GitLab

Enable ktransformers on Intel GPU with local chat backend

update norm cpu kernel

update torch MLA kernel

fix flashinfer float_workspace_buffer small

update default config

support safetensor load, delete architectures argument

Restore CPU offloading capability

modeling_deepseek_v3: fix GenerationMixin warning

fix AMX

ktransformers/utils: fix _get_logits_warper error

Fix GenerationMixin warning introduced by upgrading transformers to 4.51.3.

update-custom-flashinfer

fix-cache-lens

change inject yaml

✨ update ignore

✨ fix typo

fix-client

Fix compile

Update AMX.md

fix-hopper-flashinfer

Update AMX.md