kl: 0.0 global kl: 0.43038636445999146
it: 3650, train mse: 7.257022857666016, local kl: 0.0 global kl: 0.03677525371313095 valid mse: 6.6273579597473145, local kl: 0.0 global kl: 0.25810497999191284
it: 3700, train mse: 3.218953847885132, local kl: 0.0 global kl: 0.37948447465896606 valid mse: 8.089192390441895, local kl: 0.0 global kl: 0.29796186089515686
it: 3750, train mse: 5.663877964019775, local kl: 0.0 global kl: 0.6704846620559692 valid mse: 6.486609935760498, local kl: 0.0 global kl: 0.42541375756263733
it: 3800, train mse: 6.129737377166748, local kl: 0.0 global kl: 0.1962352693080902 valid mse: 11.257026672363281, local kl: 0.0 global kl: 0.3895520269870758
it: 3850, train mse: 4.051203727722168, local kl: 0.0 global kl: 0.10139000415802002 valid mse: 7.520037651062012, local kl: 0.0 global kl: 0.37576133012771606
it: 3900, train mse: 5.170616626739502, local kl: 0.0 global kl: 0.24590864777565002 valid mse: 7.342522621154785, local kl: 0.0 global kl: 0.28869879245758057
it: 3950, train mse: 8.378457069396973, local kl: 0.0 global kl: 0.10827463865280151 valid mse: 8.379571914672852, local kl: 0.0 global kl: 0.41714566946029663
it: 4000, train mse: 7.148488998413086, local kl: 0.0 global kl: 0.0751582607626915 valid mse: 6.872501373291016, local kl: 0.0 global kl: 0.3492758274078369
it: 4050, train mse: 4.542021751403809, local kl: 0.0 global kl: 0.26059025526046753 valid mse: 9.289677619934082, local kl: 0.0 global kl: 0.2330787181854248
it: 4100, train mse: 7.16706657409668, local kl: 0.0 global kl: 0.3492324948310852 valid mse: 5.678450584411621, local kl: 0.0 global kl: 0.2954118251800537
it: 4150, train mse: 3.7486979961395264, local kl: 0.0 global kl: 0.2796930968761444 valid mse: 7.265677452087402, local kl: 0.0 global kl: 0.45425352454185486
it: 4200, train mse: 6.847521781921387, local kl: 0.0 global kl: 0.054018598049879074 valid mse: 7.8295087814331055, local kl: 0.0 global kl: 0.3711893856525421
it: 4250, train mse: 4.685816764831543, local kl: 0.0 global kl: 0.42634710669517517 valid mse: 8.758159637451172, local kl: 0.0 global kl: 0.3825348913669586
it: 4300, train mse: 5.5344438552856445, local kl: 0.0 global kl: 0.22022227942943573 valid mse: 5.903736114501953, local kl: 0.0 global kl: 0.3595265746116638
it: 4350, train mse: 4.722398281097412, local kl: 0.0 global kl: 0.31746798753738403 valid mse: 7.319108009338379, local kl: 0.0 global kl: 0.38010507822036743
it: 4400, train mse: 5.760119438171387, local kl: 0.0 global kl: 0.10568954795598984 valid mse: 10.589418411254883, local kl: 0.0 global kl: 0.37727606296539307
it: 4450, train mse: 7.319855690002441, local kl: 0.0 global kl: 0.06301102787256241 valid mse: 8.740340232849121, local kl: 0.0 global kl: 0.3655267357826233
it: 4500, train mse: 3.1504905223846436, local kl: 0.0 global kl: 0.3285572826862335 valid mse: 7.369481086730957, local kl: 0.0 global kl: 0.2695663571357727
it: 4550, train mse: 12.340091705322266, local kl: 0.0 global kl: 0.08042008429765701 valid mse: 7.301823616027832, local kl: 0.0 global kl: 0.3903345465660095
it: 4600, train mse: 2.906344175338745, local kl: 0.0 global kl: 0.26688846945762634 valid mse: 6.935266494750977, local kl: 0.0 global kl: 0.4099596440792084
it: 4650, train mse: 2.3874430656433105, local kl: 0.0 global kl: 0.09007643908262253 valid mse: 7.919711589813232, local kl: 0.0 global kl: 0.35886842012405396
it: 4700, train mse: 12.935585021972656, local kl: 0.0 global kl: 0.4299693703651428 valid mse: 6.045656681060791, local kl: 0.0 global kl: 0.40540003776550293
it: 4750, train mse: 5.5298943519592285, local kl: 0.0 global kl: 0.1563889980316162 valid mse: 10.158427238464355, local kl: 0.0 global kl: 0.30935588479042053
it: 4800, train mse: 8.241556167602539, local kl: 0.0 global kl: 0.2652563452720642 valid mse: 11.171483993530273, local kl: 0.0 global kl: 0.3736705183982849
it: 4850, train mse: 3.1805484294891357, local kl: 0.0 global kl: 0.39079540967941284 valid mse: 9.77929401397705, local kl: 0.0 global kl: 0.33826833963394165
it: 4900, train mse: 4.4519829750061035, local kl: 0.0 global kl: 0.07853703200817108 valid mse: 6.185012340545654, local kl: 0.0 global kl: 0.35779574513435364
it: 4950, train mse: 3.2634706497192383, local kl: 0.0 global kl: 0.09261451661586761 valid mse: 8.981058120727539, local kl: 0.0 global kl: 0.30547672510147095
it: 5000, train mse: 4.802589416503906, local kl: 0.0 global kl: 0.2305392473936081 valid mse: 6.2857537269592285, local kl: 0.0 global kl: 0.2820480167865753
it: 5050, train mse: 10.452089309692383, local kl: 0.0 global kl: 0.4236159920692444 valid mse: 15.756779670715332, local kl: 0.0 global kl: 0.352389395236969
it: 5100, train mse: 4.418468952178955, local kl: 0.0 global kl: 0.2916039824485779 valid mse: 5.652602195739746, local kl: 0.0 global kl: 0.4395517408847809
it: 5150, train mse: 3.155707836151123, local kl: 0.0 global kl: 0.20065434277057648 valid mse: 6.407737731933594, local kl: 0.0 global kl: 0.3677622973918915
it: 5200, train mse: 5.155989170074463, local kl: 0.0 global kl: 0.06610047817230225 valid mse: 9.56169605255127, local kl: 0.0 global kl: 0.4589458107948303
it: 5250, train mse: 4.80708646774292, local kl: 0.0 global kl: 0.37294965982437134 valid mse: 5.848698616027832, local kl: 0.0 global kl: 0.3502209186553955
it: 5300, train mse: 4.2320966720581055, local kl: 0.0 global kl: 0.3940396308898926 valid mse: 9.30656623840332, local kl: 0.0 global kl: 0.30326324701309204
it: 5350, train mse: 3.368889093399048, local kl: 0.0 global kl: 0.11714529991149902 valid mse: 7.122623443603516, local kl: 0.0 global kl: 0.3140185475349426
it: 5400, train mse: 4.736330509185791, local kl: 0.0 global kl: 0.29291030764579773 valid mse: 8.740752220153809, local kl: 0.0 global kl: 0.33096784353256226
it: 5450, train mse: 3.0721261501312256, local kl: 0.0 global kl: 0.24153193831443787 valid mse: 7.526906967163086, local kl: 0.0 global kl: 0.23040533065795898
it: 5500, train mse: 2.251396417617798, local kl: 0.0 global kl: 0.23623883724212646 valid mse: 5.616119861602783, local kl: 0.0 global kl: 0.36190903186798096
it: 5550, train mse: 5.669631481170654, local kl: 0.0 global kl: 0.19129809737205505 valid mse: 6.927257061004639, local kl: 0.0 global kl: 0.39864856004714966
it: 5600, train mse: 4.986225128173828, local kl: 0.0 global kl: 0.1282101571559906 valid mse: 10.496126174926758, local kl: 0.0 global kl: 0.5013895630836487
it: 5650, train mse: 3.416712522506714, local kl: 0.0 global kl: 0.5132272839546204 valid mse: 7.912802219390869, local kl: 0.0 global kl: 0.2783695161342621
it: 5700, train mse: 6.409973621368408, local kl: 0.0 global kl: 0.26236265897750854 valid mse: 5.6769208908081055, local kl: 0.0 global kl: 0.330376535654068
it: 5750, train mse: 4.770817756652832, local kl: 0.0 global kl: 0.09122099727392197 valid mse: 8.434250831604004, local kl: 0.0 global kl: 0.22047221660614014
it: 5800, train mse: 5.001779556274414, local kl: 0.0 global kl: 0.12504197657108307 valid mse: 4.678805351257324, local kl: 0.0 global kl: 0.30371442437171936
Saving best model with MSE 4.6788054
it: 5850, train mse: 2.562816858291626, local kl: 0.0 global kl: 0.2263568937778473 valid mse: 8.591744422912598, local kl: 0.0 global kl: 0.27123093605041504
it: 5900, train mse: 5.965782642364502, local kl: 0.0 global kl: 0.14450441300868988 valid mse: 6.526638031005859, local kl: 0.0 global kl: 0.26184195280075073
it: 5950, train mse: 4.971260070800781, local kl: 0.0 global kl: 0.26506513357162476 valid mse: 9.050493240356445, local kl: 0.0 global kl: 0.24614223837852478
it: 6000, train mse: 2.139435052871704, local kl: 0.0 global kl: 0.21575240790843964 valid mse: 6.019064903259277, local kl: 0.0 global kl: 0.33791884779930115
it: 6050, train mse: 6.137020111083984, local kl: 0.0 global kl: 0.14495471119880676 valid mse: 7.378580570220947, local kl: 0.0 global kl: 0.40671825408935547
it: 6100, train mse: 2.989556074142456, local kl: 0.0 global kl: 0.28941601514816284 valid mse: 5.312650203704834, local kl: 0.0 global kl: 0.35394877195358276
it: 6150, train mse: 5.275032997131348, local kl: 0.0 global kl: 0.19572262465953827 valid mse: 9.568642616271973, local kl: 0.0 global kl: 0.3031867742538452
it: 6200, train mse: 2.097744941711426, local kl: 0.0 global kl: 0.11583449691534042 valid mse: 7.629317760467529, local kl: 0.0 global kl: 0.18689103424549103
it: 6250, train mse: 2.82601261138916, local kl: 0.0 global kl: 0.09567038714885712 valid mse: 4.10685920715332, local kl: 0.0 global kl: 0.25446128845214844
Saving best model with MSE 4.106859
it: 6300, train mse: 5.974459648132324, local kl: 0.0 global kl: 0.14638052880764008 valid mse: 9.20683765411377, local kl: 0.0 global kl: 0.2652411460876465
it: 6350, train mse: 1.8060892820358276, local kl: 0.0 global kl: 0.03624194115400314 valid mse: 5.081802845001221, local kl: 0.0 global kl: 0.2935090661048889
it: 6400, train mse: 4.471085071563721, local kl: 0.0 global kl: 0.10615763813257217 valid mse: 5.342784404754639, local kl: 0.0 global kl: 0.33748170733451843
it: 6450, train mse: 6.545074939727783, local kl: 0.0 global kl: 0.10726684331893921 valid mse: 6.896714687347412, local kl: 0.0 global kl: 0.323881596326828
it: 6500, train mse: 2.5383620262145996, local kl: 0.0 global kl: 0.23081830143928528 valid mse: 6.3591718673706055, local kl: 0.0 global kl: 0.34656813740730286
it: 6550, train mse: 5.719841957092285, local kl: 0.0 global kl: 0.4749657213687897 valid mse: 6.491209983825684, local kl: 0.0 global kl: 0.2878289818763733
it: 6600, train mse: 5.910828113555908, local kl: 0.0 global kl: 0.28966766595840454 valid mse: 6.139090061187744, local kl: 0.0 global kl: 0.37320029735565186
it: 6650, train mse: 3.6535539627075195, local kl: 0.0 global kl: 0.156722754240036 valid mse: 7.422572612762451, local kl: 0.0 global kl: 0.46536216139793396
it: 6700, train mse: 2.4783477783203125, local kl: 0.0 global kl: 0.5568627715110779 valid mse: 6.314305305480957, local kl: 0.0 global kl: 0.3284427225589752
it: 6750, train mse: 3.3709657192230225, local kl: 0.0 global kl: 0.050863467156887054 valid mse: 5.767014503479004, local kl: 0.0 global kl: 0.33910903334617615
it: 6800, train mse: 6.175292491912842, local kl: 0.0 global kl: 0.3491904139518738 valid mse: 4.503702163696289, local kl: 0.0 global kl: 0.32652440667152405
it: 6850, train mse: 3.8023219108581543, local kl: 0.0 global kl: 0.06765740364789963 valid mse: 6.622994422912598, local kl: 0.0 global kl: 0.3001803159713745
it: 6900, train mse: 4.160538196563721, local kl: 0.0 global kl: 0.4292406141757965 valid mse: 6.113997459411621, local kl: 0.0 global kl: 0.4045369029045105
it: 6950, train mse: 4.649363040924072, local kl: 0.0 global kl: 0.41961026191711426 valid mse: 4.276648044586182, local kl: 0.0 global kl: 0.3466185927391052
it: 7000, train mse: 3.80216646194458, local kl: 0.0 global kl: 0.24474628269672394 valid mse: 7.366701602935791, local kl: 0.0 global kl: 0.3217413127422333
it: 7050, train mse: 4.457913875579834, local kl: 0.0 global kl: 0.08160854876041412 valid mse: 8.236196517944336, local kl: 0.0 global kl: 0.3104216456413269
it: 7100, train mse: 4.036474227905273, local kl: 0.0 global kl: 0.34795302152633667 valid mse: 6.519223213195801, local kl: 0.0 global kl: 0.30121925473213196
it: 7150, train mse: 10.416582107543945, local kl: 0.0 global kl: 0.20938590168952942 valid mse: 6.476963520050049, local kl: 0.0 global kl: 0.31724563241004944
it: 7200, train mse: 4.6048431396484375, local kl: 0.0 global kl: 0.08076813071966171 valid mse: 8.524465560913086, local kl: 0.0 global kl: 0.3751816749572754
it: 7250, train mse: 3.6872944831848145, local kl: 0.0 global kl: 0.1997508853673935 valid mse: 6.197627544403076, local kl: 0.0 global kl: 0.3406832814216614
it: 7300, train mse: 3.8916993141174316, local kl: 0.0 global kl: 0.23886600136756897 valid mse: 7.22996711730957, local kl: 0.0 global kl: 0.3130055069923401
it: 7350, train mse: 2.611192464828491, local kl: 0.0 global kl: 0.3268502354621887 valid mse: 5.523129463195801, local kl: 0.0 global kl: 0.3285183012485504
it: 7400, train mse: 6.942709922790527, local kl: 0.0 global kl: 0.2782469391822815 valid mse: 7.836209297180176, local kl: 0.0 global kl: 0.31100982427597046
it: 7450, train mse: 3.344785451889038, local kl: 0.0 global kl: 0.26881664991378784 valid mse: 8.611660957336426, local kl: 0.0 global kl: 0.4491482675075531
it: 7500, train mse: 1.5923527479171753, local kl: 0.0 global kl: 0.27819710969924927 valid mse: 7.496548652648926, local kl: 0.0 global kl: 0.28556162118911743
it: 7550, train mse: 4.170473575592041, local kl: 0.0 global kl: 0.08307133615016937 valid mse: 10.237719535827637, local kl: 0.0 global kl: 0.30482497811317444
it: 7600, train mse: 6.553063869476318, local kl: 0.0 global kl: 0.17025312781333923 valid mse: 8.6735258102417, local kl: 0.0 global kl: 0.3645261526107788
it: 7650, train mse: 4.065461158752441, local kl: 0.0 global kl: 0.34236496686935425 valid mse: 8.479494094848633, local kl: 0.0 global kl: 0.36526840925216675
it: 7700, train mse: 4.961207866668701, local kl: 0.0 global kl: 0.055183954536914825 valid mse: 8.131888389587402, local kl: 0.0 global kl: 0.29826459288597107
it: 7750, train mse: 6.046692371368408, local kl: 0.0 global kl: 0.09669448435306549 valid mse: 3.9786760807037354, local kl: 0.0 global kl: 0.35963624715805054
Saving best model with MSE 3.978676
it: 7800, train mse: 6.818951606750488, local kl: 0.0 global kl: 0.24333927035331726 valid mse: 6.3762898445129395, local kl: 0.0 global kl: 0.3607324957847595
it: 7850, train mse: 2.4706339836120605, local kl: 0.0 global kl: 0.05488988757133484 valid mse: 4.414888381958008, local kl: 0.0 global kl: 0.34023556113243103
it: 7900, train mse: 2.1523842811584473, local kl: 0.0 global kl: 0.35624030232429504 valid mse: 9.29452133178711, local kl: 0.0 global kl: 0.2989594340324402
it: 7950, train mse: 3.2831366062164307, local kl: 0.0 global kl: 0.0817825049161911 valid mse: 5.546777248382568, local kl: 0.0 global kl: 0.34652942419052124
it: 8000, train mse: 2.419466018676758, local kl: 0.0 global kl: 0.15890830755233765 valid mse: 5.730711936950684, local kl: 0.0 global kl: 0.30686846375465393
it: 8050, train mse: 3.0915606021881104, local kl: 0.0 global kl: 0.18376436829566956 valid mse: 7.603568077087402, local kl: 0.0 global kl: 0.26619523763656616
it: 8100, train mse: 3.9810428619384766, local kl: 0.0 global kl: 0.18115364015102386 valid mse: 5.10783576965332, local kl: 0.0 global kl: 0.359058141708374
it: 8150, train mse: 3.3388726711273193, local kl: 0.0 global kl: 0.23145636916160583 valid mse: 5.9502339363098145, local kl: 0.0 global kl: 0.35192400217056274
it: 8200, train mse: 3.37048602104187, local kl: 0.0 global kl: 0.11535388231277466 valid mse: 8.301655769348145, local kl: 0.0 global kl: 0.36458468437194824
it: 8250, train mse: 4.109913349151611, local kl: 0.0 global kl: 0.41509026288986206 valid mse: 9.708446502685547, local kl: 0.0 global kl: 0.3672819435596466
it: 8300, train mse: 5.04306697845459, local kl: 0.0 global kl: 0.17482152581214905 valid mse: 5.784182071685791, local kl: 0.0 global kl: 0.3562691807746887
it: 8350, train mse: 5.084077835083008, local kl: 0.0 global kl: 0.36874935030937195 valid mse: 4.468667984008789, local kl: 0.0 global kl: 0.36787712574005127
it: 8400, train mse: 4.271218776702881, local kl: 0.0 global kl: 0.10807012021541595 valid mse: 6.108529567718506, local kl: 0.0 global kl: 0.46954697370529175
it: 8450, train mse: 4.777900218963623, local kl: 0.0 global kl: 0.18805329501628876 valid mse: 4.6490325927734375, local kl: 0.0 global kl: 0.3332107365131378
it: 8500, train mse: 4.776600360870361, local kl: 0.0 global kl: 0.45112594962120056 valid mse: 5.256242752075195, local kl: 0.0 global kl: 0.27138400077819824
it: 8550, train mse: 4.138899803161621, local kl: 0.0 global kl: 0.10771618783473969 valid mse: 6.240035057067871, local kl: 0.0 global kl: 0.300680935382843
it: 8600, train mse: 1.576478362083435, local kl: 0.0 global kl: 0.1757890284061432 valid mse: 6.6346845626831055, local kl: 0.0 global kl: 0.37793534994125366
it: 8650, train mse: 2.8382034301757812, local kl: 0.0 global kl: 0.21107156574726105 valid mse: 6.79932165145874, local kl: 0.0 global kl: 0.3157055377960205
it: 8700, train mse: 3.407536029815674, local kl: 0.0 global kl: 0.37959223985671997 valid mse: 9.133509635925293, local kl: 0.0 global kl: 0.5125530362129211
it: 8750, train mse: 7.821719646453857, local kl: 0.0 global kl: 0.7021588087081909 valid mse: 5.413036346435547, local kl: 0.0 global kl: 0.43339404463768005
it: 8800, train mse: 5.183409214019775, local kl: 0.0 global kl: 0.17576152086257935 valid mse: 4.541306495666504, local kl: 0.0 global kl: 0.2747305929660797
it: 8850, train mse: 2.91965651512146, local kl: 0.0 global kl: 0.27040895819664 valid mse: 7.314889907836914, local kl: 0.0 global kl: 0.3331682085990906
it: 8900, train mse: 6.030102252960205, local kl: 0.0 global kl: 0.26144424080848694 valid mse: 3.5533640384674072, local kl: 0.0 global kl: 0.3404063284397125
Saving best model with MSE 3.553364
it: 8950, train mse: 4.397461414337158, local kl: 0.0 global kl: 0.05493462085723877 valid mse: 5.9842681884765625, local kl: 0.0 global kl: 0.274897962808609
it: 9000, train mse: 2.3178467750549316, local kl: 0.0 global kl: 0.4220324456691742 valid mse: 8.240208625793457, local kl: 0.0 global kl: 0.3417636454105377
it: 9050, train mse: 3.9416871070861816, local kl: 0.0 global kl: 0.3305783271789551 valid mse: 6.444085121154785, local kl: 0.0 global kl: 0.38608676195144653
it: 9100, train mse: 3.0786702632904053, local kl: 0.0 global kl: 0.2676463723182678 valid mse: 5.586668014526367, local kl: 0.0 global kl: 0.31162920594215393
it: 9150, train mse: 4.8885111808776855, local kl: 0.0 global kl: 0.13672319054603577 valid mse: 6.070211410522461, local kl: 0.0 global kl: 0.3050900101661682
it: 9200, train mse: 5.312922477722168, local kl: 0.0 global kl: 0.2960338592529297 valid mse: 9.268487930297852, local kl: 0.0 global kl: 0.35168373584747314
it: 9250, train mse: 2.768794536590576, local kl: 0.0 global kl: 0.2655943036079407 valid mse: 5.806611061096191, local kl: 0.0 global kl: 0.3730645775794983
it: 9300, train mse: 5.302567481994629, local kl: 0.0 global kl: 0.15043455362319946 valid mse: 8.494799613952637, local kl: 0.0 global kl: 0.37083563208580017
it: 9350, train mse: 3.613917589187622, local kl: 0.0 global kl: 0.08114735037088394 valid mse: 5.149729251861572, local kl: 0.0 global kl: 0.3519126772880554
it: 9400, train mse: 3.5866565704345703, local kl: 0.0 global kl: 0.31550872325897217 valid mse: 4.176536560058594, local kl: 0.0 global kl: 0.3209688067436218
it: 9450, train mse: 3.6047780513763428, local kl: 0.0 global kl: 0.5527338981628418 valid mse: 7.673277378082275, local kl: 0.0 global kl: 0.32824787497520447
it: 9500, train mse: 3.573899984359741, local kl: 0.0 global kl: 0.41524648666381836 valid mse: 6.117753505706787, local kl: 0.0 global kl: 0.29492849111557007
it: 9550, train mse: 3.2054526805877686, local kl: 0.0 global kl: 0.1836562603712082 valid mse: 8.707452774047852, local kl: 0.0 global kl: 0.2986350357532501
it: 9600, train mse: 3.5614242553710938, local kl: 0.0 global kl: 0.16533608734607697 valid mse: 9.112504005432129, local kl: 0.0 global kl: 0.2638760507106781
it: 9650, train mse: 5.366582870483398, local kl: 0.0 global kl: 0.18603262305259705 valid mse: 5.131077766418457, local kl: 0.0 global kl: 0.3327139914035797
it: 9700, train mse: 4.284852504730225, local kl: 0.0 global kl: 0.10326925665140152 valid mse: 6.033595085144043, local kl: 0.0 global kl: 0.3148204982280731
it: 9750, train mse: 4.011636734008789, local kl: 0.0 global kl: 0.13394047319889069 valid mse: 6.1456074714660645, local kl: 0.0 global kl: 0.31065863370895386
it: 9800, train mse: 8.060441017150879, local kl: 0.0 global kl: 0.1340002864599228 valid mse: 8.053906440734863, local kl: 0.0 global kl: 0.30056092143058777
it: 9850, train mse: 3.4673500061035156, local kl: 0.0 global kl: 0.11179602146148682 valid mse: 11.445433616638184, local kl: 0.0 global kl: 0.3187212347984314
it: 9900, train mse: 6.45358419418335, local kl: 0.0 global kl: 0.20117692649364471 valid mse: 5.675044536590576, local kl: 0.0 global kl: 0.3423985242843628
it: 9950, train mse: 4.138010025024414, local kl: 0.0 global kl: 0.547824501991272 valid mse: 7.030754566192627, local kl: 0.0 global kl: 0.2879236042499542

