huggingface/transformers

Pull Requests Commits

Merge branch 'main' into cleaning_fsdp_tests

3outeille committed 66 days ago

Verified 86f6b670

Enable push event (to main) for PR CI workflow (#46240)

ydshieh committed 66 days ago

Verified 5298d0ca

fix(hrm_text): Add XPU Expectations for tests (#46214)

kaixuanliu committed 66 days ago

Verified a6cd3b4e

Fix model parallel bugs for Gemma4 (#45817)

kaixuanliu committed 66 days ago

Verified f39b5c8b

Fix bnb 4bit/8bit quantization drop chunked tensors bug (#46210)

kaixuanliu committed 66 days ago

Verified 212863bc

[deepseek_v4] keep hc_head / sinks / position_bias in fp32 (#46198)

ArthurZucker committed 66 days ago

Verified 9ded3dbb

3outeille committed 67 days ago

30e02a64

reuse some code

3outeille committed 67 days ago

bd240ca0

Merge remote-tracking branch 'origin/main' into cleaning_fsdp_tests

3outeille committed 67 days ago

aab2ed5d

init FSDP through from_pretrained (#46102)

3outeille committed 67 days ago

Verified 0588858f

separate cpu ofloadding and mixed precision tests into mocking

3outeille committed 67 days ago

40bd015c

Fix FSDP2 and distributed checkpointing imports for older PyTorch versions (#46141)

ryota-komatsu committed 67 days ago

Verified 634500b9

Fix Gemma4 Array Mask Indexing (#46203)

petecao committed 67 days ago

Verified 5a4b70f8

utils: handle flash_attn missing from importlib packages_distributions without crashing (#45524)

SAY-5 committed 67 days ago

Verified 02af5bfa

fix model parallel device mismatch issue in `create_bidirectional_mask` (#46221)

kaixuanliu committed 67 days ago

Verified 710742f1

Fix a regression in encoder-decoder generation cache initialization (#46111)

kaixuanliu committed 67 days ago

Verified 90e3c4fa

[AMD CI] revert AMD mi325 hf-workflows ref from SHA back to @main (#46213)

Abdennacer-Badaoui committed 67 days ago

Verified 6db6fdb9

Trainer.compute_loss: fix loss over-counting under TP and EP-as-TP (#45994)

AmineDiro committed 67 days ago

Verified b131ec14

[GLM-4.6V] Update with GLM-GA Processor (#46184)

zRzRzRzRzRzRzR committed 67 days ago

Verified 89c0c2ed

🚨 [ALM] Add base model without head (#45534)

eustlb committed 67 days ago

Verified 027d1a97

update xpu expectation for falcon mamba (#46086)

sywangyi committed 67 days ago

Verified 08692e3c

chore: enable Dependabot weekly GitHub Actions bumps (#46157)

hf-dependantbot-rollout[bot] committed 67 days ago

Verified 9b1bf032

Fix Gemma4 use_bidirectional_attention="all" mask behavior (#46079)

oliverholworthy committed 67 days ago

Verified ceb7ba69

Fix loading with only 1 device or distributed config (#46197)

Cyrilvallez committed 67 days ago

Verified 4a2e3652

Fix TypeError on list-typed ignore_keys_at_rope_validation in RoPE config (#46142)

Charly21r committed 68 days ago

Verified ece1ea06

Support XPU autocast dtype fallback for FlashAttention (#46199)

YangKai0616 committed 68 days ago

Verified 7bc093b7

Fix path traversal when saving named chat templates (#46191)

LinZiyuu committed 68 days ago

Verified eaaaf849

Fix is_last off-by-one in MaskGenerationPipeline for partial batches (#46136)

J3r3myPerera committed 68 days ago

Verified 47949d3a

Fix wrong variable in check_model_type isinstance check (#46080)

SebTardif committed 68 days ago

Verified 118ddfb6

Enable passing kwargs through RoFormer models (#46171)

ir2718 committed 68 days ago

Verified 0b2da85d

Older