huggingface/transformers

Pull Requests Commits

Revert temporary DeepSpeed CI workflow controls

SunMarc committed 1 day ago

20a57855

Stabilize DeepSpeed dataset and launcher checks

SunMarc committed 1 day ago

c07f873d

Allow DeepSpeed-only training CI against a selected image

SunMarc committed 1 day ago

80456011

trigger workflow

SunMarc committed 1 day ago

782ac95d

SunMarc committed 1 day ago

29c3491b

[docs] Replace examples using pipeline tasks removed in v5 (#50625)

rickygrosvenor-pramanick committed 1 day ago

Verified 0d7a6219

[CI] Fix external links (#50685)

stevhliu committed 1 day ago

Verified 701fdd25

Update A10 expected outputs of FA2 integration tests (#50738)

ydshieh committed 1 day ago

Verified 0a47202c

Add BeamForTokenClassification (#50634)

AymenKallala committed 1 day ago

Verified 029045ca

Revert mark linguistic generated (#50735)

vasqu committed 1 day ago

Verified 3a820e5c

Remove deprecated and unused older mask API file (#50728)

Cyrilvallez committed 1 day ago

Verified ceed5c87

Update dev version (#50725)

vasqu committed 1 day ago

Verified ffefdb9a

FIx audio training examples. (#50672)

ebezzam committed 1 day ago

Verified 6b511ef0

Check Flash Attention support before loading a Flash Attention kernel from the Hub (#50712)

albertvillanova committed 1 day ago

Verified e18fcb1b

🚨🚨 Rename `past_key_values`/`cache_params` to `cache` (#50585)

Cyrilvallez committed 1 day ago

Verified 6a9193de

Fix the MPS RMSNorm hub kernel repo id (#50714)

tarekziade committed 1 day ago

Verified c748111b

CI pass, CI no pass? (#50690)

zucchini-nlp committed 1 day ago

Verified 41f9d399

Add inkling2 (#50682)

guarin committed 2 days ago

Verified 109a7d19

[`Gemma4`] Fix FA slicing (#50681)

vasqu committed 2 days ago

Verified dc3bd32b

Restore the root-level PEFT adapter when resuming from a checkpoint w… (#50368)

ziywang50 committed 2 days ago

Verified acf8004e

Granite for Docling (#49012)

nassarofficial committed 2 days ago

Verified 46b887d9

🚨 Remove the `ep_router` expert-parallel style (#49615)

IlyasMoutawwakil committed 2 days ago

Verified 506ac612

🚨 Run MXFP4 GPT-OSS experts through `ep_dispatch_experts` (#49613)

IlyasMoutawwakil committed 2 days ago

Verified 95e9b5f6

Add dots3-note Preview model support (#47844)

miraclezqc committed 2 days ago

Verified b0e6e7fd

Add IQuestCoder (#50579)

hmellor committed 2 days ago

Verified bb98e254

Move the remaining router-masked MoE plans to `ep_dispatch_experts` (#49607)

IlyasMoutawwakil committed 2 days ago

Verified 290c247f

🚨 `deepgemm_megamoe` EP implementation for DeepGEMM Mega MoE (#49606)

IlyasMoutawwakil committed 2 days ago

Verified f5d8e7df

Add InternLM3 (#50578)

hmellor committed 2 days ago

Verified 1c76d04b

fix(tokenizers): stop treating a missing transformers_version as Mistral (#49733)

cdcailf committed 2 days ago

Verified 9b6e5e20

Load the released A.X-K1 checkpoints without unexpected or missing keys (#50644)

hmellor committed 2 days ago

Verified 0923f62c

Older