This page discusses the generally supported model file formats and versions for each tool in the LEIP SDK.

Depending on the operators used in a particular model architecture, you may encounter unforeseen compatibility issues with specific tools.

Model Requirements

Model

Requirement

TensorFlow

  • Compatible with TensorFLow 2.10

PyTorch

  • Compatible with PyTorch 1.10

  • When using LEIP Optimize or Compile, your input model should be traced and saved using torch.jit.save (recommended), or alternatively the whole eager model saved using torch.save (but it must be traceable anyway).

Modules

Module

Execution

LEIP Train

For leip_train command and QGT API (Input and Output)

Keras models from Tensorflow 2.x onwards, in eager execution mode only.

LEIP Optimize

LEIP Compile

The leip_optimize and leip_compile commands support the following input formats.

  • TF (SavedModel)

  • TF (Keras)

  • TF (Graph Proto)

  • TF (ckpt meta)

  • TFLite

  • PyTorch 1.10

LEIP Evaluate

LEIP Run

The leip_evaluate and leip_run commands can currently execute in the following inference frameworks, which are included in the LEIP SDK Docker Images:

  • Tensorflow 2.10

  • Tensorflow Lite 2.10

  • PyTorch 1.10

  • LRE - LEIP Runtime Environment

Supported Input Formats

Model

Description

Tensorflow 2.10

  • TF (ckpt meta)

  • TF (Graph Proto)

  • TF (SavedModel)

  • TF (Keras)

Tensorflow Lite 2.10

Any .tflite file converted by Tensorflow.

PyTorch 1.10

A complete pytorch module (eager or traced) with .pt extension.