A very simple tool for situations where optimization with onnx-simplifier would exceed the Protocol Buffers upper file size limit of 2GB, or simply to separate onnx files to any size you want.

Last update: Aug 30, 2022

Related tags

Overview

sne4onnx

A very simple tool for situations where optimization with onnx-simplifier would exceed the Protocol Buffers upper file size limit of 2GB, or simply to separate onnx files to any size you want. Simple Network Extraction for ONNX.

https://github.com/PINTO0309/simple-onnx-processing-tools

Key concept

If INPUT OP name and OUTPUT OP name are specified, the onnx graph within the range of the specified OP name is extracted and .onnx is generated.
Change backend to onnx.utils.Extractor.extract_model so that onnx.ModelProto can be specified as input.

1. Setup

1-1. HostPC

### option
$ echo export PATH="~/.local/bin:$PATH" >> ~/.bashrc \
&& source ~/.bashrc

### run
$ pip install -U onnx \
&& pip install -U sne4onnx

1-2. Docker

### docker pull
$ docker pull pinto0309/sne4onnx:latest

### docker build
$ docker build -t pinto0309/sne4onnx:latest .

### docker run
$ docker run --rm -it -v `pwd`:/workdir pinto0309/sne4onnx:latest
$ cd /workdir

2. CLI Usage

$ sne4onnx -h

usage:
    sne4onnx [-h]
    --input_onnx_file_path INPUT_ONNX_FILE_PATH
    --input_op_names INPUT_OP_NAMES
    --output_op_names OUTPUT_OP_NAMES
    [--output_onnx_file_path OUTPUT_ONNX_FILE_PATH]

optional arguments:
  -h, --help
        show this help message and exit

  --input_onnx_file_path INPUT_ONNX_FILE_PATH
        Input onnx file path.

  --input_op_names INPUT_OP_NAMES
        List of OP names to specify for the input layer of the model.
        Specify the name of the OP, separated by commas.
        e.g. --input_op_names aaa,bbb,ccc

  --output_op_names OUTPUT_OP_NAMES
        List of OP names to specify for the output layer of the model.
        Specify the name of the OP, separated by commas.
        e.g. --output_op_names ddd,eee,fff

  --output_onnx_file_path OUTPUT_ONNX_FILE_PATH
        Output onnx file path. If not specified, extracted.onnx is output.

3. In-script Usage

$ python
>>> from sne4onnx import extraction
>>> help(extraction)

Help on function extraction in module sne4onnx.onnx_network_extraction:

extraction(
    input_op_names: List[str],
    output_op_names: List[str],
    input_onnx_file_path: Union[str, NoneType] = '',
    onnx_graph: Union[onnx.onnx_ml_pb2.ModelProto, NoneType] = None,
    output_onnx_file_path: Union[str, NoneType] = ''
) -> onnx.onnx_ml_pb2.ModelProto

    Parameters
    ----------
    input_op_names: List[str]
        List of OP names to specify for the input layer of the model.
        Specify the name of the OP, separated by commas.
        e.g. ['aaa','bbb','ccc']

    output_op_names: List[str]
        List of OP names to specify for the output layer of the model.
        Specify the name of the OP, separated by commas.
        e.g. ['ddd','eee','fff']

    input_onnx_file_path: Optional[str]
        Input onnx file path.
        Either input_onnx_file_path or onnx_graph must be specified.
        onnx_graph If specified, ignore input_onnx_file_path and process onnx_graph.

    onnx_graph: Optional[onnx.ModelProto]
        onnx.ModelProto.
        Either input_onnx_file_path or onnx_graph must be specified.
        onnx_graph If specified, ignore input_onnx_file_path and process onnx_graph.

    output_onnx_file_path: Optional[str]
        Output onnx file path.
        If not specified, .onnx is not output.
        Default: ''

    Returns
    -------
    extracted_graph: onnx.ModelProto
        Extracted onnx ModelProto

4. CLI Execution

$ sne4onnx \
--input_onnx_file_path input.onnx \
--input_op_names aaa,bbb,ccc \
--output_op_names ddd,eee,fff \
--output_onnx_file_path output.onnx

5. In-script Execution

5-1. Use ONNX files

from sne4onnx import extraction

extracted_graph = extraction(
  input_op_names=['aaa', 'bbb', 'ccc'],
  output_op_names=['ddd', 'eee', 'fff'],
  input_onnx_file_path='input.onnx',
  output_onnx_file_path='output.onnx',
)

5-2. Use onnx.ModelProto

from sne4onnx import extraction

extracted_graph = extraction(
  input_op_names=['aaa', 'bbb', 'ccc'],
  output_op_names=['ddd', 'eee', 'fff'],
  onnx_graph=graph,
  output_onnx_file_path='output.onnx',
)

