alihan/light-the-torch
1
0
Fork 0
mirror of https://github.com/pmeier/light-the-torch.git synced 2024-09-08 23:29:28 +03:00
Code Issues Packages Projects Releases Wiki Activity

make it more clear that ltt helps install *all* PyTorch distributions (#103)

Browse Source
This commit is contained in:
Philip Meier
2022-10-24 10:36:14 +02:00
committed by GitHub
parent a296a8aeb3
commit 8e37c9bdd2
1 changed files with 15 additions and 12 deletions
Download Patch File Download Diff File Expand all files Collapse all files

27
README.md
View File

@@ -5,9 +5,10 @@
[![Code coverage via codecov.io](https://codecov.io/gh/pmeier/light-the-torch/branch/main/graph/badge.svg)](https://codecov.io/gh/pmeier/light-the-torch)
`light-the-torch` is a small utility that wraps `pip` to ease the installation process
for PyTorch distributions and third-party packages that depend on them. It auto-detects
compatible CUDA versions from the local setup and installs the correct PyTorch binaries
without user interference.
for PyTorch distributions like `torch`, `torchvision`, `torchaudio`, and so on as well
as third-party packages that depend on them. It auto-detects compatible CUDA versions
from the local setup and installs the correct PyTorch binaries without user
interference.
- [Why do I need it?](#why-do-i-need-it)
- [How do I install it?](#how-do-i-install-it)
@@ -17,8 +18,8 @@ without user interference.
## Why do I need it?
PyTorch distributions are fully `pip install`'able, but PyPI, the default `pip` search
index, has some limitations:
PyTorch distributions like `torch`, `torchvision`, `torchaudio`, and so on are fully
`pip install`'able, but PyPI, the default `pip` search index, has some limitations:
1. PyPI regularly only allows binaries up to a size of
[approximately 60 MB](https://github.com/pypa/packaging-problems/issues/86). One can
@@ -34,17 +35,19 @@ index, has some limitations:
hand your NVIDIA driver version simply doesn't support the CUDA version the binary
was compiled with, you can't use any of the GPU features.
To overcome this, PyTorch also hosts _most_ binaries
[on their own package indices](https://download.pytorch.org/whl). Some distributions are
not compiled against a specific computation backend and thus hosting them on PyPI is
sufficient since they work in every environment. To access PyTorch's package indices,
you can still use `pip install`, but some
To overcome this, PyTorch also hosts _most_[^1] binaries
[on their own package indices](https://download.pytorch.org/whl). To access PyTorch's
package indices, you can still use `pip install`, but some
[additional options](https://pytorch.org/get-started/locally/) are needed:
```shell
pip install torch --extra-index-url https://download.pytorch.org/whl/cu113
```
[^1]:
Some distributions are not compiled against a specific computation backend and thus
hosting them on PyPI is sufficient since they work in every environment.
While this is certainly an improvement, it still has a few downsides:
1. You need to know what computation backend, e.g. CUDA 11.3 (`cu113`), is supported on
@@ -100,7 +103,7 @@ In fact, `ltt` is `pip` with a few added options:
the computation backend you want to use:
```shell
ltt install --pytorch-computation-backend=cu102 torch
ltt install --pytorch-computation-backend=cu102 torch torchvision torchaudio
```
Borrowing from the mutex packages that PyTorch provides for `conda` installations,
@@ -114,7 +117,7 @@ In fact, `ltt` is `pip` with a few added options:
nightly, test, or LTS channels pass the `--pytorch-channel` option:
```shell
ltt install --pytorch-channel=nightly torch
ltt install --pytorch-channel=nightly torch torchvision torchaudio
```
If `--pytorch-channel` is not passed, using `pip`'s builtin `--pre` option will