Part of https://github.com/astral-sh/uv/issues/4392
We shouldn't link to PyPI, and dropping the workspace-level
documentation link should mean that we get the auto-generated `docs.rs`
links.
## Summary
These are present on the PyTorch index, but only at very old versions.
The PyPI versions are newer, and seemingly these don't need to be built
against CUDA, etc.
Closes https://github.com/astral-sh/uv/issues/16651.
This PR enables `--torch-backend=auto` to automatically detect Intel
GPUs. It follows up on
[#14386](https://github.com/astral-sh/uv/pull/14386).
On Windows, detection is implemented by querying the
`Win32_VideoController` class via the [WMI
crate](https://github.com/ohadravid/wmi-rs/tree/v0.16.0).
Currently, Intel GPUs (XPU) do not depend on specific driver or toolkit
versions to determine which PyTorch wheel to use.
## Summary
If the user explicitly authenticated to pyx, then we attempt to use the
pyx PyTorch URLs; otherwise, we stick to `download.pytorch.org` as the
default.
## Summary
This implements the iOS part of
https://github.com/astral-sh/uv/issues/8029
FYI: @freakboy3742
<!-- What's the purpose of the change? What does it do, and why? -->
## Test Plan
Create a venv with uv and run `cargo run pip install --python-platform
arm64-apple-ios pillow`. Then the iOS binary of pillow should be
installed inside the venv.
## Summary
Packages like `triton` should come from the PyTorch index, but they
don't actually vary across (e.g.) the `cu128` or `cu129` indexes.
Closes https://github.com/astral-sh/uv/issues/15446.
## Test Plan
Validate that the following pins to `cu128`, rather than `cpu`:
```
echo "vllm\ntorch==2.7.1+cu128" | cargo run pip compile --torch-backend=auto --extra-index-url https://wheels.vllm.ai/b2f6c247a9b84556a8ea0e75bb4a2db765ff3315 - --python-platform linux --python-version 3.13 -v
```
## Summary
Add torch cuda 12.9 backend
<!-- What's the purpose of the change? What does it do, and why? -->
## Test Plan
<!-- How was it tested? -->
---------
Signed-off-by: youkaichao <youkaichao@gmail.com>
Co-authored-by: Charlie Marsh <charlie.r.marsh@gmail.com>
## Summary
The PyTorch team publishes ARM Linux wheels for `triton` to the PyTorch
index, which aren't available on PyPI.
## Test Plan
```
echo "torch" | cargo run pip compile - --torch-backend=cu128 --python-platform aarch64-unknown-linux-gnu --python-version 3.13
```
Previously failed because it couldn't find a compatible `triton` wheel.
## Summary
Make the use of `Self` consistent. Mostly done by running `cargo clippy
--fix -- -A clippy::all -W clippy::use_self`.
## Test Plan
<!-- How was it tested? -->
No need.
Following a CI failure in https://github.com/astral-sh/uv/pull/15028,
ensure that all workspace crates are inheriting the MSRV and other
workspace configuration from the workspace root.
## Summary
This PR intends to enable `--torch-backend=auto` to detect Intel GPUs
automatically:
- On Linux, detection is performed using the `lspci` command via
`Display controller` id.
- On Windows, ~~detection is done via a `powershell` query to
`Win32_VideoController`~~. Skip support for now—revisit once a better
solution is available.
Currently, Intel GPUs (XPU) do not rely on specific driver or toolkit
versions to distribute different PyTorch wheels.
## Test Plan
<!-- How was it tested? -->
On Linux:
~~On Windows:
~~
---------
Co-authored-by: Charlie Marsh <charlie.r.marsh@gmail.com>
## Summary
Allows `--torch-backend=auto` to detect AMD GPUs. The approach is fairly
well-documented inline, but I opted for `rocm_agent_enumerator` over
(e.g.) `rocminfo` since it seems to be the recommended approach for
scripting:
https://rocm.docs.amd.com/projects/rocminfo/en/latest/how-to/use-rocm-agent-enumerator.html.
Closes https://github.com/astral-sh/uv/issues/14086.
## Test Plan
```
root@rocm-jupyter-gpu-mi300x1-192gb-devcloud-atl1:~# ./uv-linux-libc-11fb582c5c046bae09766ceddd276dcc5bb41218/uv pip install torch --torch-backend=auto
Resolved 11 packages in 251ms
Prepared 2 packages in 6ms
Installed 11 packages in 257ms
+ filelock==3.18.0
+ fsspec==2025.5.1
+ jinja2==3.1.6
+ markupsafe==3.0.2
+ mpmath==1.3.0
+ networkx==3.5
+ pytorch-triton-rocm==3.3.1
+ setuptools==80.9.0
+ sympy==1.14.0
+ torch==2.7.1+rocm6.3
+ typing-extensions==4.14.0
```
---------
Co-authored-by: Zanie Blue <contact@zanie.dev>
This includes some initial work on adding Pyodide support (issue
#12729). It is enough to get
```
uv pip compile -p /path/to/pyodide --extra-index-url file:/path/to/simple-index
```
to work which should already be quite useful.
## Test Plan
* added a unit test for `pyodide_platform`
* integration tested manually with:
```
cargo run pip install \
-p /home/rchatham/Documents/programming/tmp/pyodide-venv-test/.pyodide-xbuildenv-0.29.3/0.27.4/xbuildenv/pyodide-root/dist/python \
--extra-index-url file:/home/rchatham/Documents/programming/tmp/pyodide-venv-test/.pyodide-xbuildenv-0.29.3/0.27.4/xbuildenv/pyodide-root/package_index \
--index-strategy unsafe-best-match --target blah --no-build \
numpy pydantic
```
---------
Co-authored-by: konsti <konstin@mailbox.org>
Co-authored-by: Zanie Blue <contact@zanie.dev>
Rustfmt introduces a lot of formatting changes in the 2024 edition. To
not break everything all at once, we split out the set of formatting
changes compatible with both the 2021 and 2024 edition by first
formatting with the 2024 style, and then again with the currently used
2021 style.
Notable changes are the formatting of derive macro attributes and lines
with overly long strings and adding trailing semicolons after statements
consistently.
## Summary
If you use `--torch-backend=auto`, we want to avoid selecting (e.g.) a
`+cu124` build of `torch` alongside a `+cu126` build of `torchvision`.
## Summary
It's possible that the PyTorch version the user depends on isn't in the
latest index. These indexes are equally trusted, so we should override
the policy.
Closes#12357.
## Summary
This is a prototype that I'm considering shipping under `--preview`,
based on [`light-the-torch`](https://github.com/pmeier/light-the-torch).
`light-the-torch` patches pip to pull PyTorch packages from the PyTorch
indexes automatically. And, in particular, `light-the-torch` will query
the installed CUDA drivers to determine which indexes are compatible
with your system.
This PR implements equivalent behavior under `--torch-backend auto`,
though you can also set `--torch-backend cpu`, etc. for convenience.
When enabled, the registry client will fetch from the appropriate
PyTorch index when it sees a package from the PyTorch ecosystem (and
ignore any other configured indexes, _unless_ the package is explicitly
pinned to a different index).
Right now, this is only implemented in the `uv pip` CLI, since it
doesn't quite fit into the lockfile APIs given that it relies on feature
detection on the currently-running machine.
## Test Plan
On macOS, you can test this with (e.g.):
```shell
UV_TORCH_BACKEND=auto UV_CUDA_DRIVER_VERSION=450.80.2 cargo run \
pip install torch --python-platform linux --python-version 3.12
```
On a GPU-enabled EC2 machine:
```shell
ubuntu@ip-172-31-47-149:~/uv$ UV_TORCH_BACKEND=auto cargo run pip install torch -v
Finished `dev` profile [unoptimized + debuginfo] target(s) in 0.31s
Running `target/debug/uv pip install torch -v`
DEBUG uv 0.6.6 (e95ca063b 2025-03-14)
DEBUG Searching for default Python interpreter in virtual environments
DEBUG Found `cpython-3.13.0-linux-x86_64-gnu` at `/home/ubuntu/uv/.venv/bin/python3` (virtual environment)
DEBUG Using Python 3.13.0 environment at: .venv
DEBUG Acquired lock for `.venv`
DEBUG At least one requirement is not satisfied: torch
warning: The `--torch-backend` setting is experimental and may change without warning. Pass `--preview` to disable this warning.
DEBUG Detected CUDA driver version from `/sys/module/nvidia/version`: 550.144.3
...
```
