huggingface/transformers

Pull Requests Commits

Allow bi-directional attention for all models (#43705)

Cyrilvallez committed 49 days ago

Verified 83bce8d8

Docs: fix Training step by removing tokenizer from trainer initialization (#43733)

nesjett committed 49 days ago

Verified 452c179e

Fix scheduler initialization order (#43711)

SunMarc committed 49 days ago

Verified 1744f8f3

Fix accelerate integration import (#43732)

SunMarc committed 49 days ago

Verified 8dce3100

🚨 T5Gemma2 model structure (#43633)

zucchini-nlp committed 49 days ago

Verified d75266f1

Update torch minimum version to 2.4 (#41307)

cyyever committed 49 days ago

Verified a276a763

Fix dtype in image-text-to-text pipe (#43731)

zucchini-nlp committed 49 days ago

Verified d95672f8

Preventing initialization of siglip's lecun_normal_, default_flax_embed_init in ZeRO3 (#43574)

jp1924 committed 49 days ago

Verified 225254c1

🚨 Generation cache preparation (#43679)

zucchini-nlp committed 49 days ago

Verified 1687954f

fix: AttributeError for Qwen3_omni_moe (#43593)

Vallabh-1504 committed 49 days ago

Verified 257a602f

🚨 Delete duplicate code in backbone utils (#43323)

zucchini-nlp committed 49 days ago

Verified 0c4fe5c6

Improve typing/explanations for general model properties (#43712)

Cyrilvallez committed 49 days ago

Verified 480ed54e

[Kernels] kernel migration updates for activation kernels (#43518)

ariG23498 committed 49 days ago

Verified 90b7db10

[`feat`] Allow loading T5Gemma2Encoder with AutoModel (#43559)

tomaarsen committed 49 days ago

Verified 80996194

Added S110 - try-except-pass rule (#43687)

tarekziade committed 49 days ago

Verified 31d83c77

Add EXAONE-MoE implementations (#43080)

nuxlear committed 50 days ago

Verified 379ec6b9

[docs] benchmarks (#43694)

stevhliu committed 50 days ago

Verified 71ade562

fix norm_eps dtype (#43669)

fschlatt committed 50 days ago

Verified 01e860eb

Llava onevision: output align for tests and add `image_sizes` input param (#43678)

kaixuanliu committed 50 days ago

Verified 71956a81

Fix CLIPOutput attentions not being returned (#43657)

jonathan-fulton committed 50 days ago

Verified affcf459

[`Attn`] Fixup interface usage after refactor (#43706)

vasqu committed 50 days ago

Verified ab56d8a3

Fix model/processor mismatch in SigLIP2 quantization example (#43652)

jonathan-fulton committed 50 days ago

Verified c20ba194

Fix crash of custom models in Notebook or Repl (#43690)

Cyrilvallez committed 50 days ago

Verified 432ddfad

Simplify TrainingArguments docstring (#43568)

SunMarc committed 50 days ago

Verified 670f5ab6

Composite model inherit automatically all important properties from their children (#43691)

Cyrilvallez committed 50 days ago

Verified 36ec3bfa

Update configuration_qwen3.py (#43703)

francesco-bertolotti committed 50 days ago

Verified b6a202f8

fix gptoss tp crash (#43695)

sywangyi committed 50 days ago

Verified dab8734f

[CB] Keep order of incoming requests (#43626)

remi-or committed 50 days ago

Verified cf3fda19

Fix Apertus model loading (NotImplementedError: Cannot copy out of meta tensor; no data!) (#43473)

xenova committed 50 days ago

Verified c640b63a

Remove `num_frames` in ASR pipeline (#43546)

jiqing-feng committed 50 days ago

Verified bdc583d4

Newer Older