In [0]:

    

model_type = 'anp'
x_y_encoder_net_sizes = [HIDDEN_SIZE]*2
global_latent_net_sizes = [HIDDEN_SIZE]*2
local_latent_net_sizes = None

model_hparams = tf.contrib.training.HParams(activation=tf.nn.relu,
                                            output_activation=tf.nn.relu,
                                            x_encoder_net_sizes=x_encoder_net_sizes,
                                            x_y_encoder_net_sizes=x_y_encoder_net_sizes,
                                            global_latent_net_sizes=global_latent_net_sizes,
                                            local_latent_net_sizes=local_latent_net_sizes,
                                            decoder_net_sizes=decoder_net_sizes, 
                                            heteroskedastic_net_sizes=heteroskedastic_net_sizes,
                                            att_type=att_type,
                                            att_heads=att_heads,
                                            model_type=model_type,
                                            data_uncertainty=data_uncertainty)
save_path = os.path.join(savedir, 'gnp_' + model_type + '.ckpt')
training_hparams = tf.contrib.training.HParams(lr=0.01,
                                               optimizer=tf.train.RMSPropOptimizer,
                                               num_iterations=10000,
                                               batch_size=10,
                                               num_context=num_context,
                                               num_target=num_target, 
                                               print_every=50,
                                               save_path=save_path,
                                               max_grad_norm=1000.0)

train(data_hparams,
      model_hparams,
      training_hparams)


    

    




