tf_example3

tf_example3 example

Step-by-Step

This example is used to demonstrate how to convert a TensorFlow model with mix precision.

Prerequisite

1. Installation

pip install -r requirements.txt

Note: Validated TensorFlow Version.

2. Download the FP32 model

wget https://storage.googleapis.com/intel-optimized-tensorflow/models/v1_6/mobilenet_v1_1.0_224_frozen.pb

Run

1. Run Command

python test.py --dataset_location=/path/to/imagenet/

2. Introduction

We can get a BF16 model using the Mixed Precision API.

    from neural_compressor.config import MixedPrecisionConfig
    from neural_compressor import mix_precision
    from neural_compressor import Metric
    top1 = Metric(k=1)
    config = MixedPrecisionConfig()
    mix_precision_model = mix_precision.fit(
        model="./mobilenet_v1_1.0_224_frozen.pb",
        config=config,
        eval_dataloader=eval_dataloader,
        eval_metric=top1)