huggingface/transformers

Pull Requests Commits

Merge branch 'main' into add_qat_fp8

MekkCyber committed 1 year ago

Verified 58348c17

MekkCyber committed 1 year ago

8c870cd2

Fix incorrect assignment in `update_device_map` for GPTQ quantizer (#41328)

Sai-Suraj-27 committed 1 year ago

Verified ccbaa167

[v5] Bump accelerate to 1.1.0 (#41234)

SunMarc committed 1 year ago

Verified c562c5d8

MekkCyber committed 1 year ago

e8086c12

Fix early CUDA initialisation (#41409)

hmellor committed 1 year ago

Verified 88e946e0

MekkCyber committed 1 year ago

c1eda073

Prefer raising `TypeError` exception for invalid type (#41346)

Sai-Suraj-27 committed 1 year ago

Verified 93464a02

[Model] Lfm2Moe (#41401)

paulpak58 committed 1 year ago

Verified 0c9a72e4

Fix test for model with dotted name and relative imports (#41343)

st81 committed 1 year ago

Verified b4428d54

[Cache] lfm2 cache: allocate empty kv layers during init (#41396)

paulpak58 committed 1 year ago

Verified 0464d9eb

[kernels] Kernel Config (#41232)

MekkCyber committed 1 year ago

Verified da7b8ce1

Correct numerical regression in vision embeddings (#41374)

i3hz committed 1 year ago

Verified 4763b8c5

fix resample in asr pipeline (#41298)

yhzx233 committed 1 year ago

Verified caa14e7d

fix asr ut failures (#41332)

yao-matrix committed 1 year ago

Verified 73f8c4b8

[`v5`] Sync Bert and Bart eager attention (#41248)

vasqu committed 1 year ago

Verified 57e82745

Update from pretrained error when loading (#33380)

ArthurZucker committed 1 year ago

Verified 505387c0

serve: add non-streaming mode to /v1/responses; stream event parity; remove placeholder logprobs (#41353)

antznette1 committed 1 year ago

Verified e00f46f1

[`CB`] Refactors the way we access paged (#41370)

ArthurZucker committed 1 year ago

Verified 0395ed52

Remove unused function patameters (#41358)

cyyever committed 1 year ago

Verified 39b0c949

make some ut cases pass on xpu w/ latest torch (#41337)

yao-matrix committed 1 year ago

Verified 11e4b5e5

Remove unnecessary list comprehension (#41305)

cyyever committed 1 year ago

Verified fa36c973

Fixes in check_model_inputs, GPTBigCodeModel and ImageGPTModel (#40811)

IlyasMoutawwakil committed 1 year ago

Verified 7a1aeec3

Use canonical get_size_with_aspect_ratio (with max_size) from transformers.image_transforms to fix #37939 (#41284)

sonianuj287 committed 1 year ago

Verified 297a41a6

Fix flash_attention.py: wrong argument passing for attn_implementation (#41347)

TKONIY committed 1 year ago

Verified ae60c776

[testing] update `test_longcat_generation_cpu` (#41368)

ydshieh committed 1 year ago

Verified 6bf6e36d

🚨 Remove BetterTransformer (#41367)

Cyrilvallez committed 1 year ago

Verified 4903cd40

Better typehints for `apply_chat_template` (#41355)

Samoed committed 1 year ago

Verified a5700c49

Fix typo in model proposal template (#41352)

Ombucha committed 1 year ago

Verified 089d573a

:rotating_light: [`v5`] Remove relative position embeddings (for bert like models) (#41170)

vasqu committed 1 year ago

Verified c27b67f0

Older