6. Samples

6-1. Pre-extraction

6-2. Extraction

$ sne4onnx \
--input_onnx_file_path hitnet_sf_finalpass_720x1280.onnx \
--input_op_names 0,1 \
--output_op_names 497,785 \
--output_onnx_file_path hitnet_sf_finalpass_720x960_head.onnx

6-3. Extracted

7. Reference

8. Issues

https://github.com/PINTO0309/simple-onnx-processing-tools/issues

Comments

Significantly faster processing
https://github.com/PINTO0309/simple-onnx-processing-tools/issues/2

Eliminate comma-separated specifications for input_op_names and output_op_names

Add non_verbose option
opened by PINTO0309 0

Releases(1.0.10)

1.0.10(Sep 7, 2022)

Add short form parameter

$ sne4onnx -h

usage:
    sne4onnx [-h]
    -if INPUT_ONNX_FILE_PATH
    -ion INPUT_OP_NAMES
    -oon OUTPUT_OP_NAMES
    [-of OUTPUT_ONNX_FILE_PATH]
    [-n]

optional arguments:
  -h, --help
    show this help message and exit

  -if INPUT_ONNX_FILE_PATH, --input_onnx_file_path INPUT_ONNX_FILE_PATH
    Input onnx file path.

  -ion INPUT_OP_NAMES [INPUT_OP_NAMES ...], --input_op_names INPUT_OP_NAMES [INPUT_OP_NAMES ...]
    List of OP names to specify for the input layer of the model.
    e.g. --input_op_names aaa bbb ccc

  -oon OUTPUT_OP_NAMES [OUTPUT_OP_NAMES ...], --output_op_names OUTPUT_OP_NAMES [OUTPUT_OP_NAMES ...]
    List of OP names to specify for the output layer of the model.
    e.g. --output_op_names ddd eee fff

  -of OUTPUT_ONNX_FILE_PATH, --output_onnx_file_path OUTPUT_ONNX_FILE_PATH
    Output onnx file path. If not specified, extracted.onnx is output.

  -n, --non_verbose
    Do not show all information logs. Only error logs are displayed.

Source code(tar.gz)
Source code(zip)

1.0.9(Jun 7, 2022)
Added process to delete unused input OPs

Source code(tar.gz)
Source code(zip)
1.0.8(May 25, 2022)
Security update

README update

Source code(tar.gz)
Source code(zip)
1.0.7(May 8, 2022)
Fine-tuning of document and logic indentation

Source code(tar.gz)
Source code(zip)
1.0.6(May 8, 2022)
Fix: https://github.com/PINTO0309/simple-onnx-processing-tools/issues/2

Eliminate comma-separated specifications for input_op_names and output_op_names

Add non_verbose option

Source code(tar.gz)
Source code(zip)
1.0.5(Apr 11, 2022)
Support for onnx.ModelProto input. Bug fixes.

Source code(tar.gz)
Source code(zip)
1.0.4(Apr 11, 2022)
Support for onnx.ModelProto input

Source code(tar.gz)
Source code(zip)
1.0.3(Apr 11, 2022)
Support for return onnx.ModelProto

Source code(tar.gz)
Source code(zip)
1.0.2(Apr 10, 2022)
README update

Source code(tar.gz)
Source code(zip)
1.0.1(Apr 7, 2022)
Fixes for README Typo

Source code(tar.gz)
Source code(zip)
1.0.0(Apr 7, 2022)
Initial release

Source code(tar.gz)
Source code(zip)

Parasite: a tool allowing you to compress and decompress files, to reduce their size

🦠 Parasite 🦠 Parasite is a tool written in Python3 allowing you to "compress" any file, reducing its size. ⭐ Features ⭐ + Fast + Good optimization,

30 Nov 25, 2022

🍅🍅🍅YOLOv5-Lite: lighter, faster and easier to deploy. Evolved from yolov5 and the size of model is only 1.7M (int8) and 3.3M (fp16). It can reach 10+ FPS on the Raspberry Pi 4B when the input size is 320×320~

YOLOv5-Lite：lighter, faster and easier to deploy Perform a series of ablation experiments on yolov5 to make it lighter (smaller Flops, lower memory, a

1.5k Jan 5, 2023

ONNX Runtime Web demo is an interactive demo portal showing real use cases running ONNX Runtime Web in VueJS.

ONNX Runtime Web demo is an interactive demo portal showing real use cases running ONNX Runtime Web in VueJS. It currently supports four examples for you to quickly experience the power of ONNX Runtime Web.

