Commits huggingface/transformers

[docs] add quick usage snippet to Whisper. (#31289)

Vaibhavs10 committed 1 year ago

Verified 6f0ecf10

Log additional test metrics with the CometCallback (#33124)

Lothiraldan committed 1 year ago

Verified 892d51ca

Bump torch from 1.13.1 to 2.2.0 in /examples/research_projects/jax-projects/hybrid_clip (#33137)

dependabot[bot] committed 1 year ago

Verified 746e1148

CI: fix `efficientnet` pipeline timeout and prevent future similar issues due to large image size (#33123)

gante committed 1 year ago

Verified ab0ac3b9

disable scheduled daily CI temporarily (#33136)

ydshieh committed 1 year ago

Verified 3806faa1

fix: multilingual midel convert to tflite get wrong token (#32079)

Ayaa17 committed 1 year ago

Verified 7562366d

fix: Fixed CodeGenTokenizationTest::test_truncation failing test (#32850)

Sai-Suraj-27 committed 1 year ago

Verified 3bf6dd8a

Fixup py 38 type hints for mps friendly (#33128)

muellerzr committed 1 year ago

Verified 9578c259

quickfix documentation (#32566)

molbap committed 1 year ago

Verified 26f043bd

fix: Fixed `pydantic` required version in dockerfiles to make it compatible with DeepSpeed (#33105)

Sai-Suraj-27 committed 1 year ago

Verified 35627729

Add changes for uroman package to handle non-Roman characters (#32404)

nandwalritik committed 1 year ago

Verified a378a54a

mps: add `isin_mps_friendly`, a wrapper function for `torch.isin` (#33099)

gante committed 1 year ago

Verified 72d4a3f9

Test: add higher `atol` in `test_forward_with_num_logits_to_keep` (#33093)

gante committed 1 year ago

Verified 894d421e

CI: add torchvision to the consistency image (#32941)

gante committed 1 year ago

Verified 93e0e1a8

support qwen2-vl (#32318)

simonJJJ committed 1 year ago

Verified 19e6e80e

Updated the custom_models.md changed cross_entropy code (#33118)

S-M-J-I committed 1 year ago

Verified 8defc95d

Update Jinja docs with new functions and general cleanup (#33097)

Rocketknight1 committed 1 year ago

Verified 0a7af19f

added doctring to SchedulerType class (#32898)

Arunprakash-A committed 1 year ago

Verified e3a5f35c

DeviceGuard added to use Deformable Attention more safely on multi-GPU (#32910)

DonggeunYu committed 1 year ago

Verified 1dbd9d36

Enable some Jinja extensions and add datetime capabilities (#32684)

Rocketknight1 committed 1 year ago

Verified 371b9c14

Integrate Liger (Linkedin GPU Efficient Runtime) Kernel to Trainer (#32860)

JasonZhu1313 committed 1 year ago

Verified adb91179

Forbid `PretrainedConfig` from saving `generate` parameters; Update deprecations in `generate`-related code 🧹 (#32659)

gante committed 1 year ago

Verified 970a16ec

Reducing memory usage: removing useless logits computation in generate() (#31292)

Cyrilvallez committed 1 year ago

Verified 22e6f145

docs: fix outdated link to TF32 explanation (#32947)

anakin87 committed 1 year ago

Verified d806fa3e

Generate: Deprecate returning legacy cache by default; Handle `use_cache=False` (#32863)

gante committed 1 year ago

Verified a26de151

🌐 [i18n-KO] Translated `knowledge_distillation_for_image_classification.md to Korean" (#32334)

JinukHong committed 1 year ago

Verified 09e6579d

Fix regression on `Processor.save_pretrained` caused by #31691 (#32921)

leloykun committed 1 year ago

Verified 273c0afc

[run_slow] idefics2 (#32840)

andimarafioti committed 1 year ago

Verified 18199b34

Gemma2: eager attention by default (#32865)

gante committed 1 year ago

Verified 975b988b

fix: (issue #32689) `AttributeError` raised when using `Trainer` with `eval_on_start=True` in Jupyter Notebook. (#32849)

fshp971 committed 1 year ago

Verified f1d822ba