Implemented on three backbones: DeepSeek-v3, Qwen3-MoE, and Mixtral. All three families live in separate directories inside transformers, so any subset can be patched at once and they never interfere.
Build wheels for GraalPy 3.13 by @msimacek in #14136. The previous v2.14.2 release published GraalPy 3.11 wheels, which were not working due to a PyO3 version bump. A CI workflow issue made it so that ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results