it: 0, train mse: 64.7630081177, local kl: 0.0 global kl: 0.0284446384758 valid mse: 79.3176803589, local kl: 0.0 global kl: 0.0269404239953
Saving best model with MSE 79.31768
it: 50, train mse: 49.9610214233, local kl: 0.0 global kl: 0.0187800955027 valid mse: 68.77003479, local kl: 0.0 global kl: 0.00885808933526
Saving best model with MSE 68.770035
it: 100, train mse: 27.6876678467, local kl: 0.0 global kl: 0.00339605892077 valid mse: 25.6078796387, local kl: 0.0 global kl: 0.00314322533086
Saving best model with MSE 25.60788
it: 150, train mse: 16.9543037415, local kl: 0.0 global kl: 0.00767825264484 valid mse: 29.4613399506, local kl: 0.0 global kl: 0.00151743693277
it: 200, train mse: 37.5948066711, local kl: 0.0 global kl: 0.00202285428531 valid mse: 20.3689975739, local kl: 0.0 global kl: 0.00270827626809
Saving best model with MSE 20.368998
it: 250, train mse: 22.1894664764, local kl: 0.0 global kl: 0.00616811122745 valid mse: 30.6764945984, local kl: 0.0 global kl: 0.00491950521246
it: 300, train mse: 18.5624389648, local kl: 0.0 global kl: 0.0137356575578 valid mse: 34.4517326355, local kl: 0.0 global kl: 0.00311207491904
it: 350, train mse: 15.6443109512, local kl: 0.0 global kl: 0.0163160022348 valid mse: 21.6994438171, local kl: 0.0 global kl: 0.00504907220602
it: 400, train mse: 14.6541891098, local kl: 0.0 global kl: 0.00334953586571 valid mse: 16.4170608521, local kl: 0.0 global kl: 0.00715314876288
Saving best model with MSE 16.41706
it: 450, train mse: 16.6266994476, local kl: 0.0 global kl: 0.00957967154682 valid mse: 20.1749324799, local kl: 0.0 global kl: 0.00410199631006
it: 500, train mse: 14.8499917984, local kl: 0.0 global kl: 0.0160682443529 valid mse: 14.7912063599, local kl: 0.0 global kl: 0.0394393801689
Saving best model with MSE 14.791206
it: 550, train mse: 15.0168542862, local kl: 0.0 global kl: 0.0203667916358 valid mse: 24.9175224304, local kl: 0.0 global kl: 0.0371724143624
it: 600, train mse: 13.5670003891, local kl: 0.0 global kl: 0.0104093607515 valid mse: 13.8440799713, local kl: 0.0 global kl: 0.0103303324431
Saving best model with MSE 13.84408
it: 650, train mse: 18.3329257965, local kl: 0.0 global kl: 0.00731938192621 valid mse: 21.0681095123, local kl: 0.0 global kl: 0.0210833698511
it: 700, train mse: 14.7052783966, local kl: 0.0 global kl: 0.0132300201803 valid mse: 20.7271785736, local kl: 0.0 global kl: 0.00680379476398
it: 750, train mse: 12.7832841873, local kl: 0.0 global kl: 0.00759145105258 valid mse: 20.9799118042, local kl: 0.0 global kl: 0.0540205836296
it: 800, train mse: 11.6119804382, local kl: 0.0 global kl: 0.022030364722 valid mse: 14.8253450394, local kl: 0.0 global kl: 0.0165531728417
it: 850, train mse: 15.9163389206, local kl: 0.0 global kl: 0.0307333469391 valid mse: 15.124124527, local kl: 0.0 global kl: 0.0623716190457
it: 900, train mse: 12.7129774094, local kl: 0.0 global kl: 0.00782584678382 valid mse: 19.6301498413, local kl: 0.0 global kl: 0.00642355810851
it: 950, train mse: 11.8501091003, local kl: 0.0 global kl: 0.0241616275162 valid mse: 14.3880653381, local kl: 0.0 global kl: 0.0320474840701
it: 1000, train mse: 9.90816688538, local kl: 0.0 global kl: 0.0125908013433 valid mse: 16.5989322662, local kl: 0.0 global kl: 0.0209323652089
it: 1050, train mse: 10.2955036163, local kl: 0.0 global kl: 0.0244271345437 valid mse: 17.6416492462, local kl: 0.0 global kl: 0.0217897742987
it: 1100, train mse: 13.7523422241, local kl: 0.0 global kl: 0.00833088997751 valid mse: 13.5159378052, local kl: 0.0 global kl: 0.0324059315026
Saving best model with MSE 13.515938
it: 1150, train mse: 12.4388742447, local kl: 0.0 global kl: 0.0262939631939 valid mse: 12.5264701843, local kl: 0.0 global kl: 0.0229768846184
Saving best model with MSE 12.52647
it: 1200, train mse: 12.8899173737, local kl: 0.0 global kl: 0.0128693534061 valid mse: 14.204659462, local kl: 0.0 global kl: 0.0415768101811
it: 1250, train mse: 9.01453304291, local kl: 0.0 global kl: 0.0231869742274 valid mse: 13.7802190781, local kl: 0.0 global kl: 0.0410378538072
it: 1300, train mse: 16.5804271698, local kl: 0.0 global kl: 0.0411154925823 valid mse: 14.4452905655, local kl: 0.0 global kl: 0.0418199300766
it: 1350, train mse: 6.8604516983, local kl: 0.0 global kl: 0.0870117843151 valid mse: 10.6517763138, local kl: 0.0 global kl: 0.0525965876877
Saving best model with MSE 10.651776
it: 1400, train mse: 14.6389064789, local kl: 0.0 global kl: 0.0552962347865 valid mse: 14.7664690018, local kl: 0.0 global kl: 0.0588581748307
it: 1450, train mse: 14.9171218872, local kl: 0.0 global kl: 0.0196718499064 valid mse: 20.016450882, local kl: 0.0 global kl: 0.0339011400938
it: 1500, train mse: 8.17230987549, local kl: 0.0 global kl: 0.048060964793 valid mse: 11.9285383224, local kl: 0.0 global kl: 0.0303308367729
it: 1550, train mse: 9.80083084106, local kl: 0.0 global kl: 0.0448990873992 valid mse: 15.7737455368, local kl: 0.0 global kl: 0.0390437655151
it: 1600, train mse: 11.7229309082, local kl: 0.0 global kl: 0.0331151895225 valid mse: 12.2293748856, local kl: 0.0 global kl: 0.0353385768831
it: 1650, train mse: 9.19749069214, local kl: 0.0 global kl: 0.0166108366102 valid mse: 12.0874938965, local kl: 0.0 global kl: 0.0307131167501
it: 1700, train mse: 28.7632694244, local kl: 0.0 global kl: 0.0347166880965 valid mse: 11.6974611282, local kl: 0.0 global kl: 0.0202935896814
it: 1750, train mse: 12.2418041229, local kl: 0.0 global kl: 0.15851585567 valid mse: 14.7094326019, local kl: 0.0 global kl: 0.0355505906045
it: 1800, train mse: 12.5971336365, local kl: 0.0 global kl: 0.0176427271217 valid mse: 16.3794803619, local kl: 0.0 global kl: 0.0392772480845
it: 1850, train mse: 9.76136112213, local kl: 0.0 global kl: 0.0280447788537 valid mse: 19.4201507568, local kl: 0.0 global kl: 0.0455664880574
it: 1900, train mse: 7.28677272797, local kl: 0.0 global kl: 0.0248185172677 valid mse: 14.9233903885, local kl: 0.0 global kl: 0.0334033146501
it: 1950, train mse: 15.6932659149, local kl: 0.0 global kl: 0.194687604904 valid mse: 11.623544693, local kl: 0.0 global kl: 0.0322106257081
it: 2000, train mse: 7.93204307556, local kl: 0.0 global kl: 0.00810660794377 valid mse: 12.1560506821, local kl: 0.0 global kl: 0.047209803015
it: 2050, train mse: 6.16339159012, local kl: 0.0 global kl: 0.0296894852072 valid mse: 15.2900409698, local kl: 0.0 global kl: 0.0306191928685
it: 2100, train mse: 10.7622795105, local kl: 0.0 global kl: 0.0546664409339 valid mse: 11.744969368, local kl: 0.0 global kl: 0.0417997539043
it: 2150, train mse: 14.4711065292, local kl: 0.0 global kl: 0.0089347185567 valid mse: 11.8025245667, local kl: 0.0 global kl: 0.0201053209603
it: 2200, train mse: 7.81746244431, local kl: 0.0 global kl: 0.0126971825957 valid mse: 14.053527832, local kl: 0.0 global kl: 0.0368904620409
it: 2250, train mse: 8.94724273682, local kl: 0.0 global kl: 0.00310469139367 valid mse: 17.8659992218, local kl: 0.0 global kl: 0.032523393631
it: 2300, train mse: 9.54660797119, local kl: 0.0 global kl: 0.0195842441171 valid mse: 14.5248613358, local kl: 0.0 global kl: 0.0330222211778
it: 2350, train mse: 9.15106201172, local kl: 0.0 global kl: 0.0200284961611 valid mse: 12.5112524033, local kl: 0.0 global kl: 0.036079864949
it: 2400, train mse: 6.10880899429, local kl: 0.0 global kl: 0.0184751674533 valid mse: 12.5632419586, local kl: 0.0 global kl: 0.0297345612198
it: 2450, train mse: 9.48158454895, local kl: 0.0 global kl: 0.00641339411959 valid mse: 16.5380992889, local kl: 0.0 global kl: 0.0211587138474
it: 2500, train mse: 12.6500406265, local kl: 0.0 global kl: 0.061958860606 valid mse: 16.4804058075, local kl: 0.0 global kl: 0.0280452556908
it: 2550, train mse: 4.90526580811, local kl: 0.0 global kl: 0.022186184302 valid mse: 13.7471199036, local kl: 0.0 global kl: 0.0264694690704
it: 2600, train mse: 8.20464324951, local kl: 0.0 global kl: 0.0316647812724 valid mse: 13.7521467209, local kl: 0.0 global kl: 0.0314339473844
it: 2650, train mse: 8.18421363831, local kl: 0.0 global kl: 0.0154923368245 valid mse: 12.1458339691, local kl: 0.0 global kl: 0.0231662932783
it: 2700, train mse: 15.7996330261, local kl: 0.0 global kl: 0.0230476893485 valid mse: 17.4103851318, local kl: 0.0 global kl: 0.021580344066
it: 2750, train mse: 9.9872636795, local kl: 0.0 global kl: 0.0157035477459 valid mse: 10.347070694, local kl: 0.0 global kl: 0.0321888811886
Saving best model with MSE 10.347071
it: 2800, train mse: 6.45690441132, local kl: 0.0 global kl: 0.0163348354399 valid mse: 11.9602985382, local kl: 0.0 global kl: 0.0204443074763
it: 2850, train mse: 10.5977420807, local kl: 0.0 global kl: 0.0593300238252 valid mse: 10.6316270828, local kl: 0.0 global kl: 0.0244000200182
it: 2900, train mse: 7.71971035004, local kl: 0.0 global kl: 0.0205303765833 valid mse: 14.4835329056, local kl: 0.0 global kl: 0.0290428455919
it: 2950, train mse: 8.59158992767, local kl: 0.0 global kl: 0.0391322188079 valid mse: 12.2872543335, local kl: 0.0 global kl: 0.0393593423069
it: 3000, train mse: 9.33846855164, local kl: 0.0 global kl: 0.0227700509131 valid mse: 12.0833559036, local kl: 0.0 global kl: 0.010295746848
it: 3050, train mse: 9.55200958252, local kl: 0.0 global kl: 0.0137119591236 valid mse: 12.2542972565, local kl: 0.0 global kl: 0.0177311785519
it: 3100, train mse: 9.45389175415, local kl: 0.0 global kl: 0.0518323592842 valid mse: 10.1224222183, local kl: 0.0 global kl: 0.0115179959685
Saving best model with MSE 10.122422
it: 3150, train mse: 8.53475379944, local kl: 0.0 global kl: 0.00556532526389 valid mse: 10.3664722443, local kl: 0.0 global kl: 0.0167824309319
it: 3200, train mse: 14.7130670547, local kl: 0.0 global kl: 0.00725350389257 valid mse: 13.4216785431, local kl: 0.0 global kl: 0.0166822876781
it: 3250, train mse: 6.08467149734, local kl: 0.0 global kl: 0.00810975581408 valid mse: 16.4777603149, local kl: 0.0 global kl: 0.00983073562384
it: 3300, train mse: 8.09979152679, local kl: 0.0 global kl: 0.012577672489 valid mse: 9.96167469025, local kl: 0.0 global kl: 0.00848270766437
Saving best model with MSE 9.961675
it: 3350, train mse: 4.72372865677, local kl: 0.0 global kl: 0.00613280432299 valid mse: 11.0637807846, local kl: 0.0 global kl: 0.00615393230692
it: 3400, train mse: 10.315615654, local kl: 0.0 global kl: 0.00656230887398 valid mse: 11.6009082794, local kl: 0.0 global kl: 0.0132741276175
it: 3450, train mse: 8.36712646484, local kl: 0.0 global kl: 0.00555268069729 valid mse: 13.8226156235, local kl: 0.0 global kl: 0.00762661313638
it: 3500, train mse: 11.0030622482, local kl: 0.0 global kl: 0.0142816472799 valid mse: 16.2914791107, local kl: 0.0 global kl: 0.0123196747154
it: 3550, train mse: 10.3917589188, local kl: 0.0 global kl: 0.0188368223608 valid mse: 10.2939329147, local kl: 0.0 global kl: 0.00771629530936
it: 3600, train mse: 6.3807721138, local kl: 0.0 global kl: 0.00318354321644 valid mse: 14.1263093948, local kl: 0.0 global kl: 0.00837360043079
it: 3650, train mse: 10.9951467514, local kl: 0.0 global kl: 0.00202321028337 valid mse: 12.3924980164, local kl: 0.0 global kl: 0.00522702978924
it: 3700, train mse: 7.23176860809, local kl: 0.0 global kl: 0.00266489479691 valid mse: 12.7200927734, local kl: 0.0 global kl: 0.00396175170317
it: 3750, train mse: 8.55606555939, local kl: 0.0 global kl: 0.00925898831338 valid mse: 14.0137720108, local kl: 0.0 global kl: 0.00250621861778
it: 3800, train mse: 11.3390598297, local kl: 0.0 global kl: 0.00218505295925 valid mse: 14.3661346436, local kl: 0.0 global kl: 0.0047243218869
it: 3850, train mse: 7.99956226349, local kl: 0.0 global kl: 0.00250096130185 valid mse: 12.1657915115, local kl: 0.0 global kl: 0.00190852605738
it: 3900, train mse: 7.7180929184, local kl: 0.0 global kl: 0.00893716048449 valid mse: 11.8147439957, local kl: 0.0 global kl: 0.00306757772341
it: 3950, train mse: 12.9125757217, local kl: 0.0 global kl: 0.00161541963462 valid mse: 12.2687330246, local kl: 0.0 global kl: 0.00504614925012
it: 4000, train mse: 8.82065868378, local kl: 0.0 global kl: 0.00273153814487 valid mse: 9.80111789703, local kl: 0.0 global kl: 0.0115208532661
Saving best model with MSE 9.801118
it: 4050, train mse: 6.87136173248, local kl: 0.0 global kl: 0.000998712028377 valid mse: 11.104432106, local kl: 0.0 global kl: 0.00125319801737
it: 4100, train mse: 11.4629602432, local kl: 0.0 global kl: 0.00103445572313 valid mse: 9.58469200134, local kl: 0.0 global kl: 0.00390500202775
Saving best model with MSE 9.584692
it: 4150, train mse: 6.50435495377, local kl: 0.0 global kl: 0.00108770641964 valid mse: 10.7875213623, local kl: 0.0 global kl: 0.00105466297828
it: 4200, train mse: 13.2464532852, local kl: 0.0 global kl: 0.00143365247641 valid mse: 15.5973091125, local kl: 0.0 global kl: 0.005217297934
it: 4250, train mse: 5.94040822983, local kl: 0.0 global kl: 0.000811045058072 valid mse: 7.88524675369, local kl: 0.0 global kl: 0.00688512763008
Saving best model with MSE 7.8852468
it: 4300, train mse: 14.4908676147, local kl: 0.0 global kl: 0.00200159614906 valid mse: 9.86511230469, local kl: 0.0 global kl: 0.0252949949354
it: 4350, train mse: 5.12515974045, local kl: 0.0 global kl: 0.000958376564085 valid mse: 9.67568874359, local kl: 0.0 global kl: 0.000272515462711
it: 4400, train mse: 10.154255867, local kl: 0.0 global kl: 0.000526558258571 valid mse: 9.83167457581, local kl: 0.0 global kl: 0.00285768928006
it: 4450, train mse: 18.5940647125, local kl: 0.0 global kl: 0.00094478239771 valid mse: 15.1388969421, local kl: 0.0 global kl: 0.000459858740214
it: 4500, train mse: 7.80405759811, local kl: 0.0 global kl: 0.000291106523946 valid mse: 8.2906627655, local kl: 0.0 global kl: 0.00043852214003
it: 4550, train mse: 6.47125482559, local kl: 0.0 global kl: 0.000201356597245 valid mse: 7.43446540833, local kl: 0.0 global kl: 0.000381313351681
Saving best model with MSE 7.4344654
it: 4600, train mse: 3.1276049614, local kl: 0.0 global kl: 0.000623437575996 valid mse: 8.65041160583, local kl: 0.0 global kl: 0.000439082650701
it: 4650, train mse: 4.53352880478, local kl: 0.0 global kl: 0.00382654299028 valid mse: 8.11879920959, local kl: 0.0 global kl: 0.00240699434653
it: 4700, train mse: 8.64017963409, local kl: 0.0 global kl: 0.00354858161882 valid mse: 16.0375881195, local kl: 0.0 global kl: 0.0035379207693
it: 4750, train mse: 6.09897756577, local kl: 0.0 global kl: 0.000678872922435 valid mse: 10.927324295, local kl: 0.0 global kl: 0.000929847825319
it: 4800, train mse: 14.3970279694, local kl: 0.0 global kl: 0.00463562551886 valid mse: 9.73890972137, local kl: 0.0 global kl: 0.00216364115477
it: 4850, train mse: 6.58751583099, local kl: 0.0 global kl: 0.00482672266662 valid mse: 8.70272922516, local kl: 0.0 global kl: 0.000501747592352
it: 4900, train mse: 3.93574953079, local kl: 0.0 global kl: 0.000523503113072 valid mse: 9.94631195068, local kl: 0.0 global kl: 0.000965662649833
it: 4950, train mse: 4.55448627472, local kl: 0.0 global kl: 0.000526734103914 valid mse: 9.6288690567, local kl: 0.0 global kl: 0.000616802135482
it: 5000, train mse: 6.87348604202, local kl: 0.0 global kl: 0.00277106952854 valid mse: 8.04102993011, local kl: 0.0 global kl: 0.000840869382955
it: 5050, train mse: 12.1860265732, local kl: 0.0 global kl: 0.00338411075063 valid mse: 26.117105484, local kl: 0.0 global kl: 0.000942699261941
it: 5100, train mse: 6.98806428909, local kl: 0.0 global kl: 0.00151172582991 valid mse: 10.0117607117, local kl: 0.0 global kl: 0.00649570580572
it: 5150, train mse: 6.63073825836, local kl: 0.0 global kl: 0.00011873881158 valid mse: 8.07402515411, local kl: 0.0 global kl: 0.000596080324613
it: 5200, train mse: 9.67107391357, local kl: 0.0 global kl: 0.000389726832509 valid mse: 12.2125806808, local kl: 0.0 global kl: 0.00018779860693
it: 5250, train mse: 9.34505939484, local kl: 0.0 global kl: 0.000166355661349 valid mse: 10.8388776779, local kl: 0.0 global kl: 0.00107367988676
it: 5300, train mse: 5.37861633301, local kl: 0.0 global kl: 0.00280993245542 valid mse: 8.88254547119, local kl: 0.0 global kl: 7.56903682486e-05
it: 5350, train mse: 5.965072155, local kl: 0.0 global kl: 0.000146397331264 valid mse: 9.85553836823, local kl: 0.0 global kl: 0.0022602956742
it: 5400, train mse: 5.15051364899, local kl: 0.0 global kl: 0.000875322963111 valid mse: 8.52423286438, local kl: 0.0 global kl: 0.00283210678026
it: 5450, train mse: 4.30928230286, local kl: 0.0 global kl: 0.00025451098918 valid mse: 10.3733119965, local kl: 0.0 global kl: 0.000267201190582
it: 5500, train mse: 6.32313156128, local kl: 0.0 global kl: 0.000223893715884 valid mse: 6.66473913193, local kl: 0.0 global kl: 0.000835072714835
Saving best model with MSE 6.664739
it: 5550, train mse: 7.10088729858, local kl: 0.0 global kl: 0.00308933178894 valid mse: 9.02118015289, local kl: 0.0 global kl: 0.00248698634095
it: 5600, train mse: 5.73894786835, local kl: 0.0 global kl: 0.000535689468961 valid mse: 16.3189983368, local kl: 0.0 global kl: 0.000297353428323
it: 5650, train mse: 6.53498363495, local kl: 0.0 global kl: 0.000162272190209 valid mse: 10.4136037827, local kl: 0.0 global kl: 0.00157965277322
it: 5700, train mse: 5.97054195404, local kl: 0.0 global kl: 0.00122974417172 valid mse: 9.08224391937, local kl: 0.0 global kl: 0.000368570443243
it: 5750, train mse: 9.71584320068, local kl: 0.0 global kl: 0.000367580505554 valid mse: 13.6934843063, local kl: 0.0 global kl: 0.000494012085255
it: 5800, train mse: 4.64285373688, local kl: 0.0 global kl: 0.00106024951674 valid mse: 6.82003974915, local kl: 0.0 global kl: 0.00308306608349
it: 5850, train mse: 4.0269536972, local kl: 0.0 global kl: 0.000892104464583 valid mse: 10.3829870224, local kl: 0.0 global kl: 0.00051659962628
it: 5900, train mse: 6.70267677307, local kl: 0.0 global kl: 0.000206738011912 valid mse: 9.20429897308, local kl: 0.0 global kl: 0.00128472759388
it: 5950, train mse: 5.82171392441, local kl: 0.0 global kl: 0.000408051593695 valid mse: 7.0138759613, local kl: 0.0 global kl: 0.000308591988869
it: 6000, train mse: 4.40027856827, local kl: 0.0 global kl: 0.000796438078396 valid mse: 9.12188148499, local kl: 0.0 global kl: 0.000128895670059
it: 6050, train mse: 9.53746509552, local kl: 0.0 global kl: 0.00760453939438 valid mse: 7.81054210663, local kl: 0.0 global kl: 0.00175522326026
it: 6100, train mse: 4.45688772202, local kl: 0.0 global kl: 0.000260949047515 valid mse: 8.95620918274, local kl: 0.0 global kl: 0.00145903311204
it: 6150, train mse: 6.03924846649, local kl: 0.0 global kl: 0.000122955621919 valid mse: 6.76616001129, local kl: 0.0 global kl: 0.000429089646786
it: 6200, train mse: 5.48244524002, local kl: 0.0 global kl: 0.000390308821807 valid mse: 20.5469837189, local kl: 0.0 global kl: 0.000555784616154
it: 6250, train mse: 4.27764511108, local kl: 0.0 global kl: 0.000308468239382 valid mse: 7.8664188385, local kl: 0.0 global kl: 0.000310756178806
it: 6300, train mse: 11.366818428, local kl: 0.0 global kl: 0.00351597974077 valid mse: 9.82455921173, local kl: 0.0 global kl: 0.00170724373311
it: 6350, train mse: 2.46166729927, local kl: 0.0 global kl: 0.000714055553544 valid mse: 6.72135543823, local kl: 0.0 global kl: 0.00338607584126
it: 6400, train mse: 4.7444729805, local kl: 0.0 global kl: 0.00102824135683 valid mse: 9.2028875351, local kl: 0.0 global kl: 0.00624210480601
it: 6450, train mse: 5.53831386566, local kl: 0.0 global kl: 0.00308912643231 valid mse: 6.8811955452, local kl: 0.0 global kl: 0.00148018007167
it: 6500, train mse: 3.15298366547, local kl: 0.0 global kl: 0.00220530456863 valid mse: 9.01347064972, local kl: 0.0 global kl: 0.00370546942577
it: 6550, train mse: 6.04751348495, local kl: 0.0 global kl: 0.000499614805449 valid mse: 7.63133144379, local kl: 0.0 global kl: 0.000131436710944
it: 6600, train mse: 4.04490756989, local kl: 0.0 global kl: 0.000656728167087 valid mse: 10.3360509872, local kl: 0.0 global kl: 0.00073360017268
it: 6650, train mse: 5.60446500778, local kl: 0.0 global kl: 0.000285047892248 valid mse: 7.65454673767, local kl: 0.0 global kl: 0.000616347009782
it: 6700, train mse: 4.36028718948, local kl: 0.0 global kl: 0.00453044194728 valid mse: 8.8675365448, local kl: 0.0 global kl: 0.000881977204699
it: 6750, train mse: 5.67843151093, local kl: 0.0 global kl: 0.000437284907093 valid mse: 7.22657823563, local kl: 0.0 global kl: 0.000824737478979
it: 6800, train mse: 4.84189653397, local kl: 0.0 global kl: 0.024709969759 valid mse: 4.90496969223, local kl: 0.0 global kl: 0.0236764252186
Saving best model with MSE 4.9049697
it: 6850, train mse: 4.26179456711, local kl: 0.0 global kl: 0.00055028940551 valid mse: 7.25542116165, local kl: 0.0 global kl: 0.000114741618745
it: 6900, train mse: 6.36087274551, local kl: 0.0 global kl: 0.00129798124544 valid mse: 8.18745136261, local kl: 0.0 global kl: 0.0027166348882
it: 6950, train mse: 5.87736320496, local kl: 0.0 global kl: 0.0126898065209 valid mse: 7.43747520447, local kl: 0.0 global kl: 0.00440812529996
it: 7000, train mse: 5.52474021912, local kl: 0.0 global kl: 0.000309136754368 valid mse: 8.28202152252, local kl: 0.0 global kl: 0.00153665663674
it: 7050, train mse: 5.03467321396, local kl: 0.0 global kl: 0.000116998307931 valid mse: 11.8256807327, local kl: 0.0 global kl: 0.0017861276865
it: 7100, train mse: 6.45345830917, local kl: 0.0 global kl: 0.00130525627173 valid mse: 6.01131296158, local kl: 0.0 global kl: 0.000844441354275
it: 7150, train mse: 11.2650690079, local kl: 0.0 global kl: 0.000157113885507 valid mse: 6.48564767838, local kl: 0.0 global kl: 0.00114156899508
it: 7200, train mse: 8.53460502625, local kl: 0.0 global kl: 0.00205096113496 valid mse: 10.4135475159, local kl: 0.0 global kl: 0.000136847753311
it: 7250, train mse: 7.23005723953, local kl: 0.0 global kl: 0.000381057790946 valid mse: 7.72998571396, local kl: 0.0 global kl: 0.000160030424013
it: 7300, train mse: 5.12624502182, local kl: 0.0 global kl: 0.00106563407462 valid mse: 7.70452642441, local kl: 0.0 global kl: 0.00255457358435
it: 7350, train mse: 3.60031795502, local kl: 0.0 global kl: 0.00024006034073 valid mse: 7.44686508179, local kl: 0.0 global kl: 0.000285424524918
it: 7400, train mse: 6.49435758591, local kl: 0.0 global kl: 0.00106653175317 valid mse: 7.85102128983, local kl: 0.0 global kl: 0.0047991941683
it: 7450, train mse: 4.9187374115, local kl: 0.0 global kl: 0.000439968978753 valid mse: 8.18260383606, local kl: 0.0 global kl: 9.6586496511e-05
it: 7500, train mse: 2.73666954041, local kl: 0.0 global kl: 0.000841789296828 valid mse: 6.619535923, local kl: 0.0 global kl: 0.0027444276493
it: 7550, train mse: 5.79969215393, local kl: 0.0 global kl: 0.00330476043746 valid mse: 9.35495853424, local kl: 0.0 global kl: 0.00115711439867
it: 7600, train mse: 4.49835205078, local kl: 0.0 global kl: 0.00194122712128 valid mse: 5.87750768661, local kl: 0.0 global kl: 0.00189034605864
it: 7650, train mse: 22.8320846558, local kl: 0.0 global kl: 0.000949108216446 valid mse: 7.05777597427, local kl: 0.0 global kl: 0.000483049923787
it: 7700, train mse: 5.38102054596, local kl: 0.0 global kl: 0.000185582990525 valid mse: 7.47399377823, local kl: 0.0 global kl: 0.000109604516183
it: 7750, train mse: 4.68957281113, local kl: 0.0 global kl: 0.00572677329183 valid mse: 6.69128990173, local kl: 0.0 global kl: 0.0018532841932
it: 7800, train mse: 6.66092443466, local kl: 0.0 global kl: 0.00310849444941 valid mse: 10.4606180191, local kl: 0.0 global kl: 0.000243645947194
it: 7850, train mse: 3.11577033997, local kl: 0.0 global kl: 1.5451314539e-05 valid mse: 6.82742834091, local kl: 0.0 global kl: 0.00146804994438
it: 7900, train mse: 4.57215595245, local kl: 0.0 global kl: 0.00259584211744 valid mse: 8.10348033905, local kl: 0.0 global kl: 0.000585200614296
it: 7950, train mse: 4.70207834244, local kl: 0.0 global kl: 0.000530009448994 valid mse: 10.0070972443, local kl: 0.0 global kl: 0.000541696383152
it: 8000, train mse: 3.06886363029, local kl: 0.0 global kl: 0.00313439453021 valid mse: 7.23742341995, local kl: 0.0 global kl: 0.000539263477549
it: 8050, train mse: 6.78094291687, local kl: 0.0 global kl: 4.61612762592e-05 valid mse: 6.10177326202, local kl: 0.0 global kl: 0.000341662147548
it: 8100, train mse: 5.6431760788, local kl: 0.0 global kl: 0.0010191409383 valid mse: 8.63405227661, local kl: 0.0 global kl: 0.000874067656696
it: 8150, train mse: 5.46238040924, local kl: 0.0 global kl: 0.000547581643332 valid mse: 5.91333293915, local kl: 0.0 global kl: 0.00166004034691
it: 8200, train mse: 4.9228310585, local kl: 0.0 global kl: 0.00308890594169 valid mse: 16.0339565277, local kl: 0.0 global kl: 0.00500393100083
it: 8250, train mse: 3.8587975502, local kl: 0.0 global kl: 0.00069486751454 valid mse: 13.1743860245, local kl: 0.0 global kl: 0.000862545624841
it: 8300, train mse: 3.83528280258, local kl: 0.0 global kl: 0.000963471713476 valid mse: 9.77619457245, local kl: 0.0 global kl: 0.00142085982952
it: 8350, train mse: 7.03997039795, local kl: 0.0 global kl: 0.000634289113805 valid mse: 4.97336769104, local kl: 0.0 global kl: 0.00145085644908
it: 8400, train mse: 4.43544721603, local kl: 0.0 global kl: 0.000613922486082 valid mse: 6.5383181572, local kl: 0.0 global kl: 0.000189305486856
it: 8450, train mse: 14.3611030579, local kl: 0.0 global kl: 6.85721897753e-05 valid mse: 6.50342130661, local kl: 0.0 global kl: 0.00605388358235
it: 8500, train mse: 6.32074356079, local kl: 0.0 global kl: 0.00178860756569 valid mse: 5.69865942001, local kl: 0.0 global kl: 0.0014897917863
it: 8550, train mse: 4.997402668, local kl: 0.0 global kl: 0.000612749136053 valid mse: 6.93235301971, local kl: 0.0 global kl: 0.000319864106132
it: 8600, train mse: 2.14302778244, local kl: 0.0 global kl: 0.00106488354504 valid mse: 6.11469316483, local kl: 0.0 global kl: 0.00011900679965
it: 8650, train mse: 5.03133630753, local kl: 0.0 global kl: 0.00446807593107 valid mse: 6.52502918243, local kl: 0.0 global kl: 0.00233721965924
it: 8700, train mse: 5.98341035843, local kl: 0.0 global kl: 0.000268626899924 valid mse: 5.51168870926, local kl: 0.0 global kl: 0.000839194399305
it: 8750, train mse: 9.60915470123, local kl: 0.0 global kl: 0.00140433199704 valid mse: 6.29504680634, local kl: 0.0 global kl: 0.00027540238807
it: 8800, train mse: 4.92585134506, local kl: 0.0 global kl: 0.0005673514097 valid mse: 5.98855018616, local kl: 0.0 global kl: 0.000232818274526
it: 8850, train mse: 4.73342561722, local kl: 0.0 global kl: 0.0025003971532 valid mse: 5.43116807938, local kl: 0.0 global kl: 0.000823396432679
it: 8900, train mse: 4.28579854965, local kl: 0.0 global kl: 7.21645255908e-06 valid mse: 6.45592021942, local kl: 0.0 global kl: 3.19381761074e-05
it: 8950, train mse: 6.42366266251, local kl: 0.0 global kl: 0.00161480717361 valid mse: 5.7478556633, local kl: 0.0 global kl: 0.00137718149927
it: 9000, train mse: 4.9110045433, local kl: 0.0 global kl: 0.000163153672474 valid mse: 6.62253141403, local kl: 0.0 global kl: 0.000133672438096
it: 9050, train mse: 4.37285995483, local kl: 0.0 global kl: 0.0155428275466 valid mse: 7.01032114029, local kl: 0.0 global kl: 0.00603013252839
it: 9100, train mse: 2.39806365967, local kl: 0.0 global kl: 0.000801308080554 valid mse: 7.66832876205, local kl: 0.0 global kl: 0.000804585579317
it: 9150, train mse: 4.28036642075, local kl: 0.0 global kl: 0.00141456234269 valid mse: 8.42846488953, local kl: 0.0 global kl: 0.00134164281189
it: 9200, train mse: 5.39240169525, local kl: 0.0 global kl: 0.00179158279207 valid mse: 9.80750656128, local kl: 0.0 global kl: 0.00222642044537
it: 9250, train mse: 3.42189073563, local kl: 0.0 global kl: 0.00147715606727 valid mse: 8.18420410156, local kl: 0.0 global kl: 0.00136872730218
it: 9300, train mse: 6.52889823914, local kl: 0.0 global kl: 0.00460201269016 valid mse: 11.1375608444, local kl: 0.0 global kl: 0.0155175896361
it: 9350, train mse: 3.82684922218, local kl: 0.0 global kl: 0.0028404253535 valid mse: 9.99074363708, local kl: 0.0 global kl: 0.0517904981971
it: 9400, train mse: 3.70641374588, local kl: 0.0 global kl: 0.00447480473667 valid mse: 6.30994749069, local kl: 0.0 global kl: 0.00175878521986
it: 9450, train mse: 3.37405943871, local kl: 0.0 global kl: 0.0 valid mse: 7.7958483696, local kl: 0.0 global kl: 5.69016447116e-05
it: 9500, train mse: 4.44150495529, local kl: 0.0 global kl: 0.0 valid mse: 6.47423362732, local kl: 0.0 global kl: 0.0
it: 9550, train mse: 3.35752177238, local kl: 0.0 global kl: 0.00158100074623 valid mse: 8.8787984848, local kl: 0.0 global kl: 0.000295011734124
it: 9600, train mse: 2.74135279655, local kl: 0.0 global kl: 0.00153329316527 valid mse: 8.66943454742, local kl: 0.0 global kl: 0.000642638304271
it: 9650, train mse: 9.17913913727, local kl: 0.0 global kl: 0.0 valid mse: 6.89774990082, local kl: 0.0 global kl: 0.00509135704488
it: 9700, train mse: 4.65834856033, local kl: 0.0 global kl: 0.0 valid mse: 6.6122546196, local kl: 0.0 global kl: 0.0
it: 9750, train mse: 3.57403874397, local kl: 0.0 global kl: 0.0 valid mse: 7.54349708557, local kl: 0.0 global kl: 0.0
it: 9800, train mse: 6.71903657913, local kl: 0.0 global kl: 0.0 valid mse: 9.41256523132, local kl: 0.0 global kl: 0.0
it: 9850, train mse: 4.21986579895, local kl: 0.0 global kl: 0.0 valid mse: 6.43475675583, local kl: 0.0 global kl: 0.0
it: 9900, train mse: 11.3623714447, local kl: 0.0 global kl: 0.0 valid mse: 5.40795230865, local kl: 0.0 global kl: 0.0
it: 9950, train mse: 3.62081885338, local kl: 0.0 global kl: 0.0 valid mse: 7.67432403564, local kl: 0.0 global kl: 0.0