58 Dec 18, 2022

A repository that shares tuning results of trained models generated by TensorFlow / Keras. Post-training quantization (Weight Quantization, Integer Quantization, Full Integer Quantization, Float16 Quantization), Quantization-aware training. TensorFlow Lite. OpenVINO. CoreML. TensorFlow.js. TF-TRT. MediaPipe. ONNX. [.tflite,.h5,.pb,saved_model,tfjs,tftrt,mlmodel,.xml/.bin, .onnx]

PINTO_model_zoo Please read the contents of the LICENSE file located directly under each folder before using the model. My model conversion scripts ar

2.4k Jan 5, 2023

A very simple tool for situations where optimization with onnx-simplifier would exceed the Protocol Buffers upper file size limit of 2GB, or simply to separate onnx files to any size you want.

Related tags

Overview

sne4onnx

Key concept

1. Setup

1-1. HostPC

1-2. Docker

2. CLI Usage

3. In-script Usage

4. CLI Execution

5. In-script Execution

5-1. Use ONNX files

5-2. Use onnx.ModelProto

6. Samples

6-1. Pre-extraction

6-2. Extraction

6-3. Extracted

7. Reference

8. Issues

You might also like...

Parasite: a tool allowing you to compress and decompress files, to reduce their size

🍅🍅🍅YOLOv5-Lite: lighter, faster and easier to deploy. Evolved from yolov5 and the size of model is only 1.7M (int8) and 3.3M (fp16). It can reach 10+ FPS on the Raspberry Pi 4B when the input size is 320×320~

ONNX Runtime Web demo is an interactive demo portal showing real use cases running ONNX Runtime Web in VueJS.

An executor that loads ONNX models and embeds documents using the ONNX runtime.

ONNX-GLPDepth - Python scripts for performing monocular depth estimation using the GLPDepth model in ONNX

ONNX-PackNet-SfM: Python scripts for performing monocular depth estimation using the PackNet-SfM model in ONNX

Keras + Hyperopt: A very simple wrapper for convenient hyperparameter optimization

Music source separation is a task to separate audio recordings into individual sources

Comments

Significantly faster processing

Releases(1.0.10)

1.0.10(Sep 7, 2022)

1.0.9(Jun 7, 2022)

1.0.8(May 25, 2022)

1.0.7(May 8, 2022)

1.0.6(May 8, 2022)

1.0.5(Apr 11, 2022)

1.0.4(Apr 11, 2022)

1.0.3(Apr 11, 2022)

1.0.2(Apr 10, 2022)

1.0.1(Apr 7, 2022)

1.0.0(Apr 7, 2022)

Owner

Katsuya Hyodo

This is an official pytorch implementation of Lite-HRNet: A Lightweight High-Resolution Network.

code for paper "Not All Unlabeled Data are Equal: Learning to Weight Data in Semi-supervised Learning" by Zhongzheng Ren*, Raymond A. Yeh*, Alexander G. Schwing.

Pre-trained model, code, and materials from the paper "Impact of Adversarial Examples on Deep Learning Models for Biomedical Image Segmentation" (MICCAI 2019).

Official implementation of "Learning Proposals for Practical Energy-Based Regression", 2021.

Implementation of CVPR'2022:Reconstructing Surfaces for Sparse Point Clouds with On-Surface Priors

Classifying audio using Wavelet transform and deep learning

Anomaly detection analysis and labeling tool, specifically for multiple time series (one time series per category)

Implementation for paper "Towards the Generalization of Contrastive Self-Supervised Learning"

TrackTech: Real-time tracking of subjects and objects on multiple cameras

This is a simple framework to make object detection dataset very quickly

This repository focus on Image Captioning & Video Captioning & Seq-to-Seq Learning & NLP

SlotRefine: A Fast Non-Autoregressive Model forJoint Intent Detection and Slot Filling

wgan, wgan2(improved, gp), infogan, and dcgan implementation in lasagne, keras, pytorch

existing and custom freqtrade strategies supporting the new hyperstrategy format.

Beyond Image to Depth: Improving Depth Prediction using Echoes (CVPR 2021)

Implementation of the Swin Transformer in PyTorch.

Gans-in-action - Companion repository to GANs in Action: Deep learning with Generative Adversarial Networks

Prototypical python implementation of the trust-region algorithm presented in Sequential Linearization Method for Bound-Constrained Mathematical Programs with Complementarity Constraints by Larson, Leyffer, Kirches, and Manns.

Voxel Set Transformer: A Set-to-Set Approach to 3D Object Detection from Point Clouds (CVPR 2022)

Public Models considered for emotion estimation from EEG

code for paper "Not All Unlabeled Data are Equal: Learning to Weight Data in Semi-supervised Learning" by Zhongzheng Ren, Raymond A. Yeh, Alexander G. Schwing.