| commit |
1.00 |
Update AFMoE architecture to use v5-style MoE impl (#44063) |
|
Commit message contains explicit AI assi |
2026-03-19 |
| commit |
1.00 |
Sdpa for owlvit (#42136) |
|
Commit message contains explicit AI assi |
2026-03-17 |
| commit |
1.00 |
:rotating_light: Validate config attributes (#41250) |
|
Commit message contains explicit AI assi |
2026-03-16 |
| commit |
1.00 |
Fix off-by-one in decode_spans boundary check (#44584) |
|
Commit message contains explicit AI assi |
2026-03-12 |
| pr |
1.00 |
Fix #44155: [AudioFlamingo3] Batched inference produces inco |
|
PR body explicitly mentions AI collabora |
2026-02-21 |
| commit |
0.00 |
Fix glm dsa (#44564) |
|
— |
2026-03-19 |
| commit |
0.00 |
🚨🚨 Refactor Image Processors to support different backends ( |
|
— |
2026-03-19 |
| commit |
0.00 |
[generate] Never use `cache_position` anymore in generation |
|
— |
2026-03-19 |
| commit |
0.00 |
Fix KeyError in convert_to_native_format for dict vocab (#44 |
|
— |
2026-03-19 |
| commit |
0.00 |
fix: XLNet: relative_positional_encoding computes on CPU eve |
|
— |
2026-03-19 |
| commit |
0.00 |
Fix annotations reader for python 3.14 in `PreTrainedModel` |
|
— |
2026-03-19 |
| commit |
0.00 |
[CB] Better parametrization for compile (#44578) |
|
— |
2026-03-19 |
| commit |
0.00 |
Fix `KeyError` when patching mistral regex (#43376) |
|
— |
2026-03-19 |
| commit |
0.00 |
Correct code block formatting in weightconverter.md (#44839) |
|
— |
2026-03-19 |
| commit |
0.00 |
deepseek_v2, deepseek_v3, and modernbert fix for having inco |
|
— |
2026-03-18 |
| commit |
0.00 |
[Model] Add PP-OCRv5_server_rec and PP-OCRv5_mobile_rec mod |
|
— |
2026-03-18 |
| commit |
0.00 |
Add `Jina-Embeddings-V3` Model (#44251) |
|
— |
2026-03-18 |
| commit |
0.00 |
feat(ci): added a network debug report (#44636) |
|
— |
2026-03-18 |
| commit |
0.00 |
Add GreedyLR adaptive learning rate scheduler (#44271) |
|
— |
2026-03-18 |
| commit |
0.00 |
Fix unexpected `position_ids` keys when loading OwlViT model |
|
— |
2026-03-18 |
| commit |
0.00 |
Update more modular examples (#44834) |
|
— |
2026-03-18 |
| commit |
0.00 |
Fix and re-run modular converter on examples (#44833) |
|
— |
2026-03-18 |
| commit |
0.00 |
Remove cache_position in more models (4 and last one) (#4482 |
|
— |
2026-03-18 |
| commit |
0.00 |
Fix loading issue in Sam3 (#44831) |
|
— |
2026-03-18 |
| commit |
0.00 |
feat(integration): Add KubeflowCallback to enable automatic |
|
— |
2026-03-18 |
| commit |
0.00 |
Add GGUF support for MiniMax-M2.1 model (#44526) |
|
— |
2026-03-18 |
| commit |
0.00 |
Centralize AI agent templates in `.ai` (#44489) |
|
— |
2026-03-18 |
| commit |
0.00 |
support xxxFast alias in v5 tokenizers (#44766) |
|
— |
2026-03-18 |
| commit |
0.00 |
Remove cache_position in more models (3) (#44759) |
|
— |
2026-03-18 |
| commit |
0.00 |
Fix `supports_{tp/pp}_plan` (#44696) |
|
— |
2026-03-18 |
| commit |
0.00 |
[CI] Temporarily skip Mistral4 tests as they almost all fail |
|
— |
2026-03-18 |
| commit |
0.00 |
update flex attention to use `return_aux` instead of `return |
|
— |
2026-03-18 |
| commit |
0.00 |
[Gemma] Update conversion scripts for Transformers v5 Comapt |
|
— |
2026-03-18 |
| commit |
0.00 |
fix bug embedding_size mismatch with hidden_size in electra |