In [0]:

    

model_type = 'acnp'
x_y_encoder_net_sizes = [HIDDEN_SIZE]*4
global_latent_net_sizes = None
local_latent_net_sizes = None

model_hparams = tf.contrib.training.HParams(activation=tf.nn.relu,
                                            output_activation=tf.nn.relu,
                                            x_encoder_net_sizes=x_encoder_net_sizes,
                                            x_y_encoder_net_sizes=x_y_encoder_net_sizes,
                                            global_latent_net_sizes=global_latent_net_sizes,
                                            local_latent_net_sizes=local_latent_net_sizes,
                                            decoder_net_sizes=decoder_net_sizes, 
                                            heteroskedastic_net_sizes=heteroskedastic_net_sizes,
                                            att_type=att_type,
                                            att_heads=att_heads,
                                            model_type=model_type,
                                            data_uncertainty=data_uncertainty)
save_path = os.path.join(savedir, 'gnp_' + model_type + '.ckpt')
training_hparams = tf.contrib.training.HParams(lr=0.01,
                                               optimizer=tf.train.RMSPropOptimizer,
                                               num_iterations=10000,
                                               batch_size=10,
                                               num_context=num_context,
                                               num_target=num_target, 
                                               print_every=50,
                                               save_path=save_path,
                                               max_grad_norm=1000.0)

train(data_hparams,
      model_hparams,
      training_hparams)


    

    




