huggingface/transformers

Pull Requests Commits

CI: trigger AMD mi325 model tests with empty HF cache for important models

Abdennacer-Badaoui committed 43 days ago

081ff38a

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

remi-or committed 98 days ago

Verified cf3fda19

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

xenova committed 98 days ago

Verified c640b63a

Remove `num_frames` in ASR pipeline (#43546)

jiqing-feng committed 98 days ago

Verified bdc583d4

remove ipex and ccl for xpu and cpu (#42852)

yao-matrix committed 98 days ago

Verified 683c3e9f

🚨 Refactor DETR to updated standards (#41549)

yonigozlan committed 98 days ago

Verified aefa23ad

update guide with new attr name for toks (#43689)

itazap committed 98 days ago

Verified 8012fde3

🚨Fix floating-point precision in JanusImageProcessor resize (#43187)

raimbekovm committed 98 days ago

Verified 56c4b5e3

Docs: fix typos in Get started (index, quicktour) (#43666)

CodeByKodi committed 98 days ago

Verified 78e4f885

the cache class is deprecated

vasqu committed 98 days ago

a9cf5e53

custom tok init fix (#43591)

itazap committed 98 days ago

Verified f12b50a0

More export friendly rewrites and skipping the failing ones (#43436)

IlyasMoutawwakil committed 98 days ago

Verified 41771ce4

Cast byte_count to int in caching_allocator_warmup for MPS compatibility (#43608)

tobyliu2004 committed 98 days ago

Verified 559af445

[Docs] Complete missing Llama4 configuration docs (#43460)

udaymehta committed 98 days ago

Verified 02943590

Fix t5 failures (#43374)

Abdennacer-Badaoui committed 98 days ago

Verified 07aed9d7

Add EoMT with DINOv3 backbone (#41212)

NielsRogge committed 98 days ago

Verified 751cff7c

Update DBRX docs to reference re-uploaded checkpoint (#43196)

qgallouedec committed 99 days ago

Verified 645db400

[loading] Fix forced upcasting to fp32 (#43683)

Cyrilvallez committed 99 days ago

Verified fb22e73f

Fix FP8Expert for Qwen (#43670)

yiliu30 committed 99 days ago

Verified 010b9f1e

Simplify loading structure (#43589)

Cyrilvallez committed 99 days ago

Verified 0fa2c2f0

[CB] Refactor logic for inputs and outputs outside of the main API (#43569)

remi-or committed 99 days ago

Verified e5c8b0d1

Make sure hub errors are surfaced in `PreTrainedTokenizerBase` (#43675)

tarekziade committed 99 days ago

Verified 5ccea230

Fix `FP8Expert` for DeepSeek R1 (#43616)

yiliu30 committed 99 days ago

Verified 8fce0a18

Use correct sampling rate in chat template (#43674)

zucchini-nlp committed 99 days ago

Verified e4b22baa

[`HunYuan`] Fix RoPE init (#43411)

vasqu committed 99 days ago

Verified 4cd791d6

XPU now supports MoE kernel(MegaBlocks) implementation (#43435)

YangKai0616 committed 99 days ago

Verified b9f0074f

[`Sam`] Fixup training flags (#43567)

vasqu committed 99 days ago

Verified aeb728ab

remove torchao.autoquant from transformers (#43561)

vkuzo committed 99 days ago

Verified 5772c733

[DeepSpeed] properly handle MoE weight conversion (#43524)

kashif committed 99 days ago

Verified 063b4569

Tie zamba weights correctly (#43623)

zucchini-nlp committed 99 days ago

Verified d5eab539

Older