huggingface/transformers

Pull Requests Commits

Less fixes for llama

muellerzr committed 1 year ago

7c47fc22

Add language to word timestamps for Whisper (#31572)

robinderat committed 1 year ago

Verified b31d5950

Pass missing arguments to `SeamlessM4Tv2ConformerEncoderLayer.forward()` when gradient checkpointing is enabled (#31945)

anferico committed 1 year ago

Verified cb23d1b2

doc: fix broken BEiT and DiNAT model links on Backbone page (#32029)

dvrogozh committed 1 year ago

Verified bc36c26f

Fix typo in classification function selection logic to improve code consistency (#32031)

moses committed 1 year ago

Verified 63be8e6f

Fixed `log messages` that are resulting in TypeError due to too many arguments (#32017)

Sai-Suraj-27 committed 1 year ago

Verified 72fb02c4

Fix tests skip (#32012)

qubvel committed 1 year ago

Verified 691586b0

Chameleon: add model (#31534)

zucchini-nlp committed 1 year ago

Verified 24cfcc21

SpeechEncoderDecoder doesn't support param buffer assignments (#32009)

muellerzr committed 1 year ago

Verified 4037a2b5

Fix if else and *actually* enable superfast init (#32007)

muellerzr committed 1 year ago

Verified 6f40a213

Fix gather when collecting 'num_input_tokens_seen' (#31974)

CodeCreator committed 1 year ago

Verified e3917064

Bug report update -- round 2 (#32006)

gante committed 1 year ago

Verified c22efa61

fix: Fixed incorrect dictionary assignment in `src/transformers/__init__.py` (#31993)

Sai-Suraj-27 committed 1 year ago

Verified 88e0813d

add flash-attn deterministic option to flash-attn>=2.4.1 (#31961)

junrae6454 committed 1 year ago

Verified 036d3de2

Bug report update (#31983)

gante committed 1 year ago

Verified 89eec5cf

Tests: remove cuda versions when the result is the same 🧹🧹 (#31955)

gante committed 1 year ago

Verified 999981da

Fix bad test about slower init (#32002)

muellerzr committed 1 year ago

Verified 693cb828

[tests] fix deepspeed zero3 config for `test_stage3_nvme_offload` (#31881)

faaany committed 1 year ago

Verified 25e5e3fa

Speedup model init on CPU (by 10x+ for llama-3-8B as one example) (#31771)

muellerzr committed 1 year ago

Verified e0dfd7bc

Cambricon MLUs support SDPA and flash_attn (#31102)

huismiling committed 1 year ago

Verified 03a3becc

Fix the incorrect permutation of gguf (#31788)

PenutChen committed 1 year ago

Verified ac946aac

Generate: doc nits (#31982)

gante committed 1 year ago

Verified 6fbea6d2

Masking: remove flakiness from test (#31939)

gante committed 1 year ago

Verified e4682de6

Avoid race condition (#31973)

ydshieh committed 1 year ago

Verified a1a34657

Notify new docker images built for circleci (#31701)

ydshieh committed 1 year ago

Verified 11efb4fc

fix: Fixed the arguments in `create_repo()` function call (#31947)

Sai-Suraj-27 committed 1 year ago

Verified 556a4205

Generate: handle `logits_warper` update in models with custom generate fn (#31957)

gante committed 1 year ago

Verified 90750042

fix: Removed a wrong key-word argument in `sigmoid_focal_loss()` function call (#31951)

Sai-Suraj-27 committed 1 year ago

Verified 454bc14d

Whisper: move to tensor cpu before converting to np array at decode time (#31954)

gante committed 1 year ago

Verified a5c642fe

Generate: v4.42 deprecations 🧹🧹 (#31956)

gante committed 1 year ago

Verified df1c248a

Older