it: 0, train mse: 65.7367477417, local kl: 0.0 global kl: 0.0 valid mse: 81.7034683228, local kl: 0.0 global kl: 0.0
Saving best model with MSE 81.70347
it: 50, train mse: 65.3536453247, local kl: 0.0 global kl: 0.0 valid mse: 78.2391357422, local kl: 0.0 global kl: 0.0
Saving best model with MSE 78.239136
it: 100, train mse: 20.2306365967, local kl: 0.0 global kl: 0.0 valid mse: 24.6245040894, local kl: 0.0 global kl: 0.0
Saving best model with MSE 24.624504
it: 150, train mse: 25.3244667053, local kl: 0.0 global kl: 0.0 valid mse: 29.9224376678, local kl: 0.0 global kl: 0.0
it: 200, train mse: 55.0799636841, local kl: 0.0 global kl: 0.0 valid mse: 37.4522705078, local kl: 0.0 global kl: 0.0
it: 250, train mse: 19.7182826996, local kl: 0.0 global kl: 0.0 valid mse: 20.1678028107, local kl: 0.0 global kl: 0.0
Saving best model with MSE 20.167803
it: 300, train mse: 17.832529068, local kl: 0.0 global kl: 0.0 valid mse: 28.3981285095, local kl: 0.0 global kl: 0.0
it: 350, train mse: 18.4766139984, local kl: 0.0 global kl: 0.0 valid mse: 23.8278865814, local kl: 0.0 global kl: 0.0
it: 400, train mse: 18.0333061218, local kl: 0.0 global kl: 0.0 valid mse: 17.5554351807, local kl: 0.0 global kl: 0.0
Saving best model with MSE 17.555435
it: 450, train mse: 21.6048126221, local kl: 0.0 global kl: 0.0 valid mse: 23.7403793335, local kl: 0.0 global kl: 0.0
it: 500, train mse: 12.4685487747, local kl: 0.0 global kl: 0.0 valid mse: 18.5810337067, local kl: 0.0 global kl: 0.0
it: 550, train mse: 10.6943292618, local kl: 0.0 global kl: 0.0 valid mse: 21.0635414124, local kl: 0.0 global kl: 0.0
it: 600, train mse: 12.761256218, local kl: 0.0 global kl: 0.0 valid mse: 14.1990861893, local kl: 0.0 global kl: 0.0
Saving best model with MSE 14.199086
it: 650, train mse: 21.7848148346, local kl: 0.0 global kl: 0.0 valid mse: 19.3282756805, local kl: 0.0 global kl: 0.0
it: 700, train mse: 11.0993785858, local kl: 0.0 global kl: 0.0 valid mse: 14.6637878418, local kl: 0.0 global kl: 0.0
it: 750, train mse: 12.2639379501, local kl: 0.0 global kl: 0.0 valid mse: 13.1787528992, local kl: 0.0 global kl: 0.0
Saving best model with MSE 13.178753
it: 800, train mse: 20.0387477875, local kl: 0.0 global kl: 0.0 valid mse: 22.4977970123, local kl: 0.0 global kl: 0.0
it: 850, train mse: 13.2544546127, local kl: 0.0 global kl: 0.0 valid mse: 16.8391914368, local kl: 0.0 global kl: 0.0
it: 900, train mse: 17.5578575134, local kl: 0.0 global kl: 0.0 valid mse: 17.6554012299, local kl: 0.0 global kl: 0.0
it: 950, train mse: 18.2634468079, local kl: 0.0 global kl: 0.0 valid mse: 15.6240139008, local kl: 0.0 global kl: 0.0
it: 1000, train mse: 9.51635360718, local kl: 0.0 global kl: 0.0 valid mse: 12.7266283035, local kl: 0.0 global kl: 0.0
Saving best model with MSE 12.726628
it: 1050, train mse: 11.504573822, local kl: 0.0 global kl: 0.0 valid mse: 15.3926639557, local kl: 0.0 global kl: 0.0
it: 1100, train mse: 14.789689064, local kl: 0.0 global kl: 0.0 valid mse: 16.4801502228, local kl: 0.0 global kl: 0.0
it: 1150, train mse: 9.64725017548, local kl: 0.0 global kl: 0.0 valid mse: 14.1702136993, local kl: 0.0 global kl: 0.0
it: 1200, train mse: 16.9259605408, local kl: 0.0 global kl: 0.0 valid mse: 17.0866680145, local kl: 0.0 global kl: 0.0
it: 1250, train mse: 10.0683765411, local kl: 0.0 global kl: 0.0 valid mse: 15.4766168594, local kl: 0.0 global kl: 0.0
it: 1300, train mse: 13.5223140717, local kl: 0.0 global kl: 0.0 valid mse: 13.8499736786, local kl: 0.0 global kl: 0.0
it: 1350, train mse: 8.4925699234, local kl: 0.0 global kl: 0.0 valid mse: 10.1370582581, local kl: 0.0 global kl: 0.0
Saving best model with MSE 10.137058
it: 1400, train mse: 16.6004943848, local kl: 0.0 global kl: 0.0 valid mse: 12.937046051, local kl: 0.0 global kl: 0.0
it: 1450, train mse: 10.5270080566, local kl: 0.0 global kl: 0.0 valid mse: 14.3343610764, local kl: 0.0 global kl: 0.0
it: 1500, train mse: 9.12664604187, local kl: 0.0 global kl: 0.0 valid mse: 12.9889736176, local kl: 0.0 global kl: 0.0
it: 1550, train mse: 8.74339866638, local kl: 0.0 global kl: 0.0 valid mse: 14.2528152466, local kl: 0.0 global kl: 0.0
it: 1600, train mse: 9.29442882538, local kl: 0.0 global kl: 0.0 valid mse: 14.3698720932, local kl: 0.0 global kl: 0.0
it: 1650, train mse: 5.6758928299, local kl: 0.0 global kl: 0.0 valid mse: 10.1771230698, local kl: 0.0 global kl: 0.0
it: 1700, train mse: 18.8332118988, local kl: 0.0 global kl: 0.0 valid mse: 9.14881420135, local kl: 0.0 global kl: 0.0
Saving best model with MSE 9.148814
it: 1750, train mse: 10.3147468567, local kl: 0.0 global kl: 0.0 valid mse: 9.5696439743, local kl: 0.0 global kl: 0.0
it: 1800, train mse: 10.6403160095, local kl: 0.0 global kl: 0.0 valid mse: 18.7960891724, local kl: 0.0 global kl: 0.0
it: 1850, train mse: 7.81835603714, local kl: 0.0 global kl: 0.0 valid mse: 11.2514820099, local kl: 0.0 global kl: 0.0
it: 1900, train mse: 4.89335393906, local kl: 0.0 global kl: 0.0 valid mse: 10.714184761, local kl: 0.0 global kl: 0.0
it: 1950, train mse: 8.42966365814, local kl: 0.0 global kl: 0.0 valid mse: 8.15223693848, local kl: 0.0 global kl: 0.0
Saving best model with MSE 8.152237
it: 2000, train mse: 5.1705365181, local kl: 0.0 global kl: 0.0 valid mse: 5.96285486221, local kl: 0.0 global kl: 0.0
Saving best model with MSE 5.962855
it: 2050, train mse: 5.49812364578, local kl: 0.0 global kl: 0.0 valid mse: 7.16910648346, local kl: 0.0 global kl: 0.0
it: 2100, train mse: 6.26994180679, local kl: 0.0 global kl: 0.0 valid mse: 12.4936695099, local kl: 0.0 global kl: 0.0
it: 2150, train mse: 15.2274904251, local kl: 0.0 global kl: 0.0 valid mse: 7.26281261444, local kl: 0.0 global kl: 0.0
it: 2200, train mse: 11.2772979736, local kl: 0.0 global kl: 0.0 valid mse: 9.72311592102, local kl: 0.0 global kl: 0.0
it: 2250, train mse: 11.8303308487, local kl: 0.0 global kl: 0.0 valid mse: 14.8373403549, local kl: 0.0 global kl: 0.0
it: 2300, train mse: 4.75330924988, local kl: 0.0 global kl: 0.0 valid mse: 9.56557846069, local kl: 0.0 global kl: 0.0
it: 2350, train mse: 6.53345346451, local kl: 0.0 global kl: 0.0 valid mse: 7.12393903732, local kl: 0.0 global kl: 0.0
it: 2400, train mse: 4.30107116699, local kl: 0.0 global kl: 0.0 valid mse: 7.53502988815, local kl: 0.0 global kl: 0.0
it: 2450, train mse: 8.53674602509, local kl: 0.0 global kl: 0.0 valid mse: 14.6128282547, local kl: 0.0 global kl: 0.0
it: 2500, train mse: 8.40525436401, local kl: 0.0 global kl: 0.0 valid mse: 6.19256687164, local kl: 0.0 global kl: 0.0
it: 2550, train mse: 2.95968437195, local kl: 0.0 global kl: 0.0 valid mse: 9.72466754913, local kl: 0.0 global kl: 0.0
it: 2600, train mse: 6.60181808472, local kl: 0.0 global kl: 0.0 valid mse: 6.81970930099, local kl: 0.0 global kl: 0.0
it: 2650, train mse: 4.69409942627, local kl: 0.0 global kl: 0.0 valid mse: 8.59883213043, local kl: 0.0 global kl: 0.0
it: 2700, train mse: 7.50884771347, local kl: 0.0 global kl: 0.0 valid mse: 6.95531177521, local kl: 0.0 global kl: 0.0
it: 2750, train mse: 4.8391866684, local kl: 0.0 global kl: 0.0 valid mse: 5.65957832336, local kl: 0.0 global kl: 0.0
Saving best model with MSE 5.6595783
it: 2800, train mse: 3.89937472343, local kl: 0.0 global kl: 0.0 valid mse: 7.57889318466, local kl: 0.0 global kl: 0.0
it: 2850, train mse: 8.40572547913, local kl: 0.0 global kl: 0.0 valid mse: 7.75555992126, local kl: 0.0 global kl: 0.0
it: 2900, train mse: 3.79789614677, local kl: 0.0 global kl: 0.0 valid mse: 7.79356193542, local kl: 0.0 global kl: 0.0
it: 2950, train mse: 6.92382955551, local kl: 0.0 global kl: 0.0 valid mse: 8.36308097839, local kl: 0.0 global kl: 0.0
it: 3000, train mse: 14.8439455032, local kl: 0.0 global kl: 0.0 valid mse: 10.9851503372, local kl: 0.0 global kl: 0.0
it: 3050, train mse: 6.81380939484, local kl: 0.0 global kl: 0.0 valid mse: 6.92878770828, local kl: 0.0 global kl: 0.0
it: 3100, train mse: 9.34636211395, local kl: 0.0 global kl: 0.0 valid mse: 14.500787735, local kl: 0.0 global kl: 0.0
it: 3150, train mse: 9.26283359528, local kl: 0.0 global kl: 0.0 valid mse: 6.30068492889, local kl: 0.0 global kl: 0.0
it: 3200, train mse: 4.99939680099, local kl: 0.0 global kl: 0.0 valid mse: 9.01606750488, local kl: 0.0 global kl: 0.0
it: 3250, train mse: 10.3573570251, local kl: 0.0 global kl: 0.0 valid mse: 8.545835495, local kl: 0.0 global kl: 0.0
it: 3300, train mse: 4.73247909546, local kl: 0.0 global kl: 0.0 valid mse: 5.85308980942, local kl: 0.0 global kl: 0.0
it: 3350, train mse: 4.04008674622, local kl: 0.0 global kl: 0.0 valid mse: 9.22214698792, local kl: 0.0 global kl: 0.0
it: 3400, train mse: 6.24930810928, local kl: 0.0 global kl: 0.0 valid mse: 33.7341842651, local kl: 0.0 global kl: 0.0
it: 3450, train mse: 5.26230955124, local kl: 0.0 global kl: 0.0 valid mse: 9.35756778717, local kl: 0.0 global kl: 0.0
it: 3500, train mse: 12.1452112198, local kl: 0.0 global kl: 0.0 valid mse: 10.1861276627, local kl: 0.0 global kl: 0.0
it: 3550, train mse: 6.57857131958, local kl: 0.0 global kl: 0.0 valid mse: 8.76634788513, local kl: 0.0 global kl: 0.0
it: 3600, train mse: 3.73112773895, local kl: 0.0 global kl: 0.0 valid mse: 9.72912693024, local kl: 0.0 global kl: 0.0
it: 3650, train mse: 7.4620218277, local kl: 0.0 global kl: 0.0 valid mse: 8.77695274353, local kl: 0.0 global kl: 0.0
it: 3700, train mse: 4.84601593018, local kl: 0.0 global kl: 0.0 valid mse: 8.81804561615, local kl: 0.0 global kl: 0.0
it: 3750, train mse: 8.16280651093, local kl: 0.0 global kl: 0.0 valid mse: 9.90767192841, local kl: 0.0 global kl: 0.0
it: 3800, train mse: 10.0433502197, local kl: 0.0 global kl: 0.0 valid mse: 13.6299219131, local kl: 0.0 global kl: 0.0
it: 3850, train mse: 3.3756570816, local kl: 0.0 global kl: 0.0 valid mse: 8.17988681793, local kl: 0.0 global kl: 0.0
it: 3900, train mse: 10.0067844391, local kl: 0.0 global kl: 0.0 valid mse: 8.59128952026, local kl: 0.0 global kl: 0.0
it: 3950, train mse: 5.77128887177, local kl: 0.0 global kl: 0.0 valid mse: 10.5845918655, local kl: 0.0 global kl: 0.0
it: 4000, train mse: 4.51383495331, local kl: 0.0 global kl: 0.0 valid mse: 7.068338871, local kl: 0.0 global kl: 0.0
it: 4050, train mse: 4.98091030121, local kl: 0.0 global kl: 0.0 valid mse: 8.5066318512, local kl: 0.0 global kl: 0.0
it: 4100, train mse: 7.20468521118, local kl: 0.0 global kl: 0.0 valid mse: 6.62613773346, local kl: 0.0 global kl: 0.0
it: 4150, train mse: 5.98456907272, local kl: 0.0 global kl: 0.0 valid mse: 8.89725589752, local kl: 0.0 global kl: 0.0
it: 4200, train mse: 6.81597948074, local kl: 0.0 global kl: 0.0 valid mse: 10.3917827606, local kl: 0.0 global kl: 0.0
it: 4250, train mse: 6.9283914566, local kl: 0.0 global kl: 0.0 valid mse: 9.05327510834, local kl: 0.0 global kl: 0.0
it: 4300, train mse: 4.36524629593, local kl: 0.0 global kl: 0.0 valid mse: 8.44386863708, local kl: 0.0 global kl: 0.0
it: 4350, train mse: 6.68196630478, local kl: 0.0 global kl: 0.0 valid mse: 9.33015727997, local kl: 0.0 global kl: 0.0
it: 4400, train mse: 6.17279624939, local kl: 0.0 global kl: 0.0 valid mse: 9.19670295715, local kl: 0.0 global kl: 0.0
it: 4450, train mse: 12.6597633362, local kl: 0.0 global kl: 0.0 valid mse: 10.5631847382, local kl: 0.0 global kl: 0.0
it: 4500, train mse: 9.09540271759, local kl: 0.0 global kl: 0.0 valid mse: 8.79087543488, local kl: 0.0 global kl: 0.0
it: 4550, train mse: 4.73377656937, local kl: 0.0 global kl: 0.0 valid mse: 6.90797185898, local kl: 0.0 global kl: 0.0
it: 4600, train mse: 4.62642383575, local kl: 0.0 global kl: 0.0 valid mse: 8.55520439148, local kl: 0.0 global kl: 0.0
it: 4650, train mse: 3.4700961113, local kl: 0.0 global kl: 0.0 valid mse: 7.51751422882, local kl: 0.0 global kl: 0.0
it: 4700, train mse: 9.23295211792, local kl: 0.0 global kl: 0.0 valid mse: 7.77554512024, local kl: 0.0 global kl: 0.0
it: 4750, train mse: 9.05116653442, local kl: 0.0 global kl: 0.0 valid mse: 12.036233902, local kl: 0.0 global kl: 0.0
it: 4800, train mse: 13.5407152176, local kl: 0.0 global kl: 0.0 valid mse: 12.3419723511, local kl: 0.0 global kl: 0.0
it: 4850, train mse: 5.84860372543, local kl: 0.0 global kl: 0.0 valid mse: 7.18619537354, local kl: 0.0 global kl: 0.0
it: 4900, train mse: 3.20920991898, local kl: 0.0 global kl: 0.0 valid mse: 7.34103727341, local kl: 0.0 global kl: 0.0
it: 4950, train mse: 5.38743114471, local kl: 0.0 global kl: 0.0 valid mse: 7.44813299179, local kl: 0.0 global kl: 0.0
it: 5000, train mse: 6.46536684036, local kl: 0.0 global kl: 0.0 valid mse: 7.49541425705, local kl: 0.0 global kl: 0.0
it: 5050, train mse: 11.1308479309, local kl: 0.0 global kl: 0.0 valid mse: 10.9846305847, local kl: 0.0 global kl: 0.0
it: 5100, train mse: 4.49386548996, local kl: 0.0 global kl: 0.0 valid mse: 8.73712730408, local kl: 0.0 global kl: 0.0
it: 5150, train mse: 6.00617408752, local kl: 0.0 global kl: 0.0 valid mse: 9.01704216003, local kl: 0.0 global kl: 0.0
it: 5200, train mse: 3.69821691513, local kl: 0.0 global kl: 0.0 valid mse: 8.64171886444, local kl: 0.0 global kl: 0.0
it: 5250, train mse: 5.92057132721, local kl: 0.0 global kl: 0.0 valid mse: 8.74329471588, local kl: 0.0 global kl: 0.0
it: 5300, train mse: 5.06529951096, local kl: 0.0 global kl: 0.0 valid mse: 9.09030628204, local kl: 0.0 global kl: 0.0
it: 5350, train mse: 2.99715685844, local kl: 0.0 global kl: 0.0 valid mse: 7.02201032639, local kl: 0.0 global kl: 0.0
it: 5400, train mse: 5.16487693787, local kl: 0.0 global kl: 0.0 valid mse: 7.07584142685, local kl: 0.0 global kl: 0.0
it: 5450, train mse: 7.45202493668, local kl: 0.0 global kl: 0.0 valid mse: 9.08118343353, local kl: 0.0 global kl: 0.0
it: 5500, train mse: 2.25443792343, local kl: 0.0 global kl: 0.0 valid mse: 6.78090143204, local kl: 0.0 global kl: 0.0
it: 5550, train mse: 5.21571016312, local kl: 0.0 global kl: 0.0 valid mse: 7.7740187645, local kl: 0.0 global kl: 0.0
it: 5600, train mse: 4.68369007111, local kl: 0.0 global kl: 0.0 valid mse: 8.0690536499, local kl: 0.0 global kl: 0.0
it: 5650, train mse: 3.80479168892, local kl: 0.0 global kl: 0.0 valid mse: 8.23646068573, local kl: 0.0 global kl: 0.0
it: 5700, train mse: 3.77411437035, local kl: 0.0 global kl: 0.0 valid mse: 6.97674369812, local kl: 0.0 global kl: 0.0
it: 5750, train mse: 9.92887687683, local kl: 0.0 global kl: 0.0 valid mse: 14.2863779068, local kl: 0.0 global kl: 0.0
it: 5800, train mse: 5.3835144043, local kl: 0.0 global kl: 0.0 valid mse: 6.73290014267, local kl: 0.0 global kl: 0.0
it: 5850, train mse: 6.5707154274, local kl: 0.0 global kl: 0.0 valid mse: 6.61964511871, local kl: 0.0 global kl: 0.0
it: 5900, train mse: 3.90734815598, local kl: 0.0 global kl: 0.0 valid mse: 8.63213920593, local kl: 0.0 global kl: 0.0
it: 5950, train mse: 5.53853178024, local kl: 0.0 global kl: 0.0 valid mse: 9.92138004303, local kl: 0.0 global kl: 0.0
it: 6000, train mse: 3.68294143677, local kl: 0.0 global kl: 0.0 valid mse: 7.86432504654, local kl: 0.0 global kl: 0.0
it: 6050, train mse: 6.50118684769, local kl: 0.0 global kl: 0.0 valid mse: 7.08808135986, local kl: 0.0 global kl: 0.0
it: 6100, train mse: 2.67820477486, local kl: 0.0 global kl: 0.0 valid mse: 7.2803735733, local kl: 0.0 global kl: 0.0
it: 6150, train mse: 5.71847248077, local kl: 0.0 global kl: 0.0 valid mse: 7.99496030807, local kl: 0.0 global kl: 0.0
it: 6200, train mse: 8.11187934875, local kl: 0.0 global kl: 0.0 valid mse: 7.48106098175, local kl: 0.0 global kl: 0.0
it: 6250, train mse: 3.74325585365, local kl: 0.0 global kl: 0.0 valid mse: 7.17589521408, local kl: 0.0 global kl: 0.0
it: 6300, train mse: 8.93766689301, local kl: 0.0 global kl: 0.0 valid mse: 8.45359420776, local kl: 0.0 global kl: 0.0
it: 6350, train mse: 2.83439540863, local kl: 0.0 global kl: 0.0 valid mse: 7.32174921036, local kl: 0.0 global kl: 0.0
it: 6400, train mse: 3.30885505676, local kl: 0.0 global kl: 0.0 valid mse: 8.55918312073, local kl: 0.0 global kl: 0.0
it: 6450, train mse: 6.51359319687, local kl: 0.0 global kl: 0.0 valid mse: 7.42704200745, local kl: 0.0 global kl: 0.0
it: 6500, train mse: 5.09794044495, local kl: 0.0 global kl: 0.0 valid mse: 7.59022331238, local kl: 0.0 global kl: 0.0
it: 6550, train mse: 4.79481267929, local kl: 0.0 global kl: 0.0 valid mse: 7.18137645721, local kl: 0.0 global kl: 0.0
it: 6600, train mse: 5.79938745499, local kl: 0.0 global kl: 0.0 valid mse: 6.95350933075, local kl: 0.0 global kl: 0.0
it: 6650, train mse: 4.77199840546, local kl: 0.0 global kl: 0.0 valid mse: 8.89945316315, local kl: 0.0 global kl: 0.0
it: 6700, train mse: 4.92524623871, local kl: 0.0 global kl: 0.0 valid mse: 7.51850700378, local kl: 0.0 global kl: 0.0
it: 6750, train mse: 5.1519947052, local kl: 0.0 global kl: 0.0 valid mse: 7.640645504, local kl: 0.0 global kl: 0.0
it: 6800, train mse: 6.27360773087, local kl: 0.0 global kl: 0.0 valid mse: 7.50144052505, local kl: 0.0 global kl: 0.0
it: 6850, train mse: 5.40743923187, local kl: 0.0 global kl: 0.0 valid mse: 6.82595491409, local kl: 0.0 global kl: 0.0
it: 6900, train mse: 6.17084169388, local kl: 0.0 global kl: 0.0 valid mse: 6.41033554077, local kl: 0.0 global kl: 0.0
it: 6950, train mse: 4.87587594986, local kl: 0.0 global kl: 0.0 valid mse: 6.71524381638, local kl: 0.0 global kl: 0.0
it: 7000, train mse: 6.31531858444, local kl: 0.0 global kl: 0.0 valid mse: 8.15464115143, local kl: 0.0 global kl: 0.0
it: 7050, train mse: 4.50356626511, local kl: 0.0 global kl: 0.0 valid mse: 7.9999704361, local kl: 0.0 global kl: 0.0
it: 7100, train mse: 15.3462190628, local kl: 0.0 global kl: 0.0 valid mse: 9.22660827637, local kl: 0.0 global kl: 0.0
it: 7150, train mse: 8.91292858124, local kl: 0.0 global kl: 0.0 valid mse: 10.7640733719, local kl: 0.0 global kl: 0.0
it: 7200, train mse: 4.02579927444, local kl: 0.0 global kl: 0.0 valid mse: 9.46789073944, local kl: 0.0 global kl: 0.0
it: 7250, train mse: 7.17653417587, local kl: 0.0 global kl: 0.0 valid mse: 7.53027677536, local kl: 0.0 global kl: 0.0
it: 7300, train mse: 11.4841499329, local kl: 0.0 global kl: 0.0 valid mse: 12.4329023361, local kl: 0.0 global kl: 0.0
it: 7350, train mse: 4.13797283173, local kl: 0.0 global kl: 0.0 valid mse: 8.46900844574, local kl: 0.0 global kl: 0.0
it: 7400, train mse: 5.51351976395, local kl: 0.0 global kl: 0.0 valid mse: 7.60308456421, local kl: 0.0 global kl: 0.0
it: 7450, train mse: 4.57208108902, local kl: 0.0 global kl: 0.0 valid mse: 5.99515771866, local kl: 0.0 global kl: 0.0
it: 7500, train mse: 3.1107583046, local kl: 0.0 global kl: 0.0 valid mse: 7.75345945358, local kl: 0.0 global kl: 0.0
it: 7550, train mse: 5.13208627701, local kl: 0.0 global kl: 0.0 valid mse: 9.4518661499, local kl: 0.0 global kl: 0.0
it: 7600, train mse: 4.18381738663, local kl: 0.0 global kl: 0.0 valid mse: 9.03519153595, local kl: 0.0 global kl: 0.0
it: 7650, train mse: 12.787563324, local kl: 0.0 global kl: 0.0 valid mse: 9.58496570587, local kl: 0.0 global kl: 0.0
it: 7700, train mse: 5.00921916962, local kl: 0.0 global kl: 0.0 valid mse: 7.42214679718, local kl: 0.0 global kl: 0.0
it: 7750, train mse: 3.61999249458, local kl: 0.0 global kl: 0.0 valid mse: 7.50241470337, local kl: 0.0 global kl: 0.0
it: 7800, train mse: 6.80306720734, local kl: 0.0 global kl: 0.0 valid mse: 8.26316452026, local kl: 0.0 global kl: 0.0
it: 7850, train mse: 2.29181981087, local kl: 0.0 global kl: 0.0 valid mse: 8.85214424133, local kl: 0.0 global kl: 0.0
it: 7900, train mse: 4.77021932602, local kl: 0.0 global kl: 0.0 valid mse: 7.3787946701, local kl: 0.0 global kl: 0.0
it: 7950, train mse: 4.80023431778, local kl: 0.0 global kl: 0.0 valid mse: 8.70431995392, local kl: 0.0 global kl: 0.0
it: 8000, train mse: 3.49266529083, local kl: 0.0 global kl: 0.0 valid mse: 6.65678119659, local kl: 0.0 global kl: 0.0
it: 8050, train mse: 5.73452997208, local kl: 0.0 global kl: 0.0 valid mse: 6.83925485611, local kl: 0.0 global kl: 0.0
it: 8100, train mse: 3.84844756126, local kl: 0.0 global kl: 0.0 valid mse: 7.88852977753, local kl: 0.0 global kl: 0.0
it: 8150, train mse: 3.90409803391, local kl: 0.0 global kl: 0.0 valid mse: 9.50749397278, local kl: 0.0 global kl: 0.0
it: 8200, train mse: 9.59656238556, local kl: 0.0 global kl: 0.0 valid mse: 15.4808483124, local kl: 0.0 global kl: 0.0
it: 8250, train mse: 4.3342962265, local kl: 0.0 global kl: 0.0 valid mse: 10.4726543427, local kl: 0.0 global kl: 0.0
it: 8300, train mse: 5.53980493546, local kl: 0.0 global kl: 0.0 valid mse: 7.69279909134, local kl: 0.0 global kl: 0.0
it: 8350, train mse: 8.85827541351, local kl: 0.0 global kl: 0.0 valid mse: 8.22350692749, local kl: 0.0 global kl: 0.0
it: 8400, train mse: 5.76292324066, local kl: 0.0 global kl: 0.0 valid mse: 8.1316242218, local kl: 0.0 global kl: 0.0
it: 8450, train mse: 7.45373106003, local kl: 0.0 global kl: 0.0 valid mse: 7.86936712265, local kl: 0.0 global kl: 0.0
it: 8500, train mse: 9.06640148163, local kl: 0.0 global kl: 0.0 valid mse: 7.36895084381, local kl: 0.0 global kl: 0.0
it: 8550, train mse: 12.149848938, local kl: 0.0 global kl: 0.0 valid mse: 9.50592136383, local kl: 0.0 global kl: 0.0
it: 8600, train mse: 3.46420431137, local kl: 0.0 global kl: 0.0 valid mse: 7.24126815796, local kl: 0.0 global kl: 0.0
it: 8650, train mse: 4.292304039, local kl: 0.0 global kl: 0.0 valid mse: 9.26040744781, local kl: 0.0 global kl: 0.0
it: 8700, train mse: 9.71055030823, local kl: 0.0 global kl: 0.0 valid mse: 8.82184696198, local kl: 0.0 global kl: 0.0
it: 8750, train mse: 8.47222423553, local kl: 0.0 global kl: 0.0 valid mse: 9.01480960846, local kl: 0.0 global kl: 0.0
it: 8800, train mse: 9.16973781586, local kl: 0.0 global kl: 0.0 valid mse: 7.96605300903, local kl: 0.0 global kl: 0.0
it: 8850, train mse: 3.52850151062, local kl: 0.0 global kl: 0.0 valid mse: 9.2842874527, local kl: 0.0 global kl: 0.0
it: 8900, train mse: 5.84380292892, local kl: 0.0 global kl: 0.0 valid mse: 7.89853143692, local kl: 0.0 global kl: 0.0
it: 8950, train mse: 3.67911863327, local kl: 0.0 global kl: 0.0 valid mse: 10.7754878998, local kl: 0.0 global kl: 0.0
it: 9000, train mse: 8.42910194397, local kl: 0.0 global kl: 0.0 valid mse: 11.3819723129, local kl: 0.0 global kl: 0.0
it: 9050, train mse: 10.7663040161, local kl: 0.0 global kl: 0.0 valid mse: 10.9134588242, local kl: 0.0 global kl: 0.0
it: 9100, train mse: 7.19693374634, local kl: 0.0 global kl: 0.0 valid mse: 11.2703056335, local kl: 0.0 global kl: 0.0
it: 9150, train mse: 9.39244174957, local kl: 0.0 global kl: 0.0 valid mse: 10.4909248352, local kl: 0.0 global kl: 0.0
it: 9200, train mse: 9.37042045593, local kl: 0.0 global kl: 0.0 valid mse: 9.98014831543, local kl: 0.0 global kl: 0.0
it: 9250, train mse: 6.7469420433, local kl: 0.0 global kl: 0.0 valid mse: 9.88978290558, local kl: 0.0 global kl: 0.0
it: 9300, train mse: 11.1691074371, local kl: 0.0 global kl: 0.0 valid mse: 11.3918952942, local kl: 0.0 global kl: 0.0
it: 9350, train mse: 6.81262731552, local kl: 0.0 global kl: 0.0 valid mse: 11.9581575394, local kl: 0.0 global kl: 0.0
it: 9400, train mse: 6.94437170029, local kl: 0.0 global kl: 0.0 valid mse: 9.65116596222, local kl: 0.0 global kl: 0.0
it: 9450, train mse: 3.52075266838, local kl: 0.0 global kl: 0.0 valid mse: 9.37579536438, local kl: 0.0 global kl: 0.0
it: 9500, train mse: 8.23788261414, local kl: 0.0 global kl: 0.0 valid mse: 10.3723373413, local kl: 0.0 global kl: 0.0
it: 9550, train mse: 5.80496072769, local kl: 0.0 global kl: 0.0 valid mse: 11.5373182297, local kl: 0.0 global kl: 0.0
it: 9600, train mse: 6.65742826462, local kl: 0.0 global kl: 0.0 valid mse: 10.2984313965, local kl: 0.0 global kl: 0.0
it: 9650, train mse: 10.0597848892, local kl: 0.0 global kl: 0.0 valid mse: 10.5743751526, local kl: 0.0 global kl: 0.0
it: 9700, train mse: 6.400370121, local kl: 0.0 global kl: 0.0 valid mse: 9.38891601562, local kl: 0.0 global kl: 0.0
it: 9750, train mse: 3.95457386971, local kl: 0.0 global kl: 0.0 valid mse: 9.85726833344, local kl: 0.0 global kl: 0.0
it: 9800, train mse: 8.6476764679, local kl: 0.0 global kl: 0.0 valid mse: 9.93773555756, local kl: 0.0 global kl: 0.0
it: 9850, train mse: 11.5563812256, local kl: 0.0 global kl: 0.0 valid mse: 9.13677597046, local kl: 0.0 global kl: 0.0
it: 9900, train mse: 12.7001924515, local kl: 0.0 global kl: 0.0 valid mse: 8.21686267853, local kl: 0.0 global kl: 0.0
it: 9950, train mse: 6.21834993362, local kl: 0.0 global kl: 0.0 valid mse: 10.6142158508, local kl: 0.0 global kl: 0.0

