intel-gpu: pin unsloth_zoo>=2026.5.2 (fixes #5494)#5499
Merged
Conversation
Fixes #5494: installing any intel-gpu-torch* extra without also pulling `huggingface` or `colab-new` lets the resolver silently fall back to a stale unsloth_zoo (2026.3.6 in the original report) because no version floor is enforced on `unsloth_zoo[intelgpu]` in those blocks. unsloth_zoo 2026.5.2 (just released) also relaxes its own torch upper bound from <2.11.0 to <2.13.0, which is what unblocks the resolver for the intel-gpu-torch2110 and intel-gpu-torch2120 extras shipped in #5484. Adding the floor in `huggingfacenotorch` propagates it to every extra that includes the HF-without-torch base: amd, huggingface, and all nine intelgputorch* blocks. Single line, single source of truth. Requires unsloth_zoo 2026.5.2 to be on PyPI for end-to-end resolution.
![gemini-code-assist[bot]](https://avatars.githubusercontent.com/in/956858?s=60&v=4){kind=small}
rhsCZ
pushed a commit
to rhsCZ/unsloth
that referenced
this pull request
May 18, 2026
The merge of main into ci/frozen-lockfile-installs in 888347d resolved pyproject.toml against the pre-merge branch state, silently dropping the intel-gpu xpu extras that landed on main in f7069b2 and were pinned in e20bbef (unslothai#5499) after this branch was created. None of the lockfile-pin feature commits touch pyproject.toml; this restore reverts only the unintended deletion and keeps the supply-chain hardening intact. Restored extras: intelgputorch271 + intel-gpu-torch271 intelgputorch291 + intel-gpu-torch291 intelgputorch2110 + intel-gpu-torch2110 intelgputorch2120 + intel-gpu-torch2120 Restored the triton-xpu 3.6.0 pin section inside intelgputorch210.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Fixes #5494. Installing any
intel-gpu-torch*extra currently lets the resolver fall back to a staleunsloth_zoo(the reporter saw2026.3.6from.[intelgputorch2120]) because no version floor is enforced onunsloth_zoo[intelgpu]in those blocks, and theunslothpackage itself has nounsloth_zooconstraint in its base dependencies.This PR adds a single
unsloth_zoo>=2026.5.2line at the top ofhuggingfacenotorch, which every intel-gpu-torch* extra, plusamdandhuggingface, already includes. One source of truth, propagates to every affected path.Why 2026.5.2
unsloth_zoo 2026.5.2 (unslothai/unsloth-zoo#658, merged today) relaxes its own torch upper cap from
<2.11.0to<2.13.0. Without that release, pinningunsloth_zoo>=2026.5.1would make the resolver fail outright on the newintelgputorch2110/intelgputorch2120extras shipped in #5484 (because their pinnedtorch-2.11.0+xpu/torch-2.12.0+xpuwheels violate unsloth_zoo 2026.5.1'storch<2.11.0cap).The chain:
Test plan
Once 2026.5.2 is on PyPI, run
uv pip compile pyproject.toml --extra <EXTRA> --python-version 3.11 --python-platform x86_64-unknown-linux-gnufor every intel-gpu-torch* extra and confirm:unsloth_zoo==2026.5.2(or newer), not a stale older release.Draft until 2026.5.2 lands on PyPI so CI can run on a satisfiable graph.
Closes #5494.