huggingface/transformers

Pull Requests Commits

Fix inaccurate train_tokens_per_second when resuming from checkpoint (#41156)

lilin-1 committed 1 year ago

Verified a3fa1d39

[v5] Remove `model_parallel` deprecated feature (#41166)

SunMarc committed 1 year ago

Verified ad74fba0

More typing fixes (#41102)

cyyever committed 1 year ago

Verified 38a08b6e

[tests] `CausalLMTester` automatically infers other test classes from `base_model_class` 🐛 🔫 (#41066)

gante committed 1 year ago

Verified 4fade114

[XPU] Add MXFP4 support for XPU (#41117)

YangKai0616 committed 1 year ago

Verified cdba28c3

CI Runners - move amd runners mi355 and 325 to runner group (#41193)

glegendre01 committed 1 year ago

Verified 2dcb20dc

Modernbert fix (#41056)

remi-or committed 1 year ago

Verified d0d574b1

handle flash slow tests (#41072)

itazap committed 1 year ago

Verified 071eb533

Enable fa in amd docker (#41069)

remi-or committed 1 year ago

Verified 50d2448a

Remove data from examples (#41168)

LysandreJik committed 1 year ago

Verified 10f6891f

Fix flash-attn for paged_attention when no kernels (#41078)

remi-or committed 1 year ago

Verified 97ca0b47

Improve `add_dates` script (#41167)

yonigozlan committed 1 year ago

Verified 53838edd

Add language specifiers to code blocks of markdown files (#41114)

cyyever committed 1 year ago

Verified 449533af

Force new vision models addition to include a fast image processor (#40802)

yonigozlan committed 1 year ago

Verified e691f844

Simplify and improve model loading logic (#41103)

Cyrilvallez committed 1 year ago

Verified e54bb62a

Fix format of compressed_tensors.md (#41155)

cyyever committed 1 year ago

Verified 6dc9ed87

Add Parakeet (#39062)

nithinraok committed 1 year ago

Verified a579de7f

extend gemma3n integration ut cases on XPU (#41071)

yao-matrix committed 1 year ago

Verified 1dd22a23

Fix single quotes in markdown (#41154)

cyyever committed 1 year ago

Verified 05fb90c9

Adapt and test huggingface_hub v1.0.0 (#40889)

Wauplin committed 1 year ago

Verified 44682e71

Fix: align Qwen2.5-VL inference rope index with training by passing s… (#41153)

Xqle committed 1 year ago

Verified 750dd2a4

Fix loading logic flaw with regards to unexpected and missing keys (#40850)

LysandreJik committed 1 year ago

Verified 7258ea44

dummy commit (#41133)

ydshieh committed 1 year ago

Verified 2c4caa19

Fixed loading LongT5 from legacy checkpoints (#40724)

Szustarol committed 1 year ago

Verified 6d187592

Fixed MXFP4 model storage issue (#41118)

YangKai0616 committed 1 year ago

Verified 3ca43d34

🚨Refactor: Update text2text generation pipelines to use max_new_tokens… (#40928)

lilin-1 committed 1 year ago

Verified b33cb700

Remove self-assignment (#41062)

cyyever committed 1 year ago

Verified b0c7034d

Fix broken `` expressions in markdown files (#41113)

cyyever committed 1 year ago

Verified 04a0bb56

Fix the error where a keyword argument appearing before *args (#41099)

cyyever committed 1 year ago

Verified 071c7b14

[Qwen3-next] Fix dimension mismatch in torch_chunk_gated_delta_rule and torch_recurrent_gated_delta_rule (#40963) (#41036)

notkisk committed 1 year ago

Verified 80f20e0f

Newer Older