In [0]:

    

model_type = 'acns'
x_y_encoder_net_sizes = [HIDDEN_SIZE]*2
global_latent_net_sizes = [HIDDEN_SIZE]*2
local_latent_net_sizes = [HIDDEN_SIZE]*2


model_hparams = tf.contrib.training.HParams(activation=tf.nn.relu,
                                            output_activation=tf.nn.relu,
                                            x_encoder_net_sizes=x_encoder_net_sizes,
                                            x_y_encoder_net_sizes=x_y_encoder_net_sizes,
                                            global_latent_net_sizes=global_latent_net_sizes,
                                            local_latent_net_sizes=local_latent_net_sizes,
                                            decoder_net_sizes=decoder_net_sizes, 
                                            heteroskedastic_net_sizes=heteroskedastic_net_sizes,
                                            att_type=att_type,
                                            att_heads=att_heads,
                                            model_type=model_type,
                                            data_uncertainty=data_uncertainty)
save_path = os.path.join(savedir, 'gnp_' + model_type + '.ckpt')
training_hparams = tf.contrib.training.HParams(lr=0.01,
                                               optimizer=tf.train.RMSPropOptimizer,
                                               num_iterations=10000,
                                               batch_size=10,
                                               num_context=num_context,
                                               num_target=num_target, 
                                               print_every=50,
                                               save_path=save_path,
                                               max_grad_norm=1000.0)

train(data_hparams,
      model_hparams,
      training_hparams)


    

    