|
— |
2026-03-18 |
| commit |
0.00 |
Fix pegasus conversion (#44571) |
|
— |
2026-03-18 |
| commit |
0.00 |
Fix repo-check bot (#44812) |
|
— |
2026-03-18 |
| commit |
0.00 |
[docs] is_causal feature (#44777) |
|
— |
2026-03-17 |
| commit |
0.00 |
docs(tasks): remove references to removed question-answering |
|
— |
2026-03-17 |
| commit |
0.00 |
Fix configs with `@strict` (#44770) |
|
— |
2026-03-17 |
| commit |
0.00 |
[AMD CI] Fix test failures across important models (#44632) |
|
— |
2026-03-17 |
| commit |
0.00 |
Move VLM conversions to the main mapping (#44627) |
|
— |
2026-03-17 |
| commit |
0.00 |
Fix config loading issues (type issues) (#44789) |
|
— |
2026-03-17 |
| commit |
0.00 |
Remove `is_causal` from `EuroBertConfig` (#44774) |
|
— |
2026-03-17 |
| commit |
0.00 |
model-linter: Added rule 10 (#44761) |
|
— |
2026-03-17 |
| commit |
0.00 |
[fix] mistral 4 docs (#44776) |
|
— |
2026-03-16 |
| commit |
0.00 |
Add Mistral 4 (#44760) |
|
— |
2026-03-16 |
| commit |
0.00 |
Fix: Eurobert model was missing @strict decorator and invali |
|
— |
2026-03-16 |
| commit |
0.00 |
fix: sig lip import (#44764) |
|
— |
2026-03-16 |
| commit |
0.00 |
Disable async loading when quantizing on the fly (#44576) |
|
— |
2026-03-16 |
| commit |
0.00 |
Bump torchao >=0.15 and fix quantization CI (#44604) |
|
— |
2026-03-16 |
| commit |
0.00 |
Fix tensor indexing crash in serve generate_response KV cach |
|
— |
2026-03-16 |
| commit |
0.00 |
[MistralCommonBackend] Upgrade mistral-common to v1.10.0 (#4 |
|
— |
2026-03-16 |
| commit |
0.00 |
Fix `mlcd` auto config/model/mapping issues (#44730) |
|
— |
2026-03-16 |
| commit |
0.00 |
Fix bug and add XPU Expectations for qwen2 and jamba tests ( |
|
— |
2026-03-16 |
| commit |
0.00 |
Add model lerobot PI0 to transformers (#44160) |
|
— |
2026-03-16 |
| commit |
0.00 |
[medasr] doc update (#44633) |
|
— |
2026-03-16 |
| commit |
0.00 |
Idefics3 without cache fix (#44607) |
|
— |
2026-03-16 |
| commit |
0.00 |
Add XPU Expectations for vibe voice acoustic tokenizer tests |
|
— |
2026-03-16 |
| commit |
0.00 |
Fix transformers serve's 422 unprocessable entity (#44620) |
|
— |
2026-03-16 |
| commit |
0.00 |
Fix missing / incorrect `config` class in some model class d |
|
— |
2026-03-15 |
| commit |
0.00 |
Update Nvidia CI docker file to use torch 2.10 (#44712) |
|
— |
2026-03-14 |
| commit |
0.00 |
[`FA`] Fix fa detection (#44703) |
|
— |
2026-03-14 |
| commit |
0.00 |
Fix `set_encoder` (#44698) |
|
— |
2026-03-14 |
| commit |
0.00 |
[docs] cb config (#44675) |
|
— |
2026-03-13 |
| commit |
0.00 |
Fix more model tester missing `parent` issue (#44685) |
|
— |
2026-03-13 |
| commit |
0.00 |
:rotating_light: [`FA4`] Initial support (#42435) |
|
— |
2026-03-13 |
| commit |
0.00 |
Add register method for `ParallelInterface` (#44640) |
|
— |
2026-03-13 |
| commit |
0.00 |
[CB] [Bug] Fix crashes when running without cuda (#44673) |
|
— |
2026-03-13 |
| commit |
0.00 |
Another (small) set of fixes required for tiny model creatio |
|
— |
2026-03-13 |
| commit |
0.00 |
Fix CookieCutter (#44334) |
|
— |
2026-03-13 |
| commit |
0.00 |
Fix AWQ tests for GPTQModel migration (#44654) |
|
— |
2026-03-13 |
| commit |
0.00 |
[Model] Add PP-OCRV5_mobile_det Model Support (#43247) |
|
— |
2026-03-13 |
| commit |
0.00 |
pipelines do not have modelcard (#44621) |
|
— |
2026-03-13 |
| commit |
0.00 |
[`Chmv2`] Fix conversion after capture refactor (#44665) |
|
— |
2026-03-13 |
| commit |
0.00 |
fix(models, testing): Fix Llama4 vision rotary meta tensor i |
|
— |
2026-03-13 |
| commit |
0.00 |
[CB] Add dedicated config (#44434) |
|
— |
2026-03-13 |
| commit |
0.00 |
fix(models): Forward timm model kwargs to timm.create_model |
|
— |
2026-03-13 |
| commit |
0.00 |
Ensure same `dtype` for subconfig when `_from_config` (#4462 |
|
— |
2026-03-13 |
| commit |
0.00 |
Remove `cache_position` in more models (2) (#44602) |
|
— |
2026-03-12 |
| commit |
0.00 |
fix: cast to proper dtype in EmbeddingParallel (#44612) |
|
— |
2026-03-12 |
| commit |
0.00 |
Allow to disable stdout hiding for TP (#44608) |
|
— |
2026-03-12 |
| commit |
0.00 |
Remove many output_attentions and other traced outputs on 10 |
|
— |
2026-03-12 |
| commit |
0.00 |
[Model] Add PP-OCRV5_server_det Model Support (#43274) |
|
— |
2026-03-12 |
| commit |
0.00 |
fix: raise error if mm_token_type_ids not supplied (#44433) |
|
— |
2026-03-12 |
| commit |
0.00 |
Fix output capturing for Backbones (#44638) |
|
— |
2026-03-12 |
| commit |
0.00 |
Fix lfm2 kernel path (#44634) |
|
— |
2026-03-12 |
| commit |
0.00 |
Fix for `VibeVoiceAcousticTokenizer` (#44628) |
|
— |
2026-03-12 |
| commit |
0.00 |
Add an integration test for LASR using pipe and chunked deco |
|
— |
2026-03-12 |
| commit |
0.00 |
Fix more wrong HF hub checkpoint names (#44624) |
|
— |
2026-03-12 |
| commit |
0.00 |
Update agentic contributions guidelines in AGENTS.md to forc |
|
— |
2026-03-12 |
| commit |
0.00 |
Expand model-structure lint rules with a fast AST-based, ruf |
|
— |
2026-03-12 |
| commit |
0.00 |
feat: add neuron in tensor parallelism initialization (#4449 |
|
— |
2026-03-11 |
| commit |
0.00 |
[WIP] FIX Make Mixtral LoRA loading work (#44478) |
|
— |
2026-03-11 |
| commit |
0.00 |
Fix Llava tests for torch too! (#44476) |
|
— |
2026-03-11 |
| commit |
0.00 |
Fix training ci and clean some tests (#44491) |
|
— |
2026-03-11 |
| commit |
0.00 |
Add CHMv2 (#44595) |
|
— |
2026-03-11 |
| commit |
0.00 |
Remove useless identity assignment (#44600) |
|
— |
2026-03-11 |
| commit |
0.00 |
Add Yoni to run-slow workflow (#44598) |
|
— |
2026-03-11 |
| commit |
0.00 |
Add shared VLM tests (#42964) |
|
— |
2026-03-11 |
| commit |
0.00 |
Fix wrong (non-existing) checkpoints (#44549) |
|
— |
2026-03-11 |
| commit |
0.00 |
Remove `cache_position` in more models (#44330) |
|
— |
2026-03-11 |
| pr |
0.00 |
Switch FP8 per tensor quant to use `torch._scaled_mm` |
|
— |
2026-03-19 |
| pr |
0.00 |
DeepGEMM |
|
— |
2026-03-18 |
| pr |
0.00 |
Update some type hints |
|
— |
2026-03-19 |
| pr |
0.00 |
Proposal to add Qwen3-ASR support [WIP] |
|
— |
2026-02-08 |
| pr |
0.00 |
[Model] Add PP-Chart2Table Model Support |
|
— |
2026-02-05 |
| pr |
0.00 |
Dequant fix |
|
— |
2026-03-18 |
| pr |
0.00 |
[Model] Add SLANeXt Model Support |
|
— |
2026-02-03 |
| pr |
0.00 |
🚨 Refactor ViT to updated standards |
|
— |
2025-10-17 |
| pr |
0.00 |
Add THD support in ESM |
|
— |
2026-02-19 |
| pr |
0.00 |
[Model] Add UVDoc Model Support |
|
— |
2026-01-21 |
| pr |
0.00 |
feat: added cache to the model linter |
|
— |
2026-03-17 |
| pr |
0.00 |
