younesbelkada/transformers

Pull Requests Commits

Merge remote-tracking branch 'upstream/main'

younesbelkada committed 1 year ago

7e7810ab

Revert parallelism temporarily (#38240)

LysandreJik committed 1 year ago

Verified 711d78d1

CI reporting improvements (#38230)

ydshieh committed 1 year ago

Verified feec294d

Protect ParallelInterface

LysandreJik committed 1 year ago

cb513e35

LysandreJik committed 1 year ago

f4ef41c4

[gemma3] fix bidirectional attention mask (#38080)

zucchini-nlp committed 1 year ago

Verified f834d368

[mllama] fix loading and inference (#38223)

zucchini-nlp committed 1 year ago

Verified 2edb0e4b

Add padding-free to bamba (#35861)

garrett361 committed 1 year ago

Verified 390f1534

Fixing Bitnet after use_rms_norm introduction (#38229)

MekkCyber committed 1 year ago

Verified 2a794713

Enable Quantize KV Cache for Mistral Model (#35042)

Bojun-Feng committed 1 year ago

Verified 96618960

parallelism goes brrr (#37877)

nmntz committed 1 year ago

Verified 1c2f36b4

Fix Llama4 (#38222)

Cyrilvallez committed 1 year ago

Verified b591d925

Mamba2 remove unecessary test parameterization (#38227)

ivarflakstad committed 1 year ago

Verified 3f0b7d0f

Minor llama4 fixes (#38123)

molbap committed 1 year ago

Verified 9cde2f5d

fix dead flax links modeling_flax_pytorch_utils.py (#38212)

dizer-ti committed 1 year ago

Verified 856f034f

Make `train_dataset` attribute in `_get_train_sampler` optional (#38226)

SunMarc committed 1 year ago

Verified bb3c6426

In Llama4 fix wrongly inverted causal attention mask when using SDPA implementation (#38094)

sogartar committed 1 year ago

Verified 2ad152f8

Disable torchscript tests for AriaForConditionalGenerationModelTest (#38225)

ivarflakstad committed 1 year ago

Verified de70c842

Add support to Marimo Notebooks and Enverge.ai (#38210)

brenoca committed 1 year ago

Verified 8ea61c45

New cache tests and refactored Hybrid Cache (#37972)

manueldeprada committed 1 year ago

Verified d34e21e7

Add `Llama4TextModel` to `AutoModel` mapping (#38162)

Matthew Hoffman committed 1 year ago

Verified 183fb363

Remove trust_remote_code=True tests from bnb quantization tests (MPT now integrated) (#38206)

Titus-von-Koeller committed 1 year ago

Verified f022bf93

[fix] sliding window attention mask (#38045)

zucchini-nlp committed 1 year ago

Verified 0a52bd24

Fix broken example generation script for Llama3 (#38062)

shfoss committed 1 year ago

Verified 555715f4

Fix: make docs work better with doc builder (#38213)

S1ro1 committed 1 year ago

Verified 7a611f0a

enable misc cases on XPU & use device agnostic APIs for cases in tests (#38192)

yao-matrix committed 1 year ago

Verified 3bd1c201

Qwen2.5-Omni: Update modeling_qwen2_5_omni.py to fix error when loading quantized weights with AutoAWQ. (#38013)

tiger-of-shawn committed 1 year ago

Verified dbc4b91d

Feat: save_pretrained for tensor parallel (and other parallelisms) models (#37919)

S1ro1 committed 1 year ago

Verified 46a4b7c9

[doc] fix bugs in `how_to_hack_models.md` (#38198)

faaany committed 1 year ago

Verified 9ecee143

Translating model_doc/bert.md to Chinese (#37806)

Nanji-Huaji committed 1 year ago

Verified f524439c

Older