Morph MORPH
SIGN IN SIGN UP
dhruvhead / transformers UNCLAIMED
0 0 0 Python

COMMITS

main
August 21, 2026
B
[VITS] Un-skip test_model_forward (#46375)
Bob Ok committed
bb593bf
Q
Apply context parallelism to the evaluation path (#48167)
Quentin Gallouédec committed
695bd68
T
Fix `gpt_oss` runs on GPU (#48118)
Tarek Ziade committed
275b524
F
🚨 TP dtensor API inference + training (#47579)
Ferdinand Mom committed
861f4c4
Y
[EsmFold2] Fix stale expected distogram logit values (#48182)
Yih-Dar committed
d4c297d
Y
Fix Apr 05 integration test regressions (cuda sm_86) (#48170)
Yih-Dar committed
cc33c37
Z
Add MLU support to is_flash_linear_attention_available (#46995)
zsl2549 committed
6a9e2d3
B
Fix ROCm SDPA-flash skip guard that crashes on RDNA GPUs (#47965)
BADAOUI Abdennacer committed
3cb4f80
Y
Fix integration test expected values for cuda sm_86 (Mar 15 regressions) (#48168)
Yih-Dar committed
251c2d4
E
Fix DynamicCache reconstruction during ExecuTorch export (#47900)
Elad Segal committed
3cd1a23
C
Revert "Support per-layer cache configuration and attention-mask selection" (#48175)
Cyril Vallez committed
6b3a9de
E
Support per-layer cache configuration and attention-mask selection (#47901)
Elad Segal committed
c84ac34
August 20, 2026
Y
[LLaVA] Fix pixtral integration tests for cuda sm_86 (#48166)
Yih-Dar committed
945dac9
M
Port ESMC and ESMFold2 to Transformers (#46419)
Matt committed
c8df3fc
Y
[Qwen2.5-Omni] Update stale expected values for cuda sm_86 (#48164)
Yih-Dar committed
cdea840
Y
[Mistral3] Fix batched integration tests: padding_side=left + update expected values (#48161)
Yih-Dar committed
dd6c64b
H
Fix DeepSeek V2 default vocab size (#48159)
Harry Mellor committed
e4dd56a
Y
[InternVL] Fix stale expected values for Llama integration tests (cuda sm_80) (#48153)
Yih-Dar committed
0c92811
Y
Retry transient network errors (RemoteDisconnected) in github_utils (#48124)
Yih-Dar committed
9511951
P
Revert "[Quantization]: Refactor is_quantization_compressed for format-based detection" (#48072)
Park Soobin committed
238a2db
Y
[Whisper] Fix speculative decoding: preserve cleared suppress tokens through super().generate() (#48108)
Yih-Dar committed
b668804
Y
🚨[wav2vec2] Support attn_implementation=sdpa dispatch (#46196)
YangKai0616 committed
2b29630
K
fix bugs for clvp model (#47127)
kaixuanliu committed
9ad4b85
J
Always tie embeddings for LongT5 and Pop2Piano (#47620)
jiqing-feng committed
5909a46
August 19, 2026
P
Use generator with seed for LengthGroupedSampler in Trainer._get_eval_sampler for deterministic eval order with per_device_eval_batch_size > 1 (#48025)
Philip Shurpik committed
94f09cf
Y
Enable mlinter findings artifact for inline PR reviews (#48117)
Yih-Dar committed
decba1d
K
Fix ASR pipeline mono conversion for channels-last audio (fixes #47886) (#47888)
Kayvan Zahiri committed
bb8f235
J
Fix CpmAnt loading: size lm_head to vocab_size (#48012)
jiqing-feng committed
e872ca2
Y
Delete old mlinter review comments before posting new ones (#48107)
Yih-Dar committed
de795d9
J
Fix Qwen2.5-Omni / Qwen3-Omni-MoE generation with a compilable cache (#47872)
jiqing-feng committed
279dbcf