Propagate the model loading from transformers serve to chat |
|
— |
2026-03-16 |
| pr |
0.00 |
chore(typing): extend typing to `src/transformers/cli` |
|
— |
2026-03-10 |
| pr |
0.00 |
Fix core dumped when `NemotronH` is torch compiled |
|
— |
2026-03-19 |
| pr |
0.00 |
Officially launch parse_response |
|
— |
2026-03-13 |
| pr |
0.00 |
[CB] Add an option to return logprobs |
|
— |
2026-03-18 |
| pr |
0.00 |
fix: handle list-type _tied_weights_keys in _get_tied_weight |
|
— |
2026-03-19 |
| pr |
0.00 |
Fix glm dsa |
|
— |
2026-03-10 |
| pr |
0.00 |
[PoC] HF exporters |
|
— |
2025-11-03 |
| pr |
0.00 |
[Mistral] Fix query scaling for Mistral4 and Ministral3 |
|
— |
2026-03-19 |
| pr |
0.00 |
Fix several based models' pipeline parallel support |
|
— |
2026-03-14 |
| pr |
0.00 |
Support Modular (!!) + Configs in `check_auto_docstrings` |
|
— |
2026-03-17 |
| pr |
0.00 |
Fix failing `Qwen3OmniModelIntegrationTests` |
|
— |
2026-03-19 |
| pr |
0.00 |
🚨🚨 Refactor Image Processors to support different backends |
|
— |
2026-01-27 |
| pr |
0.00 |
Dynamic weight conversion is recursive |
|
— |
2026-02-26 |
| pr |
0.00 |
FSDP2 native support in transformers |
|
— |
2026-02-17 |
| pr |
0.00 |
[generate] Never use `cache_position` anymore in generation |
|
— |
2026-03-18 |
| pr |
0.00 |
add HyperClovaX Vision |
|
— |
2026-02-27 |
| pr |
0.00 |
perceptron: Isaac-0.1 implementation |
|
— |
2025-09-18 |
| pr |
0.00 |
refactor: rope in model, flatten vision, rely on qwen3 backo |
|
— |
2026-03-19 |
| pr |
0.00 |
enable tp for benchmark |
|
— |
2026-02-05 |
| pr |
0.00 |
Update AFMoE architecture to use v5-style MoE impl |
|
— |
2026-02-17 |
| pr |
0.00 |
Fix KeyError in convert_to_native_format for dict vocab |
|
— |
2026-03-05 |
| pr |
0.00 |
Use `index_select` instead of advanced indexing in `batched_ |
|
— |
2026-03-13 |
| pr |
0.00 |
fix: XLNet: relative_positional_encoding computes on CPU eve |
|
— |
2026-03-17 |
| pr |
0.00 |
Fix annotations reader for python 3.14 in `PreTrainedModel` |
|
— |
2026-03-13 |
| pr |
0.00 |
fix: allow AutoImageProcessor to load from URL |
|
— |
2026-03-18 |
| pr |
0.00 |
Add Music Flamingo |
|
— |
2026-01-27 |
| pr |
0.00 |
[CB] [Minor] Simplify test suite |
|
— |
2026-03-19 |
| pr |
0.00 |
fix(testing): Fix PaliGemma 2 and PaddleOCR-VL test failures |
|
— |
2026-03-16 |
| pr |
0.00 |
fix: Add MXFP4 MoE/attention backward kernels |
|
— |
2026-02-05 |
| pr |
0.00 |
fix: handle unpicklable tokenizers in ProcessorMixin.to_dict |
|
— |
2026-03-19 |
| pr |
0.00 |
deepseek_v2, deepseek_v3, and modernbert fix for having inco |
|
— |
2026-03-17 |
| pr |
0.00 |
fix: move comments before @torch.jit.script decorator for Py |
|
— |
2026-03-19 |
| pr |
0.00 |
Fix DEIM config export and public API |
|
— |
2026-03-19 |
| pr |
0.00 |
Add /v1/completions endpoint (OpenAI legacy completions API) |
|
— |
2026-03-10 |
| pr |
0.00 |
[Misc] add enable_thinking to template kwargs |
|
— |
2026-03-18 |
| pr |
0.00 |
model: Add DEIMv2 to Transformers |
|
— |
2026-02-27 |
| pr |
0.00 |
Add xcodec2 model |
|
— |
2026-02-20 |
| pr |
0.00 |
[`Mllama`] Fix workaround compile |
|
— |
2026-03-19 |
| pr |
0.00 |
Fix Zamba2MambaMixer ignoring use_mamba_kernels=False |
|
— |
2026-03-19 |
| pr |
0.00 |
Fix AutoImageProcessor URL loading regression |
|
— |
