Getting Caffe1 Models and Datasets

This tutorial will help you acquire a variety of models and datasets and put them into places that the other tutorials will expect. We will primarily utilize Caffe's pre-trained models and the scripts that come in that repo. If you don't already have it, then clone it like so:

git clone https://github.com/BVLC/caffe.git

Start by importing the required modules.



In [ ]:

    
import os
print("Required modules imported.")

Now you can setup your root folder for Caffe below if you put it somewhere else. You should only be changing the path that's being set for CAFFE_ROOT.



In [ ]:

    
# You should have checked out original Caffe
# git clone https://github.com/BVLC/caffe.git
# change the CAFFE_ROOT directory below accordingly
CAFFE_ROOT = os.path.expanduser('~/caffe')

if not os.path.exists(CAFFE_ROOT):
    print("Houston, you may have a problem.") 
    print("Did you change CAFFE_ROOT to point to your local Caffe repo?")
    print("Try running: git clone https://github.com/BVLC/caffe.git")

Here's where you pick your model. There are several listed below such as AlexNet, GoogleNet, and Flickr Style. Uncomment the model you want to download.



In [ ]:

    
# Pick a model, and if you don't have it, it will be downloaded
# format below is the model's folder, model's dataset inside that folder
#MODEL = 'bvlc_alexnet', 'bvlc_alexnet.caffemodel' 
#MODEL = 'bvlc_googlenet', 'bvlc_googlenet.caffemodel'
#MODEL = 'finetune_flickr_style', 'finetune_flickr_style.caffemodel'
#MODEL = 'bvlc_reference_caffenet', 'bvlc_reference_caffenet.caffemodel'
MODEL = 'bvlc_reference_rcnn_ilsvrc13', 'bvlc_reference_rcnn_ilsvrc13.caffemodel'

# scripts to download the models reside here (~/caffe/models)
# after downloading the data will exist with the script
CAFFE_MODELS = os.path.join(CAFFE_ROOT, 'models')

# this is like: ~/caffe/models/bvlc_alexnet/deploy.prototxt
CAFFE_MODEL_FILE = os.path.join(CAFFE_MODELS, MODEL[0], 'deploy.prototxt')
# this is like: ~/caffe/models/bvlc_alexnet/bvlc_alexnet.caffemodel
CAFFE_PRETRAINED = os.path.join(CAFFE_MODELS, MODEL[0], MODEL[1])
    
# if the model folder doesn't have the goods, then download it
# this is usually a pretty big file with the .caffemodel extension
if not os.path.exists(CAFFE_PRETRAINED):
    print(CAFFE_PRETRAINED + " not found. Attempting download. Be patient...")
    os.system(
        os.path.join(CAFFE_ROOT, 'scripts/download_model_binary.py') +
        ' ' +
        os.path.join(CAFFE_ROOT, 'models', MODEL[0]))
else:
    print("You already have " + CAFFE_PRETRAINED)

# if the .prototxt file was missing then you're in trouble; cannot continue
if not os.path.exists(CAFFE_MODEL_FILE):
    print("Caffe model file, " + CAFFE_MODEL_FILE + " was not found!")
else:
    print("Now we can test the model!")



In [ ]: