Accelerate was created for PyTorch users who like to write the training loop of PyTorch models but are reluctant to write and maintain the boilerplate code needed to use multi-GPUs/TPU/fp16. Accelerate abstracts exactly and only the boilerplate code related to multi-GPUs/TPU/fp16 and leaves the rest of your code unchanged.
| Uploaded | Sun Mar 30 23:20:46 2025 |
| md5 checksum | 4bd03cbadfaa6c34853b315bdecbef4b |
| arch | x86_64 |
| build | py39h06a4308_0 |
| depends | numpy >=1.17,<2.0a0, packaging >=20.0, psutil, python >=3.9,<3.10.0a0, pytorch >=1.10.0, pyyaml, setuptools |
| license | Apache-2.0 |
| license_family | Apache |
| md5 | 4bd03cbadfaa6c34853b315bdecbef4b |
| name | huggingface_accelerate |
| platform | linux |
| sha1 | cea0c1153e5e8b9766fd913d888989460645e412 |
| sha256 | 912be3aa61ff2db6b02eab4f95d55df7822fa90cfd656d3b85ab336fb2ddda80 |
| size | 332991 |
| subdir | linux-64 |
| timestamp | 1689792120440 |
| version | 0.21.0 |