BSD Portsai/py-accelerate 1.12.0_2 MidnightBSD
PyTorch: simple way to launch, train, use PyTorch models on any device
- System
- MidnightBSD
- Repository
- mports
- Maintainer
- [email protected]
- Port added
- Last update
- Commit
b807d69740- License
- Apache-2.0
- Categories
- ai
Description
The accelerate module 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.
Install
Binary package:
mport install ${PYTHON_PKGNAMEPREFIX}accelerateFrom mports:
cd /usr/mports/ai/py-accelerate && make install cleanTo depend on this port in another port's Makefile:
RUN_DEPENDS= ${PYTHON_PKGNAMEPREFIX}accelerate>0:ai/py-accelerateDependencies
As this port's Makefile names them.
Build 1
${PYTHON_PKGNAMEPREFIX}wheel>0: devel/py-wheel
Runtime 7
${PYTHON_PKGNAMEPREFIX}huggingface-hub>=0.21.0: ai/py-huggingface-hub${PYTHON_PKGNAMEPREFIX}numpy>=1.16: math/py-numpy${PYTHON_PKGNAMEPREFIX}packaging>=20.0: devel/py-packaging${PYTHON_PKGNAMEPREFIX}psutil>0: sysutils/py-psutil${PYTHON_PKGNAMEPREFIX}pytorch>=2.0.0: ai/py-pytorch${PYTHON_PKGNAMEPREFIX}safetensors>=0.4.3: ai/py-safetensors${PYTHON_PKGNAMEPREFIX}pyyaml>0: devel/py-pyyaml
Test 12
${PYTHON_PKGNAMEPREFIX}bitsandbytes>=0: ai/py-bitsandbytes${PYTHON_PKGNAMEPREFIX}datasets>=0: ai/py-datasets${PYTHON_PKGNAMEPREFIX}diffusers>=0: ai/py-diffusers${PYTHON_PKGNAMEPREFIX}evaluate>=0: ai/py-evaluate${PYTHON_PKGNAMEPREFIX}parameterized>0: devel/py-parameterized${PYTHON_PKGNAMEPREFIX}scikit-learn>=0: science/py-scikit-learn${PYTHON_PKGNAMEPREFIX}scipy>=0: science/py-scipy${PYTHON_PKGNAMEPREFIX}timm>=0: ai/py-timm${PYTHON_PKGNAMEPREFIX}torchdata>=0.8.0: ai/py-torchdata${PYTHON_PKGNAMEPREFIX}torchvision>0: ai/py-torchvision${PYTHON_PKGNAMEPREFIX}tqdm>0: misc/py-tqdm${PYTHON_PKGNAMEPREFIX}transformers>=0: ai/py-transformers
Build dependencies 1
Run dependencies 7
Required by 1
Configuration
No options to configure.
- USES
- python
- Service scripts
- none
Files
distinfo 1
| File | Size | SHA256 |
|---|---|---|
| accelerate-1.12.0.tar.gz | 398399 | 70988c352feb481887077d2ab845125024b2a137a5090d6d7a32b57d03a45df6 |
Master sites 1
- PYPI
Commit history 1
| Date | Commit | Credits | Log message |
|---|---|---|---|
b807d69740 |
Lucas Holt | Add LangChain AI ports and dependencies AI-Assisted-by: OpenAI Codex <[email protected]>
Signed-off-by: Lucas Holt <[email protected]> |