it: 0, train mse: 62.2864456177, local kl: 5499.07910156 global kl: 0.0314277894795 valid mse: 79.5546112061, local kl: 9059.21679688 global kl: 0.0417891927063
Saving best model with MSE 79.55461
it: 50, train mse: 59.0019950867, local kl: 0.181515619159 global kl: 0.125239357352 valid mse: 62.9373626709, local kl: 0.0683052390814 global kl: 0.284775853157
Saving best model with MSE 62.937363
it: 100, train mse: 35.8874435425, local kl: 1.09855043888 global kl: 0.0940564647317 valid mse: 45.1217079163, local kl: 0.506570696831 global kl: 0.269403219223
Saving best model with MSE 45.121708
it: 150, train mse: 19.0480823517, local kl: 0.255944609642 global kl: 0.00604878785089 valid mse: 17.5246601105, local kl: 0.260885983706 global kl: 0.0064511760138
Saving best model with MSE 17.52466
it: 200, train mse: 42.1990890503, local kl: 0.22990898788 global kl: 0.00508384406567 valid mse: 41.239238739, local kl: 0.213793352246 global kl: 0.00565440673381
it: 250, train mse: 17.5339736938, local kl: 0.235136836767 global kl: 0.00288481661119 valid mse: 16.8867225647, local kl: 0.227825641632 global kl: 0.000600380997639
Saving best model with MSE 16.886723
it: 300, train mse: 16.5918273926, local kl: 0.208127766848 global kl: 0.000307633017655 valid mse: 29.2487335205, local kl: 0.212506964803 global kl: 0.000373042537831
it: 350, train mse: 17.502614975, local kl: 0.201738864183 global kl: 0.00271587911993 valid mse: 28.1016807556, local kl: 0.195763468742 global kl: 0.00129545794334
it: 400, train mse: 13.7512168884, local kl: 0.167487606406 global kl: 0.000340012280503 valid mse: 14.7038440704, local kl: 0.150560200214 global kl: 0.000437716400484
Saving best model with MSE 14.703844
it: 450, train mse: 16.6416320801, local kl: 0.261064648628 global kl: 0.000371543603251 valid mse: 20.7625579834, local kl: 0.260425567627 global kl: 0.000652839255054
it: 500, train mse: 13.233171463, local kl: 0.189720466733 global kl: 3.51550042978e-05 valid mse: 21.7345294952, local kl: 0.183998435736 global kl: 6.37213961454e-05
it: 550, train mse: 12.563533783, local kl: 0.247618794441 global kl: 0.000859328429215 valid mse: 19.7601108551, local kl: 0.259217262268 global kl: 0.000149289873661
it: 600, train mse: 11.9918498993, local kl: 0.199627891183 global kl: 0.00115686643403 valid mse: 12.9552364349, local kl: 0.205121561885 global kl: 0.00204820442013
Saving best model with MSE 12.955236
it: 650, train mse: 13.5922231674, local kl: 0.223497599363 global kl: 0.000469038583105 valid mse: 16.5404701233, local kl: 0.255981951952 global kl: 0.000832651741803
it: 700, train mse: 13.0148458481, local kl: 0.211920827627 global kl: 0.000268536154181 valid mse: 17.5483703613, local kl: 0.209437698126 global kl: 0.000199613205041
it: 750, train mse: 9.59166526794, local kl: 0.305997431278 global kl: 0.000414068403188 valid mse: 26.0945320129, local kl: 0.293031334877 global kl: 0.000410465756431
it: 800, train mse: 11.4301939011, local kl: 0.231869548559 global kl: 6.16666366113e-05 valid mse: 16.1652774811, local kl: 0.251746952534 global kl: 0.000101796023955
it: 850, train mse: 14.9488830566, local kl: 0.205727055669 global kl: 0.000432380096754 valid mse: 16.7459487915, local kl: 0.191242113709 global kl: 0.000535973464139
it: 900, train mse: 11.7353782654, local kl: 0.229306012392 global kl: 0.00127480563242 valid mse: 15.6939811707, local kl: 0.225796431303 global kl: 0.00314550776966
it: 950, train mse: 12.0039491653, local kl: 0.180436655879 global kl: 0.000431271357229 valid mse: 14.0342378616, local kl: 0.190131500363 global kl: 0.000474397093058
it: 1000, train mse: 9.61172199249, local kl: 0.185292005539 global kl: 5.40146247658e-05 valid mse: 13.8875808716, local kl: 0.197395443916 global kl: 8.89055518201e-05
it: 1050, train mse: 11.7347249985, local kl: 0.196140706539 global kl: 0.00020074872009 valid mse: 18.7991828918, local kl: 0.194784283638 global kl: 0.000424031721195
it: 1100, train mse: 12.83685112, local kl: 0.208857983351 global kl: 0.00015404415899 valid mse: 18.0423240662, local kl: 0.216441959143 global kl: 0.000235080486163
it: 1150, train mse: 14.4554738998, local kl: 0.222754865885 global kl: 0.000702666875441 valid mse: 12.3889408112, local kl: 0.232970684767 global kl: 0.00207480741665
Saving best model with MSE 12.388941
it: 1200, train mse: 14.2174091339, local kl: 0.207763791084 global kl: 9.10029411898e-05 valid mse: 19.2457771301, local kl: 0.19766882062 global kl: 0.000342262646882
it: 1250, train mse: 9.21660327911, local kl: 0.22493326664 global kl: 5.82957472943e-05 valid mse: 13.7225141525, local kl: 0.220113232732 global kl: 0.000147652433952
it: 1300, train mse: 13.7159690857, local kl: 0.212064623833 global kl: 0.000131730834255 valid mse: 11.7260122299, local kl: 0.205979555845 global kl: 0.00013297170517
Saving best model with MSE 11.726012
it: 1350, train mse: 6.14710330963, local kl: 0.155215471983 global kl: 0.000721328251529 valid mse: 12.6710777283, local kl: 0.159771487117 global kl: 0.000384280458093
it: 1400, train mse: 15.6029529572, local kl: 0.208580344915 global kl: 0.000132009168738 valid mse: 11.8472070694, local kl: 0.211565539241 global kl: 0.00032362883212
it: 1450, train mse: 11.4534645081, local kl: 0.194436848164 global kl: 1.98922534764e-05 valid mse: 14.7958450317, local kl: 0.211511328816 global kl: 3.14733115374e-05
it: 1500, train mse: 7.47599220276, local kl: 0.205817267299 global kl: 1.11395356726e-05 valid mse: 10.5089359283, local kl: 0.204343646765 global kl: 4.13962188759e-05
Saving best model with MSE 10.508936
it: 1550, train mse: 7.94155836105, local kl: 0.130705684423 global kl: 0.000151233078213 valid mse: 11.4270973206, local kl: 0.157115235925 global kl: 0.000194414242287
it: 1600, train mse: 9.6622467041, local kl: 0.229495242238 global kl: 3.36931552738e-05 valid mse: 10.4748601913, local kl: 0.235062986612 global kl: 0.000113999354653
Saving best model with MSE 10.47486
it: 1650, train mse: 8.01535415649, local kl: 0.204646706581 global kl: 0.000352506351192 valid mse: 14.7654361725, local kl: 0.210696607828 global kl: 2.67364721367e-05
it: 1700, train mse: 18.0324192047, local kl: 0.195043668151 global kl: 0.000162184544024 valid mse: 11.4501113892, local kl: 0.201332971454 global kl: 0.000245754083153
it: 1750, train mse: 13.1328620911, local kl: 0.21343293786 global kl: 0.00055938801961 valid mse: 11.3735542297, local kl: 0.200544431806 global kl: 0.000690056942403
it: 1800, train mse: 16.8493804932, local kl: 0.199044287205 global kl: 2.01654711418e-05 valid mse: 18.0696773529, local kl: 0.193605318666 global kl: 4.78681686218e-05
it: 1850, train mse: 6.27882146835, local kl: 0.177061200142 global kl: 0.000155749963596 valid mse: 12.0800857544, local kl: 0.184064686298 global kl: 0.000269472424407
it: 1900, train mse: 8.35207939148, local kl: 0.194784596562 global kl: 8.07575634099e-05 valid mse: 12.3058328629, local kl: 0.189502209425 global kl: 5.00819878653e-05
it: 1950, train mse: 10.1546058655, local kl: 0.195674985647 global kl: 0.000114724505693 valid mse: 11.1402826309, local kl: 0.175238698721 global kl: 9.87148596323e-05
it: 2000, train mse: 9.71391677856, local kl: 0.184523075819 global kl: 2.72192755801e-05 valid mse: 19.2698097229, local kl: 0.184009864926 global kl: 0.000174013548531
it: 2050, train mse: 5.85767221451, local kl: 0.198381736875 global kl: 3.89629276469e-05 valid mse: 11.0791387558, local kl: 0.202745318413 global kl: 3.68729888578e-05
it: 2100, train mse: 10.3723154068, local kl: 0.244671225548 global kl: 0.000718769500963 valid mse: 15.2143955231, local kl: 0.297279387712 global kl: 0.000583651650231
it: 2150, train mse: 8.00554084778, local kl: 0.213366776705 global kl: 9.18525838642e-05 valid mse: 13.3416957855, local kl: 0.197001487017 global kl: 9.8868025816e-05
it: 2200, train mse: 8.35381031036, local kl: 0.189758300781 global kl: 0.000452268577646 valid mse: 9.05846309662, local kl: 0.191650912166 global kl: 0.000171812105691
Saving best model with MSE 9.058463
it: 2250, train mse: 6.20619916916, local kl: 0.185409933329 global kl: 1.80296265171e-05 valid mse: 17.478471756, local kl: 0.18143863976 global kl: 1.40241518238e-05
it: 2300, train mse: 5.19033145905, local kl: 0.241165965796 global kl: 9.82982601272e-05 valid mse: 21.1380348206, local kl: 0.180183723569 global kl: 0.000243219983531
it: 2350, train mse: 11.5683698654, local kl: 0.160744294524 global kl: 0.000156785565196 valid mse: 9.36454486847, local kl: 0.171741783619 global kl: 0.000274230435025
it: 2400, train mse: 7.6970500946, local kl: 0.206619784236 global kl: 3.07011578116e-05 valid mse: 10.0996074677, local kl: 0.220478564501 global kl: 3.30720140482e-05
it: 2450, train mse: 7.91366195679, local kl: 0.199934303761 global kl: 7.42882621125e-05 valid mse: 16.9489192963, local kl: 0.201361864805 global kl: 4.90603779326e-05
it: 2500, train mse: 10.496301651, local kl: 0.185610115528 global kl: 4.13609486714e-05 valid mse: 9.14803028107, local kl: 0.191321969032 global kl: 0.000112169269414
it: 2550, train mse: 9.80136394501, local kl: 0.200378030539 global kl: 1.50333726197e-05 valid mse: 10.0073289871, local kl: 0.191331416368 global kl: 1.7887850845e-05
it: 2600, train mse: 5.87964296341, local kl: 0.163324683905 global kl: 0.00011043518316 valid mse: 13.8247919083, local kl: 0.167953327298 global kl: 0.000129121515783
it: 2650, train mse: 5.77988529205, local kl: 0.209986448288 global kl: 4.62380812678e-05 valid mse: 9.15426635742, local kl: 0.216263920069 global kl: 2.4042377845e-05
it: 2700, train mse: 6.55376005173, local kl: 0.187710464001 global kl: 8.40131469886e-05 valid mse: 11.4402856827, local kl: 0.193765625358 global kl: 9.41070029512e-05
it: 2750, train mse: 9.90991783142, local kl: 0.187223300338 global kl: 1.48968974827e-05 valid mse: 8.32199573517, local kl: 0.19471129775 global kl: 2.72327779385e-05
Saving best model with MSE 8.321996
it: 2800, train mse: 4.47887086868, local kl: 0.218562334776 global kl: 1.72740947164e-05 valid mse: 11.2349691391, local kl: 0.217895716429 global kl: 1.25649430629e-05
it: 2850, train mse: 11.7218160629, local kl: 0.209961295128 global kl: 5.70307784074e-05 valid mse: 10.5831851959, local kl: 0.210470199585 global kl: 5.47119707335e-05
it: 2900, train mse: 7.86523532867, local kl: 0.220303058624 global kl: 0.000199033500394 valid mse: 9.62556743622, local kl: 0.220271915197 global kl: 0.000801957794465
it: 2950, train mse: 6.97862243652, local kl: 0.201114565134 global kl: 4.87749857712e-05 valid mse: 9.56313705444, local kl: 0.202073886991 global kl: 8.14594313852e-05
it: 3000, train mse: 8.8740644455, local kl: 0.188528388739 global kl: 1.5110044842e-05 valid mse: 21.4559211731, local kl: 0.189378410578 global kl: 9.69689244812e-06
it: 3050, train mse: 7.03028917313, local kl: 0.19141459465 global kl: 9.59005228651e-06 valid mse: 8.81662940979, local kl: 0.194222420454 global kl: 1.04411028587e-05
it: 3100, train mse: 9.81300926208, local kl: 0.193172439933 global kl: 5.76641177759e-05 valid mse: 14.8049049377, local kl: 0.188325718045 global kl: 7.11048633093e-05
it: 3150, train mse: 6.90439128876, local kl: 0.180392578244 global kl: 1.02057583717e-05 valid mse: 9.03603649139, local kl: 0.179503619671 global kl: 5.34810851605e-06
it: 3200, train mse: 8.45822715759, local kl: 0.198240339756 global kl: 1.9026138034e-05 valid mse: 20.668384552, local kl: 0.187919050455 global kl: 1.504773536e-05
it: 3250, train mse: 4.85704040527, local kl: 0.189047560096 global kl: 0.000424016208854 valid mse: 14.6032829285, local kl: 0.18906763196 global kl: 0.000353548268322
it: 3300, train mse: 7.94846963882, local kl: 0.196227431297 global kl: 3.78010845452e-05 valid mse: 9.80549812317, local kl: 0.188450723886 global kl: 8.73560638865e-05
it: 3350, train mse: 6.47555732727, local kl: 0.192690134048 global kl: 7.86643977335e-06 valid mse: 11.6332654953, local kl: 0.191756099463 global kl: 7.85651172919e-06
it: 3400, train mse: 7.11293506622, local kl: 0.153776749969 global kl: 0.000408967403928 valid mse: 11.2847976685, local kl: 0.155571997166 global kl: 0.000781441107392
it: 3450, train mse: 7.51800918579, local kl: 0.188413590193 global kl: 9.01086241356e-05 valid mse: 9.013215065, local kl: 0.181772083044 global kl: 7.83407158451e-05
it: 3500, train mse: 6.44578456879, local kl: 0.182172894478 global kl: 1.57462636707e-05 valid mse: 14.3486738205, local kl: 0.175641864538 global kl: 4.4226533646e-05
it: 3550, train mse: 7.82734394073, local kl: 0.189527079463 global kl: 1.6315101675e-05 valid mse: 11.4594144821, local kl: 0.1916449368 global kl: 1.39887324622e-05
it: 3600, train mse: 6.94467353821, local kl: 0.183115661144 global kl: 1.638325557e-05 valid mse: 13.2554016113, local kl: 0.190079718828 global kl: 2.39580513153e-05
it: 3650, train mse: 11.7792301178, local kl: 0.175505384803 global kl: 7.91393995314e-06 valid mse: 10.5477638245, local kl: 0.178382307291 global kl: 1.01554642242e-05
it: 3700, train mse: 5.03458738327, local kl: 0.17453250289 global kl: 7.65228833188e-05 valid mse: 11.5814733505, local kl: 0.177529469132 global kl: 0.000203158982913
it: 3750, train mse: 7.57782030106, local kl: 0.184569880366 global kl: 1.35120308187e-05 valid mse: 8.97232818604, local kl: 0.177784264088 global kl: 2.13305938814e-05
it: 3800, train mse: 7.86161708832, local kl: 0.183272495866 global kl: 7.76099477662e-05 valid mse: 11.1136856079, local kl: 0.196970760822 global kl: 6.39793579467e-05
it: 3850, train mse: 3.52167510986, local kl: 0.0763047561049 global kl: 0.00038708280772 valid mse: 13.943526268, local kl: 0.0779080018401 global kl: 0.00169337762054
it: 3900, train mse: 10.1745815277, local kl: 0.172607302666 global kl: 1.35334512379e-05 valid mse: 14.6620168686, local kl: 0.169423088431 global kl: 2.13882740354e-05
it: 3950, train mse: 10.4093580246, local kl: 0.190277531743 global kl: 2.70861091849e-05 valid mse: 12.4551239014, local kl: 0.181545451283 global kl: 2.94832771033e-05
it: 4000, train mse: 8.88862609863, local kl: 0.180435627699 global kl: 1.39763433253e-05 valid mse: 15.7158842087, local kl: 0.188109979033 global kl: 1.55928828462e-05
it: 4050, train mse: 6.1186413765, local kl: 0.184854760766 global kl: 1.85771295946e-05 valid mse: 11.1089572906, local kl: 0.183924734592 global kl: 3.46396336681e-05
it: 4100, train mse: 8.82454109192, local kl: 0.177076593041 global kl: 4.33977875218e-05 valid mse: 7.97175455093, local kl: 0.172404706478 global kl: 0.000297067861538
Saving best model with MSE 7.9717546
it: 4150, train mse: 6.34860420227, local kl: 0.162010878325 global kl: 9.85342994682e-05 valid mse: 12.369395256, local kl: 0.167153149843 global kl: 7.4473915447e-05
it: 4200, train mse: 7.53153419495, local kl: 0.182381600142 global kl: 1.56422811415e-05 valid mse: 16.4092941284, local kl: 0.173273682594 global kl: 3.39125508617e-05
it: 4250, train mse: 6.88647174835, local kl: 0.164635419846 global kl: 0.000161061805557 valid mse: 12.9157037735, local kl: 0.162893265486 global kl: 0.000489870610181
it: 4300, train mse: 5.45740413666, local kl: 0.162686288357 global kl: 7.4758354458e-05 valid mse: 9.62774276733, local kl: 0.176862359047 global kl: 0.000188359554159
it: 4350, train mse: 7.80259227753, local kl: 0.182912126184 global kl: 1.03151833173e-05 valid mse: 12.8808517456, local kl: 0.184578195214 global kl: 2.77859762718e-05
it: 4400, train mse: 9.74481964111, local kl: 0.166723147035 global kl: 5.16272484674e-06 valid mse: 13.1909751892, local kl: 0.169156551361 global kl: 4.06018534704e-06
it: 4450, train mse: 19.7384624481, local kl: 0.182188779116 global kl: 9.24905907596e-05 valid mse: 12.6983003616, local kl: 0.182521477342 global kl: 8.40723987494e-06
it: 4500, train mse: 8.4246339798, local kl: 0.197802737355 global kl: 0.000117323688755 valid mse: 18.5331497192, local kl: 0.202409535646 global kl: 2.94871133519e-05
it: 4550, train mse: 5.71407413483, local kl: 0.18195746839 global kl: 9.32962393563e-06 valid mse: 12.6888341904, local kl: 0.177881971002 global kl: 2.7113215765e-05
it: 4600, train mse: 5.14988183975, local kl: 0.181840986013 global kl: 2.85308351522e-06 valid mse: 10.3307952881, local kl: 0.179967314005 global kl: 2.87236161967e-06
it: 4650, train mse: 4.8168721199, local kl: 0.182570651174 global kl: 0.000195461281692 valid mse: 10.4505548477, local kl: 0.175097644329 global kl: 0.000433001958299
it: 4700, train mse: 8.04023551941, local kl: 0.176999509335 global kl: 8.02979702712e-05 valid mse: 18.5742874146, local kl: 0.173989325762 global kl: 0.000150845720782
it: 4750, train mse: 11.4744195938, local kl: 0.16509744525 global kl: 0.000124083046103 valid mse: 16.5717334747, local kl: 0.162129223347 global kl: 0.000308655638946
it: 4800, train mse: 12.1576948166, local kl: 0.167233258486 global kl: 1.29124837258e-05 valid mse: 11.5347909927, local kl: 0.16774058342 global kl: 0.000129175648908
it: 4850, train mse: 8.38600349426, local kl: 0.153317779303 global kl: 6.67489221087e-05 valid mse: 11.7866687775, local kl: 0.153114706278 global kl: 0.000180639108294
it: 4900, train mse: 3.21910691261, local kl: 0.169876545668 global kl: 0.0010361217428 valid mse: 11.1520633698, local kl: 0.155870139599 global kl: 0.000623930362053
it: 4950, train mse: 6.04536485672, local kl: 0.170680835843 global kl: 4.78718138766e-05 valid mse: 11.3616313934, local kl: 0.169227316976 global kl: 0.000110296030471
it: 5000, train mse: 8.9482793808, local kl: 0.169889390469 global kl: 2.52152585745e-05 valid mse: 11.1270942688, local kl: 0.166797310114 global kl: 2.00250997295e-05
it: 5050, train mse: 11.3058900833, local kl: 0.174637392163 global kl: 0.000699086114764 valid mse: 27.5990848541, local kl: 0.179007783532 global kl: 0.000706872786395
it: 5100, train mse: 7.65875101089, local kl: 0.157522708178 global kl: 0.000124557787785 valid mse: 9.43071365356, local kl: 0.153747051954 global kl: 0.000112856490887
it: 5150, train mse: 9.1182050705, local kl: 0.163491517305 global kl: 0.000154434543219 valid mse: 12.0373334885, local kl: 0.155947715044 global kl: 0.000344854779541
it: 5200, train mse: 9.18964862823, local kl: 0.152857005596 global kl: 0.00143023813143 valid mse: 16.9932384491, local kl: 0.163087978959 global kl: 0.000440021511167
it: 5250, train mse: 6.01309585571, local kl: 0.179077938199 global kl: 2.64677400992e-05 valid mse: 12.1017608643, local kl: 0.174540519714 global kl: 4.30109867011e-05
it: 5300, train mse: 4.77178812027, local kl: 0.165937364101 global kl: 8.13349615782e-05 valid mse: 11.6792173386, local kl: 0.165314465761 global kl: 0.000144883728353
it: 5350, train mse: 7.32762479782, local kl: 0.165659800172 global kl: 7.687287507e-06 valid mse: 8.95659542084, local kl: 0.158450752497 global kl: 4.26353926741e-06
it: 5400, train mse: 6.34603977203, local kl: 0.179711833596 global kl: 5.75849444431e-05 valid mse: 15.5309028625, local kl: 0.189024552703 global kl: 4.22955126851e-05
it: 5450, train mse: 6.24010038376, local kl: 0.167452663183 global kl: 0.000206074590096 valid mse: 13.0466413498, local kl: 0.171764299273 global kl: 0.000857957289554
it: 5500, train mse: 5.77881002426, local kl: 0.15962806344 global kl: 0.000212247905438 valid mse: 11.0627565384, local kl: 0.166175410151 global kl: 0.000126539496705
it: 5550, train mse: 9.18290042877, local kl: 0.168331980705 global kl: 0.000276995648164 valid mse: 12.9016752243, local kl: 0.16469681263 global kl: 0.000320095190546
it: 5600, train mse: 5.83190917969, local kl: 0.157300010324 global kl: 1.11429162644e-05 valid mse: 11.7632465363, local kl: 0.15959303081 global kl: 1.19033020383e-05
it: 5650, train mse: 6.86527013779, local kl: 0.155584096909 global kl: 1.32030288569e-06 valid mse: 15.9009313583, local kl: 0.158354029059 global kl: 3.48681578544e-06
it: 5700, train mse: 7.36887979507, local kl: 0.0904340818524 global kl: 0.000838835723698 valid mse: 9.4283914566, local kl: 0.106506124139 global kl: 0.00265169329941
it: 5750, train mse: 4.43889284134, local kl: 0.16387706995 global kl: 5.67135757592e-05 valid mse: 10.347243309, local kl: 0.163949847221 global kl: 0.000292610551696
it: 5800, train mse: 5.10575914383, local kl: 0.163015350699 global kl: 1.2639776287e-05 valid mse: 11.0404853821, local kl: 0.160298362374 global kl: 3.10055766022e-05
it: 5850, train mse: 7.93599224091, local kl: 0.164032399654 global kl: 9.76514638751e-06 valid mse: 9.24170303345, local kl: 0.160416394472 global kl: 7.27446740711e-06
it: 5900, train mse: 10.3776683807, local kl: 0.155757471919 global kl: 0.000148498860653 valid mse: 11.2604455948, local kl: 0.139265626669 global kl: 0.000131549168145
it: 5950, train mse: 6.99791193008, local kl: 0.155542641878 global kl: 8.37909174152e-05 valid mse: 12.5463619232, local kl: 0.154730439186 global kl: 0.000100867051515
it: 6000, train mse: 2.80449819565, local kl: 0.15799960494 global kl: 6.60302248434e-05 valid mse: 9.62040042877, local kl: 0.161188334227 global kl: 9.63202037383e-05
it: 6050, train mse: 7.58068847656, local kl: 0.17973998189 global kl: 7.63358093536e-06 valid mse: 10.7727108002, local kl: 0.39361128211 global kl: 8.77454294823e-06
it: 6100, train mse: 4.37095022202, local kl: 0.129637137055 global kl: 0.000295173813356 valid mse: 8.87510585785, local kl: 0.130389377475 global kl: 0.000368953828001
it: 6150, train mse: 13.0428972244, local kl: 0.180932596326 global kl: 7.94688530732e-05 valid mse: 14.0383653641, local kl: 0.169941172004 global kl: 0.000142312812386
it: 6200, train mse: 3.86742687225, local kl: 0.191163524985 global kl: 0.000825366179924 valid mse: 13.3492841721, local kl: 0.215146064758 global kl: 0.000655144336633
it: 6250, train mse: 6.14539051056, local kl: 0.150733694434 global kl: 0.000112489506137 valid mse: 8.45994377136, local kl: 0.151587605476 global kl: 0.000117344352475
it: 6300, train mse: 12.2361354828, local kl: 0.171092897654 global kl: 3.71481655748e-05 valid mse: 10.7879056931, local kl: 0.160644814372 global kl: 2.32396578213e-05
it: 6350, train mse: 3.93730115891, local kl: 0.159200921655 global kl: 6.57950158711e-06 valid mse: 12.6560211182, local kl: 0.156636044383 global kl: 5.76627553528e-06
it: 6400, train mse: 5.05245685577, local kl: 0.150464624166 global kl: 2.92798849841e-05 valid mse: 9.31887722015, local kl: 0.154184162617 global kl: 9.22215840546e-05
it: 6450, train mse: 8.65824508667, local kl: 0.139878675342 global kl: 0.000123513309518 valid mse: 9.79777622223, local kl: 0.147403880954 global kl: 0.000321254687151
it: 6500, train mse: 8.79775714874, local kl: 0.158790230751 global kl: 0.0013421253534 valid mse: 11.4710283279, local kl: 0.155175417662 global kl: 0.00147563638166
it: 6550, train mse: 8.3409986496, local kl: 0.157603576779 global kl: 0.000827490352094 valid mse: 12.466884613, local kl: 0.15668669343 global kl: 0.00065808481304
it: 6600, train mse: 10.023443222, local kl: 0.164022535086 global kl: 6.63371611154e-05 valid mse: 11.240776062, local kl: 0.161977499723 global kl: 6.69923174428e-05
it: 6650, train mse: 5.84286308289, local kl: 0.135977804661 global kl: 0.000588231778238 valid mse: 10.3490142822, local kl: 0.139404952526 global kl: 0.00124695664272
it: 6700, train mse: 7.79529237747, local kl: 0.161768972874 global kl: 8.71413067216e-05 valid mse: 12.915512085, local kl: 0.156867772341 global kl: 8.26998875709e-05
it: 6750, train mse: 4.4223613739, local kl: 0.153377771378 global kl: 0.000851777731441 valid mse: 9.33907604218, local kl: 0.150921344757 global kl: 0.000546670460608
it: 6800, train mse: 5.22042131424, local kl: 0.165615186095 global kl: 0.000191618761164 valid mse: 9.78408336639, local kl: 0.165097162127 global kl: 0.000253304227954
it: 6850, train mse: 6.8529291153, local kl: 0.15522107482 global kl: 1.28298388518e-05 valid mse: 12.663277626, local kl: 0.148694962263 global kl: 2.6373356377e-05
it: 6900, train mse: 7.3229227066, local kl: 0.147295638919 global kl: 6.45858090138e-05 valid mse: 14.1147203445, local kl: 0.151011615992 global kl: 2.75388538284e-05
it: 6950, train mse: 8.94937229156, local kl: 0.152104094625 global kl: 1.35031223181e-05 valid mse: 9.4467716217, local kl: 0.149850592017 global kl: 9.75493730948e-06
it: 7000, train mse: 5.87000179291, local kl: 0.156834766269 global kl: 5.02180082549e-05 valid mse: 11.5433082581, local kl: 0.157472625375 global kl: 0.000119592506962
it: 7050, train mse: 7.10248851776, local kl: 0.0582807436585 global kl: 0.0022969651036 valid mse: 9.538189888, local kl: 0.0704146325588 global kl: 0.00558504648507
it: 7100, train mse: 5.58441591263, local kl: 0.153402760625 global kl: 0.000151888103574 valid mse: 8.38754081726, local kl: 0.147980704904 global kl: 0.000126773331431
it: 7150, train mse: 5.36900377274, local kl: 0.142618402839 global kl: 5.35093713552e-05 valid mse: 9.48592090607, local kl: 0.14356918633 global kl: 1.71731317096e-05
it: 7200, train mse: 6.75559043884, local kl: 0.157653778791 global kl: 1.6737723854e-05 valid mse: 10.0738840103, local kl: 0.155689790845 global kl: 1.64196080732e-05
it: 7250, train mse: 13.5063591003, local kl: 0.152521848679 global kl: 0.000124966885778 valid mse: 10.2838106155, local kl: 0.139713719487 global kl: 0.000119913151138
it: 7300, train mse: 12.8346481323, local kl: 0.145553559065 global kl: 4.45383848273e-05 valid mse: 11.1513166428, local kl: 0.143805414438 global kl: 0.000218251603656
it: 7350, train mse: 4.71422243118, local kl: 0.163474798203 global kl: 6.55323538012e-06 valid mse: 12.6680641174, local kl: 0.171087622643 global kl: 1.09966258606e-05
it: 7400, train mse: 8.75109100342, local kl: 0.144853174686 global kl: 2.23541019295e-06 valid mse: 9.26444721222, local kl: 0.146147102118 global kl: 1.71804731508e-06
it: 7450, train mse: 9.65029430389, local kl: 0.0938978567719 global kl: 0.00510507402942 valid mse: 11.9314336777, local kl: 0.105226799846 global kl: 0.0039403392002
it: 7500, train mse: 7.05823898315, local kl: 0.147047132254 global kl: 0.000365128798876 valid mse: 10.3021564484, local kl: 0.151151150465 global kl: 0.000124887374113
it: 7550, train mse: 7.21150064468, local kl: 0.144608169794 global kl: 7.76116285124e-05 valid mse: 18.865694046, local kl: 0.143986493349 global kl: 2.55140130321e-05
it: 7600, train mse: 4.63713693619, local kl: 0.143643081188 global kl: 1.42506814882e-05 valid mse: 11.0520296097, local kl: 0.151505768299 global kl: 1.08079229904e-05
it: 7650, train mse: 6.62421226501, local kl: 0.156915053725 global kl: 0.00103625841439 valid mse: 12.4888029099, local kl: 0.158286601305 global kl: 0.000845342408866
it: 7700, train mse: 7.09777355194, local kl: 0.132181510329 global kl: 5.4767318943e-05 valid mse: 9.40377140045, local kl: 0.13168373704 global kl: 6.69968285365e-05
it: 7750, train mse: 12.8806676865, local kl: 0.14612019062 global kl: 3.73053298972e-05 valid mse: 9.36151218414, local kl: 0.145089387894 global kl: 2.93655411951e-05
it: 7800, train mse: 7.24145555496, local kl: 0.142914623022 global kl: 0.00020567336469 valid mse: 12.9594707489, local kl: 0.14439304173 global kl: 3.78989388992e-05
it: 7850, train mse: 3.79354763031, local kl: 2.51562976837 global kl: 0.00186953204684 valid mse: 10.5988922119, local kl: 0.287441015244 global kl: 0.00743788480759
it: 7900, train mse: 4.15324735641, local kl: 0.151354402304 global kl: 0.000421430449933 valid mse: 12.1838312149, local kl: 0.153726786375 global kl: 0.000245421135332
it: 7950, train mse: 10.1673030853, local kl: 0.163292139769 global kl: 3.27058296534e-05 valid mse: 9.33770942688, local kl: 0.148788899183 global kl: 1.29541858769e-05
it: 8000, train mse: 7.23459005356, local kl: 0.153546407819 global kl: 0.000528972072061 valid mse: 7.39068031311, local kl: 0.156291335821 global kl: 0.00058592192363
Saving best model with MSE 7.3906803
it: 8050, train mse: 9.7094707489, local kl: 0.14072625339 global kl: 5.79909246881e-05 valid mse: 12.6864643097, local kl: 0.138993248343 global kl: 0.000136868009577
it: 8100, train mse: 5.76923704147, local kl: 0.147702544928 global kl: 0.000173183274455 valid mse: 13.2140436172, local kl: 0.151837170124 global kl: 7.43215932744e-05
it: 8150, train mse: 7.11488294601, local kl: 0.135644152761 global kl: 6.98517897035e-06 valid mse: 13.373213768, local kl: 0.125946491957 global kl: 2.61362383753e-05
it: 8200, train mse: 7.3710474968, local kl: 0.154638558626 global kl: 0.000283575151116 valid mse: 12.5485668182, local kl: 0.151715278625 global kl: 0.000126111684949
it: 8250, train mse: 10.2699127197, local kl: 0.175456091762 global kl: 0.00035999622196 valid mse: 13.0149850845, local kl: 0.170132800937 global kl: 0.000276482402114
it: 8300, train mse: 5.36184024811, local kl: 0.142473906279 global kl: 3.6942132283e-05 valid mse: 8.05149459839, local kl: 0.138821065426 global kl: 3.59935293091e-05
it: 8350, train mse: 5.52964353561, local kl: 0.142111957073 global kl: 2.03032104764e-05 valid mse: 10.1354637146, local kl: 0.142030358315 global kl: 4.13071247749e-05
it: 8400, train mse: 7.98436307907, local kl: 0.138236567378 global kl: 6.1674340941e-06 valid mse: 8.74956130981, local kl: 0.134166389704 global kl: 2.62371395365e-06
it: 8450, train mse: 6.79815769196, local kl: 0.145803377032 global kl: 9.25313713651e-07 valid mse: 9.41932201385, local kl: 0.140564382076 global kl: 4.0652221287e-06
it: 8500, train mse: 14.0185270309, local kl: 0.165524974465 global kl: 1.35494428832e-05 valid mse: 9.77439498901, local kl: 0.639012694359 global kl: 5.66056587559e-06
it: 8550, train mse: 7.55923604965, local kl: 0.13653075695 global kl: 3.18540587614e-05 valid mse: 11.3600597382, local kl: 0.131503984332 global kl: 0.000110970227979
it: 8600, train mse: 4.98573637009, local kl: 0.145700931549 global kl: 0.00036049814662 valid mse: 9.89783096313, local kl: 0.132877677679 global kl: 0.000414356123656
it: 8650, train mse: 7.07016086578, local kl: 0.133537605405 global kl: 3.37221608788e-05 valid mse: 8.86915206909, local kl: 0.133188679814 global kl: 4.50488187198e-05
it: 8700, train mse: 7.00452518463, local kl: 0.156055539846 global kl: 1.83983775059e-05 valid mse: 10.1276893616, local kl: 0.157985791564 global kl: 1.82798648893e-05
it: 8750, train mse: 9.24389839172, local kl: 0.141346931458 global kl: 9.24480355025e-06 valid mse: 8.94682884216, local kl: 0.145218014717 global kl: 0.000131562090246
it: 8800, train mse: 6.49127388, local kl: 0.0302859954536 global kl: 0.0011610185029 valid mse: 7.95368528366, local kl: 0.0339957848191 global kl: 0.00266025098972
it: 8850, train mse: 8.74621295929, local kl: 0.137569263577 global kl: 0.00036646676017 valid mse: 9.36471748352, local kl: 0.143545821309 global kl: 0.000584313296713
it: 8900, train mse: 13.8607254028, local kl: 0.136427432299 global kl: 0.000111846973596 valid mse: 11.7811965942, local kl: 0.126517921686 global kl: 0.000128867526655
it: 8950, train mse: 6.49915075302, local kl: 0.136309593916 global kl: 9.65873005043e-06 valid mse: 8.29043292999, local kl: 0.137589931488 global kl: 9.17388297239e-06
it: 9000, train mse: 8.0687713623, local kl: 0.130992978811 global kl: 8.10091805761e-05 valid mse: 9.92141437531, local kl: 0.132230728865 global kl: 5.9854133724e-05
it: 9050, train mse: 7.88770723343, local kl: 0.135272428393 global kl: 8.52267385199e-06 valid mse: 15.9149198532, local kl: 0.140545323491 global kl: 6.64302497171e-05
it: 9100, train mse: 6.70527267456, local kl: 0.145545959473 global kl: 0.000207574033993 valid mse: 12.1707963943, local kl: 0.129893913865 global kl: 0.00045126065379
it: 9150, train mse: 5.13072490692, local kl: 0.134013235569 global kl: 0.000155693647685 valid mse: 9.42814254761, local kl: 0.135368213058 global kl: 0.000322600011714
it: 9200, train mse: 5.32765960693, local kl: 0.150059968233 global kl: 0.000258281186689 valid mse: 12.03997612, local kl: 0.15044516325 global kl: 0.00015540223103
it: 9250, train mse: 4.01918888092, local kl: 0.135800868273 global kl: 4.88570294692e-05 valid mse: 7.91719770432, local kl: 0.13393239677 global kl: 2.43412760028e-05
it: 9300, train mse: 6.97529315948, local kl: 0.130949035287 global kl: 7.87030876381e-06 valid mse: 9.13849544525, local kl: 0.130585312843 global kl: 1.25865144582e-05
it: 9350, train mse: 4.86174201965, local kl: 0.130952328444 global kl: 2.32191196119e-05 valid mse: 10.1411685944, local kl: 0.129562824965 global kl: 2.79472296825e-05
it: 9400, train mse: 3.92789363861, local kl: 0.133890971541 global kl: 9.54261759034e-06 valid mse: 9.51064491272, local kl: 6.71730709076 global kl: 9.54419374466e-06
it: 9450, train mse: 5.2092294693, local kl: 0.0834173411131 global kl: 0.000369794288417 valid mse: 7.56921625137, local kl: 0.10375187546 global kl: 0.000424044119427
it: 9500, train mse: 5.70608043671, local kl: 0.138566657901 global kl: 5.64437868888e-05 valid mse: 11.6481761932, local kl: 0.138447001576 global kl: 2.97133192362e-05
it: 9550, train mse: 4.98551464081, local kl: 0.132258862257 global kl: 4.95701306136e-06 valid mse: 9.68653583527, local kl: 0.133085146546 global kl: 5.51645189262e-06
it: 9600, train mse: 6.08932209015, local kl: 0.135249704123 global kl: 7.74161380832e-05 valid mse: 9.20733070374, local kl: 0.127095490694 global kl: 2.62147168542e-05
it: 9650, train mse: 8.64238262177, local kl: 0.0805865302682 global kl: 0.000476220913697 valid mse: 8.76637268066, local kl: 0.0941435024142 global kl: 0.00127581658307
it: 9700, train mse: 10.1298732758, local kl: 0.141800433397 global kl: 0.000152748514665 valid mse: 9.92366123199, local kl: 0.136656463146 global kl: 0.000167956968653
it: 9750, train mse: 7.0702881813, local kl: 0.141756251454 global kl: 9.54003626248e-05 valid mse: 8.7110710144, local kl: 0.14146348834 global kl: 6.5360654844e-05
it: 9800, train mse: 3.95690584183, local kl: 0.121491596103 global kl: 0.000567470153328 valid mse: 9.23634243011, local kl: 0.131315231323 global kl: 0.000604229571763
it: 9850, train mse: 3.16061973572, local kl: 0.0983759760857 global kl: 0.00740187335759 valid mse: 6.29996156693, local kl: 0.124669887125 global kl: 0.00469342339784
Saving best model with MSE 6.2999616
it: 9900, train mse: 7.9342751503, local kl: 0.132523849607 global kl: 0.000469230028102 valid mse: 8.61402702332, local kl: 0.130768999457 global kl: 0.000539026281331
it: 9950, train mse: 5.58060598373, local kl: 0.1309132725 global kl: 4.53634056612e-05 valid mse: 9.56974506378, local kl: 0.134068414569 global kl: 7.80251066317e-05

