huggingface/transformers

Pull Requests Commits

Merge branch 'main' into kimi-k2-native

qgallouedec committed 30 days ago

Verified 8e459cba

Skip the expert-parallel sentinel masking when expert parallelism is off (#48201)

qgallouedec committed 31 days ago

Verified 606e6e8b

Synthetic test assets (#48589)

tarekziade committed 31 days ago

Verified c6947074

Fix deepspeed ci (#48640)

SunMarc committed 31 days ago

Verified 29723434

[MusicgenMelody] Fix conditioning silently dropped at generation step 0 (#48679)

ydshieh committed 31 days ago

Verified 5bed6296

[`MoE`] Fix eager EP (#48653)

vasqu committed 31 days ago

Verified 93ebf6b1

[higgs_audio_v2] Fix: use config.num_codebooks in audio labels tensor (Simple Fix!) (#48560)

b-re-w committed 31 days ago

Verified 455bcd0a

Assert cached decode matches recomputing without a cache (#48289)

IlyasMoutawwakil committed 31 days ago

Verified fd798ba1

Fix `AutoImageProcessor` requiring torchvision when only Pillow is installed (#48616)

blipbyte committed 31 days ago

Verified 1ee31851

Mirror the registered pytree flatten when building automatic dynamic shapes (#48578)

IlyasMoutawwakil committed 31 days ago

Verified c583a3a1

Register activation kernel layers on XPU (#47858)

jiqing-feng committed 31 days ago

Verified cb541c21

[GLM 5.3 Flash] Preserve original names when saving text checkpoints (#48676)

Dovis01 committed 31 days ago

Verified e71d8c53

[fix] Fix GlmOcr integration tests: wrong token IDs and image token decode bug (#48650)

ydshieh committed 32 days ago

Verified 4815a0a6

[tests] Fix NougatModelIntegrationTest: pin artifact revision and update golden values (#48638)

ydshieh committed 32 days ago

Verified bc984f9f

Update dev (#48654)

vasqu committed 32 days ago

Verified 293af3d2

[CI] Deduplicate Nvidia/AMD CI reply comments and add headers (#48655)

ydshieh committed 32 days ago

Verified d7d7fff6

fix incorrect hub tokenizer class (#48641)

itazap committed 32 days ago

Verified d6e4e478

MRoPE continued (#48594)

zucchini-nlp committed 32 days ago

Verified 5b7dcb0d

[fix] Update stale expected strings in HunYuanVL integration tests (#48646)

ydshieh committed 32 days ago

Verified 50bbcc63

[Quantizaiton]support 5/6/7 bits in AutoRound (#48481)

wenhuach21 committed 32 days ago

Verified e8bcd79c

[fix] Update stale golden values and fix expected_logits shape in FlavaForPreTraining integration tests (#48639)

ydshieh committed 32 days ago

Verified 3283d5f7

[tests] Fix integration test golden values broken by fast image processor default (PR #41388) (#48637)

ydshieh committed 32 days ago

Verified 5f47b5aa

Add Fun-ASR-Nano model (#46180)

LauraGPT committed 32 days ago

Verified fc501343

Fix YOLOS device mismatch with device_map="auto" (#46886)

swankystark committed 32 days ago

Verified d9fe823d

[`Generate`] Avoid unconditionally downloading remote hub file (#48620)

vasqu committed 32 days ago

Verified cbc1651a

Honor `shift_labels` in decoder-only LLM/VLM losses (#48493)

qgallouedec committed 33 days ago

Verified bd05a4b2

[KimiLinear] Fix test_cpu_offload: set num_local_experts=4 in model tester (#48624)

ydshieh committed 33 days ago

Verified 338b33dd

[docs] Per-layer config (#48601)

stevhliu committed 33 days ago

Verified bcd59bc4

Fix expert-parallel training: NaN gradients and missing gradient contributions (#48205)

qgallouedec committed 33 days ago

Verified 2ca0705f

Fix GPU memory teardown in CLI serve tests (#48618)

ydshieh committed 33 days ago

Verified 0a959de1

Older