transformers

cli

Show the actual accelerate version for transformers env (#49668 )

2026-10-20 15:57:17 +00:00

conftest_tests

Cleaned up tokenizer tests (#49669 )

2026-10-19 10:02:07 +00:00

contracts

Remove the resolved prototype Whisper MIC expected failure (#49785 )

2026-10-20 13:31:53 +00:00

distributed

FSDP x TP Saving / Loading, Optimizer (#49256 )

2026-10-15 15:09:11 +00:00

exporters

Add Unlimited OCR (#46836 )

2026-10-15 14:43:30 +00:00

fixtures

Add MOSS-Transcribe-Diarize model (#48386 )

2026-10-15 19:04:11 +02:00

generation

test: add parity check to the CPU-only continuous batching test (#49728 )

2026-10-23 03:35:08 +00:00

heterogeneity

Support per-layer MTP configuration (#48264 )

2026-09-02 14:50:35 +09:00

integrations

Add Muse Spark Open (#49644 )

2026-10-15 19:35:25 +02:00

kernels

Fix two stale hub-kernel tests (#49795 )

2026-10-20 15:57:40 +00:00

models

moe sparseblock canonicals

2026-10-27 17:29:39 -07:00

optimization

Add GreedyLR adaptive learning rate scheduler (#44271 )

2026-03-18 16:49:07 +00:00

peft_integration

QA: applied ruff rule PLW1514 (#48990 )

2026-09-22 06:49:19 +00:00

pipeline_parallel

Second phase of torch 2.6 cleanups (#49396 )

2026-10-09 17:01:40 +00:00

pipelines

Fix multi-zoom crops in the SAM mask-generation pipeline and related bugs (Fixes #35530 ) (#49689 )

2026-10-21 15:43:12 +00:00

quantization

Match a skip-list entry from the start only up to a module boundary (#49771 )

2026-10-23 13:22:36 +00:00

repo_utils

[Fetcher] Fix missing test deps for dynamically-resolved image processors (#49826 )

2026-10-22 12:10:09 +02:00

sagemaker

Remove examples (#49357 )

2026-10-13 15:11:48 +00:00

tensor_parallel

🚨 TP: give embeddings their own style, shard input embeddings by default, and shard widths that don't divide by the TP size (#49570 )

2026-10-19 08:01:56 +00:00

tokenization

Add chat input sanitization (#47386 )

2026-10-13 13:49:51 +00:00

trainer

Save the PEFT adapter of a load-time sharded model from its own parameters (#49500 )

2026-10-14 22:04:32 +00:00

utils

Fix torch_float casting to int instead of float (#47468 )

2026-10-22 12:40:11 +00:00

__init__.py

…

alm_tester.py

ALM testx fIxing... (#49355 )

2026-10-06 13:42:34 +00:00

causal_lm_tester.py

Add Granite-swa and Granitemoe-swa model support (#47179 )

2026-07-29 21:31:43 +00:00

multimodal_tester.py

🚨 MoE: shard the experts in every TP/EP plan, and test for it (#49206 )

2026-10-16 22:19:39 +00:00

test_backbone_common.py

🚨 Delete duplicate code in backbone utils (#43323 )

2026-02-04 11:37:56 +01:00

test_configuration_common.py

🚨 SubConfigSpec for centralized init of subconfigs (#48282 )

2026-10-08 10:11:43 +00:00

test_executorch.py

fix(flaky): enforce manual seed to reduce flakiness (#43794 )

2026-02-06 17:30:28 +01:00

test_feature_extraction_common.py

…

test_fsdp_mixin.py

Fix zero expert gradients in bfloat16 with expert parallelism and ep_size < fsdp_size (#49688 )

2026-10-19 04:45:19 +02:00

test_image_processing_common.py

Add image processing tester init (#48829 )

2026-09-24 17:47:37 +00:00

test_image_transforms.py

…

test_memory_cleanup_mixin.py

QA: Added a MemoryCleanupMixin class for tests (#48681 )

2026-09-11 09:16:28 +00:00

test_modeling_common.py

Fixes a dtype mismatch in flash_attn test (#49912 )

2026-10-23 11:40:42 +02:00

test_monkey_patching.py

[Fix] Remove old automatic cross attn pattern from output recorders (#47117 )

2026-07-09 00:14:10 +00:00

test_pipeline_mixin.py

QA: applied ruff rule PLW1514 (#48990 )

2026-09-22 06:49:19 +00:00

test_processing_common.py

Add Muse Spark Open (#49644 )

2026-10-15 19:35:25 +02:00

test_sentencepiece_backend_mixin.py

rm slow tokenizers (#40936 )

2025-11-27 19:24:50 +01:00

test_sequence_feature_extraction_common.py

…

test_tensor_parallel_mixin.py

Fix expert-parallel token dispatch gradients on torch 2.7 to 2.10 (#49710 )

2026-10-20 09:14:04 +02:00

test_tokenization_common.py

Cleaned up tokenizer tests (#49669 )

2026-10-19 10:02:07 +00:00

test_tokenization_mistral_common.py

CI: Point test fixtures at hf-internal-testing copies we already host (#48521 )

2026-09-04 12:49:34 +00:00

test_tokenizers_backend_mixin.py

Cleaned up tokenizer tests (#49669 )

2026-10-19 10:02:07 +00:00

test_training_mixin.py

[style] Rework ruff rules and update all files (#43144 )

2026-01-08 10:26:14 +01:00

test_video_processing_common.py

Smthn smthn (#49364 )

2026-10-06 18:17:31 +02:00

vlm_tester.py

Why override when you can fix in Mixin (#49105 )

2026-10-05 13:00:35 +00:00