In [0]:

    

model_type = 'fully_connected'
x_y_encoder_net_sizes = [HIDDEN_SIZE]*2
global_latent_net_sizes = [HIDDEN_SIZE]*2
local_latent_net_sizes = [HIDDEN_SIZE]*2

model_hparams = tf.contrib.training.HParams(activation=tf.nn.relu,
                                            output_activation=tf.nn.relu,
                                            x_encoder_net_sizes=x_encoder_net_sizes,
                                            x_y_encoder_net_sizes=x_y_encoder_net_sizes,
                                            global_latent_net_sizes=global_latent_net_sizes,
                                            local_latent_net_sizes=local_latent_net_sizes,
                                            decoder_net_sizes=decoder_net_sizes, 
                                            heteroskedastic_net_sizes=heteroskedastic_net_sizes,
                                            att_type=att_type,
                                            att_heads=att_heads,
                                            model_type=model_type,
                                            data_uncertainty=data_uncertainty)
save_path = os.path.join(savedir, 'gnp_' + model_type + '.ckpt')
training_hparams = tf.contrib.training.HParams(lr=0.01,
                                               optimizer=tf.train.RMSPropOptimizer,
                                               num_iterations=10000,
                                               batch_size=10,
                                               num_context=num_context,
                                               num_target=num_target, 
                                               print_every=50,
                                               save_path=save_path,
                                               max_grad_norm=1000.0)

train(data_hparams,
      model_hparams,
      training_hparams)


    

In [0]: