FrozenGene commented on a change in pull request #6206: URL: https://github.com/apache/incubator-tvm/pull/6206#discussion_r465441677
########## File path: tutorials/frontend/from_caffe.py ########## @@ -0,0 +1,144 @@ +# Licensed to the Apache Software Foundation (ASF) under one +# or more contributor license agreements. See the NOTICE file +# distributed with this work for additional information +# regarding copyright ownership. The ASF licenses this file +# to you under the Apache License, Version 2.0 (the +# "License"); you may not use this file except in compliance +# with the License. You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, +# software distributed under the License is distributed on an +# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY +# KIND, either express or implied. See the License for the +# specific language governing permissions and limitations +# under the License. +""" +Compile Caffe Models +==================== +**Author**: `Chen Feng <https://github.com/fernchen>`_ + +This article is an introductory tutorial to deploy Caffe models with relay. + +To get started, Caffe package needs to be installed. + +In order to install Caffe, please refer to https://caffe.berkeleyvision.org/installation.html. + +If your os is ubuntu version >= 17,04, pre-compiled caffe can be install by: + +.. code-block:: bash + + # install caffe + sudo apt install caffe-cpu + + +Below you can file an example on how to compile Caffe model using TVM. +""" +###################################################################### +# Download pretrained Caffe model +# ------------------------------- +from tvm.contrib.download import download_testdata + +proto_file_url = ("https://github.com/shicai/MobileNet-Caffe/raw/"; + "master/mobilenet_v2_deploy.prototxt") +blob_file_url = ("https://github.com/shicai/MobileNet-Caffe/blob/"; + "master/mobilenet_v2.caffemodel?raw=true") + +proto_file = download_testdata(proto_file_url, "mobilenetv2.prototxt", module="model") +blob_file = download_testdata(blob_file_url, "mobilenetv2.caffemodel", module="model") + +###################################################################### +# Load a test image +# ----------------- +# A single cat dominates the examples! +from PIL import Image +import numpy as np +from matplotlib import pyplot as plt + +image_url = "https://github.com/dmlc/mxnet.js/blob/master/data/cat.png?raw=true"; +image_path = download_testdata(image_url, 'cat.png', module='data') +resized_image = Image.open(image_path).resize((224, 224)) +plt.imshow(resized_image) +plt.show() +image_data = np.asarray(resized_image).astype(np.float32) + +# Preprocess image +image_data -= np.array([103.939, 116.779, 123.68], dtype=np.float32) +image_data /= 58.8 + +# RGB2BGR for Caffe model +image_data = image_data[..., ::-1] + +# Add a dimension to the image and transpose it so that we have NCHW format layout +image_data = np.expand_dims(image_data, axis=0) +image_data = np.transpose(image_data, (0, 3, 1, 2)) + +###################################################################### +# Compile the model with relay +# ---------------------------- +from google.protobuf import text_format +import caffe +from caffe.proto import caffe_pb2 as pb + +# Get Caffe model from buffer +init_net = pb.NetParameter() +predict_net = pb.NetParameter() +with open(proto_file, 'r') as f: + text_format.Merge(f.read(), predict_net) +with open(blob_file, 'rb') as f: + init_net.ParseFromString(f.read()) +# Caffe model input layer name, shape and dtype +shape_dict = {'data': image_data.shape} +dtype_dict = {'data': 'float32'} + +import tvm +from tvm import relay +mod, params = relay.frontend.from_caffe(init_net, predict_net, shape_dict, dtype_dict) + +# Build the module against to x86 CPU +target = 'llvm' +with tvm.transform.PassContext(opt_level=3): + lib = relay.build(mod, target=target,params=params) + +###################################################################### +# Execute on TVM +# -------------- +from tvm.contrib import graph_runtime + +# Create a runtime executor module +ctx = tvm.cpu(0) +m = graph_runtime.GraphModule(lib['default'](ctx)) + +# Feed input data +m.set_input('data', tvm.nd.array(image_data)) + +# Feed related params +m.set_input(**params) Review comment: we don't need this in the new model based interface. Please remove this line. ---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org
![https://github.com/dmlc/mxnet.js/blob/master/data/cat.png?raw=true"](https://github.com/dmlc/mxnet.js/blob/master/data/cat.png?raw=true"){kind=link}