Skip to main content

Use safetensors with ONNX

Project description

onnx-safetensors

CI PyPI - Version PyPI - Python Version Ruff Ruff

ONNX extension for saving to and loading from safetensors 🤗.

Install

pip install --upgrade onnx-safetensors

Usage

Load tensors to an ONNX model

import onnx_safetensors

# Provide your ONNX model here
model: onnx.ModelProto
tensor_file = "model.safetensors"
# Apply weights from the safetensors file to the model
onnx_safetensors.load_file(model, tensor_file)

Save weights to a safetensors file

import onnx_safetensors

# Provide your ONNX model here
model: onnx.ModelProto
tensor_file = "model.safetensors"
# Save weights from to the safetensors file
onnx_safetensors.save_file(model, tensor_file, convert_attributes=True)

# Save weights from to the safetensors file and clear the raw_data fields of the ONNX model to reduce its size
# model will be updated inplace
onnx_safetensors.save_file(model, tensor_file, convert_attributes=True, strip_data=True)

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

onnx-safetensors-0.1.1.tar.gz (7.7 kB view details)

Uploaded Source

Built Distribution

onnx_safetensors-0.1.1-py3-none-any.whl (8.1 kB view details)

Uploaded Python 3

File details

Details for the file onnx-safetensors-0.1.1.tar.gz.

File metadata

  • Download URL: onnx-safetensors-0.1.1.tar.gz
  • Upload date:
  • Size: 7.7 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? Yes
  • Uploaded via: twine/4.0.1 CPython/3.11.4

File hashes

Hashes for onnx-safetensors-0.1.1.tar.gz
Algorithm Hash digest
SHA256 3fea650ac07af6b126b485e555f70b23ed31d8e2b409e8c110e56e29dd0fffcd
MD5 0eae628bae28a967b23283eb552f53de
BLAKE2b-256 e76815730fd72c7ecd52f1a7947af2eb77ecc860f05b0164b2ee944b81bf3235

See more details on using hashes here.

File details

Details for the file onnx_safetensors-0.1.1-py3-none-any.whl.

File metadata

File hashes

Hashes for onnx_safetensors-0.1.1-py3-none-any.whl
Algorithm Hash digest
SHA256 e6456595859678fe0aef9d2c13d900f799cf229f455558cd95c13c82f05ac44f
MD5 9c13be39ef9d94deb7da1ad64eb477ed
BLAKE2b-256 b8f4447ebf56eed958ff01799f53bb49c64f18d75863f2b52e06c91f9d332d89

See more details on using hashes here.

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page