NielsRogge/transformers

Pull Requests Commits

Merge branch 'main' into fix_rt_detr_docs

NielsRogge committed 1 year ago

Verified 43d84174

Fix: Correct tensor shape comment in Mamba modeling (#37801)

ShadyPi committed 1 year ago

Verified 97e57b25

[doc] fix the code examples in qwen doc (#37803)

jiangyukunok committed 1 year ago

Verified 33493542

Fix typos in strings and comments (#37799)

co63oc committed 1 year ago

Verified d5fa7d2d

Define warmup allocator for torchao quantization (#37764)

MekkCyber committed 1 year ago

Verified f4666039

Fix the fsdp config cannot work issue. (#37549)

yuanwu2017 committed 1 year ago

Verified a41b6d9b

Gemma3 is Torch Exportable (#37728)

guangy10 committed 1 year ago

Verified 816b3701

NielsRogge committed 1 year ago

12818160

Fix error message in `hub.py` (#37796)

srai9 committed 1 year ago

Verified 397a5ede

fix performance issue in convert_ids_to_tokens (#37773)

martin-harmonic committed 1 year ago

Verified 6ce675ee

chore: update SigLIP2 model card (#37624)

saswatmeher committed 1 year ago

Verified 57c620bf

[i18n-KO] Translated `keypoint_detection.md` to Korean (#36649)

rlaalsrl0922 committed 1 year ago

Verified eb4afdd1

fix mpt test of different outputs from cuda (#37691)

jiqing-feng committed 1 year ago

Verified 555693fb

Force torch>=2.6 with torch.load to avoid vulnerability issue (#37785)

Cyrilvallez committed 1 year ago

Verified 0cfbf9c9

Fix tensor parallel with non-floating dtypes (#37790)

Cyrilvallez committed 1 year ago

Verified eefc86aa

Fix typos in strings and comments (#37784)

co63oc committed 1 year ago

Verified 21406220

Align gpt2 mask preparation to #37612 (#37787)

Cyrilvallez committed 1 year ago

Verified ba3bd372

unpin pytest<8 (#37768)

ydshieh committed 1 year ago

Verified 50d231a8

[causal mask] fix preparation with multi-gpu (#37612)

zucchini-nlp committed 1 year ago

Verified 79d4bc76

🌐 [i18n-KO] Translated `roberta.md` to Korean (#37069)

garongkim committed 1 year ago

Verified 7bb619d7

Update model card for Gemma (#37674)

afafelwafi committed 1 year ago

Verified cfe66691

Fix auto-round hfoption (#37759)

MekkCyber committed 1 year ago

Verified b2d70e9c

Guard DeepSpeed imports (#37755)

lewtun committed 1 year ago

Verified acdbe627

[deps] pin max `torch` version (#37760)

gante committed 1 year ago

Verified af6d2756

Fix typos in comments (#37694)

co63oc committed 1 year ago

Verified 0302aa1c

Fix load of rng state for resuming training from checkpoint (#37162)

winglian committed 1 year ago

Verified af000ceb

Fix tied weight loading with TP and loading sub state_dicts (#37758)

Cyrilvallez committed 1 year ago

Verified 0af0a5f9

Refine parameter type annotations (#37666)

flashJd committed 1 year ago

Verified 3af24f7e

Fix wrong input shapes in doc-string of models (#37729)

kkew3 committed 1 year ago

Verified 22e3da92

[generate] fix default autocompile case on gpu (#37756)

gante committed 1 year ago

Verified 4d64c385

Older