mehrdadh commented on a change in pull request #8715: URL: https://github.com/apache/tvm/pull/8715#discussion_r688147235
########## File path: tutorials/micro/micro_autotune.py ########## @@ -0,0 +1,269 @@ +# 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. + +""" +.. _tutorial-micro-autotune: + +Autotuning with micro TVM +========================= +**Author**: `Andrew Reusch <https://github.com/areusch>`_, `Mehrdad Hessar <https://github.com/mehrdadh>` + +This tutorial explains how to autotune a model using the C runtime. +""" + +import argparse +from tvm.contrib import utils + + +PLATFORMS = { + "host": ("host", None), + "qemu_x86": ("host", "qemu_x86"), + "nrf5340dk": ("nrf5340dk", "nrf5340dk_nrf5340_cpuapp"), + "stm32f746xx_disco": ("stm32f746xx", "stm32f746g_disco"), + "stm32f746xx_nucleo": ("stm32f746xx", "nucleo_f746zg"), + "stm32l4r5zi_nucleo": ("stm32l4r5zi", "nucleo_l4r5zi"), +} + + +def main(args): + #################### + # Defining the model + #################### + # + # To begin with, define a model in Keras to be executed on-device. This shouldn't look any different + # from a usual Keras model definition. Let's define a relatively small model here for efficiency's + # sake. + + import tensorflow as tf + from tensorflow import keras + + model = keras.models.Sequential() + model.add(keras.layers.Conv2D(2, 3, input_shape=(16, 16, 3))) + model.build() + + model.summary() + + #################### + # Importing into TVM + #################### + # Now, use `from_keras <https://tvm.apache.org/docs/api/python/relay/frontend.html#tvm.relay.frontend.from_keras>`_ to import the Keras model into TVM. + + import tvm + from tvm import relay + import numpy as np + + inputs = { + i.name.split(":", 2)[0]: [x if x is not None else 1 for x in i.shape.as_list()] + for i in model.inputs + } + inputs = {k: [v[0], v[3], v[1], v[2]] for k, v in inputs.items()} + tvm_model, params = relay.frontend.from_keras(model, inputs, layout="NCHW") + print(tvm_model) Review comment: changed it. -- 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. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