2026-03-19 |
| pr |
0.00 |
Goodbye cache position |
|
— |
2026-03-13 |
| pr |
0.00 |
[CB] Better parametrization for compile |
|
— |
2026-03-10 |
| pr |
0.00 |
Allow kernel modules to declare their preferred mask functio |
|
— |
2026-03-13 |
| pr |
0.00 |
[Model] Add PP-OCRV5_mobile_rec Model Support |
|
— |
2026-02-06 |
| pr |
0.00 |
Fix AutoImageProcessor.from_pretrained failing with URL inpu |
|
— |
2026-03-18 |
| pr |
0.00 |
Fix whisper return language |
|
— |
2025-11-16 |
| pr |
0.00 |
fix(flaky): use a fixture for `set_seed` and single-threadin |
|
— |
2026-02-07 |
| pr |
0.00 |
Add `Jina-Embeddings-V3` Model |
|
— |
2026-02-24 |
| pr |
0.00 |
[docs] training on specific hardware |
|
— |
2026-03-17 |
| pr |
0.00 |
Fix `AutoImageProcessor` to correctly detect local implement |
|
— |
2026-03-13 |
| pr |
0.00 |
Use doc-builder runnable example for GLM-ASR |
|
— |
2026-02-25 |
| pr |
0.00 |
Fix Mllama torch.compile failure caused by new attention mas |
|
— |
2026-03-19 |
| pr |
0.00 |
Fix `KeyError` when patching mistral regex |
|
— |
2026-01-20 |
| pr |
0.00 |
ci: add anti-slop action |
|
— |
2026-03-19 |
| pr |
0.00 |
Correct code block formatting in weightconverter.md |
|
— |
2026-03-19 |
| pr |
0.00 |
[Docs] Update DeiT model card to new format |
|
— |
2026-03-19 |
| pr |
0.00 |
Fix llama4 bnb mode |
|
— |
2026-03-11 |
| pr |
0.00 |
Add cu_seqlens support to OlmoHybridGatedDeltaNet for packed |
|
— |
2026-03-18 |
| pr |
0.00 |
Internalise the NomicBERT model |
|
— |
2025-12-29 |
| pr |
0.00 |
[docs] optimizers, hyperparam search, training features |
|
— |
2026-02-26 |
| pr |
0.00 |
[docs] model cards |
|
— |
2026-03-18 |
| pr |
0.00 |
Fix Mistral4 tests |
|
— |
2026-03-18 |
| pr |
0.00 |
[Model] Add PP-OCRv5_server_rec and PP-OCRv5_mobile_rec mod |
|
— |
2026-03-18 |
| pr |
0.00 |
small cleaning of quantization class |
|
— |
2025-12-04 |
| pr |
0.00 |
feat(ci): added a network debug report |
|
— |
2026-03-12 |
| pr |
0.00 |
Add GreedyLR adaptive learning rate scheduler |
|
— |
2026-02-25 |
| pr |
0.00 |
Fix unexpected `position_ids` keys when loading OwlViT model |
|
— |
2026-03-06 |
| pr |
0.00 |
Add Mistral 4 |
|
— |
2026-03-16 |
| pr |
0.00 |
Add `base_model_tp_plan` to `OlmoeConfig` |
|
— |
2026-03-13 |
| pr |
0.00 |
Update more modular examples |
|
— |
2026-03-18 |
| pr |
0.00 |
fix(gpt2): Resolve NaN/Inf issue in lm_head on Python 3.13 w |
|
— |
2026-03-13 |
| pr |
0.00 |
Fix and re-run modular converter on examples |
|
— |
2026-03-18 |
| pr |
0.00 |
[Model] Add PP-OCRv5_server_rec Model Support |
|
— |
2026-02-06 |
| pr |
0.00 |
fix: add Float8 dtype fallback in modeling_utils.py |
|
— |
2026-03-11 |
| pr |
0.00 |
Remove cache_position in more models (4 and last one) |
|
— |
2026-03-18 |
| pr |
0.00 |
docs(pipelines): remove outdated question-answering example |
|
— |
2026-03-17 |
| pr |
0.00 |
Fix loading issue in Sam3 |
|
— |
2026-03-18 |
| pr |
0.00 |
docs(quicktour): remove question-answering pipeline from qui |
|
— |
2026-03-18 |
| pr |
0.00 |
fix: handle dict vocab in CamembertTokenizer for tokenizer.j |
|
— |
2026-03-17 |
| pr |
0.00 |
Add MPS (Apple Silicon) example and documentation |
|
— |
2026-03-17 |
| pr |
0.00 |
fix: Cache XLNet relative_positional_encoding to avoid CPU c |
|
— |
2026-03-16 |
| pr |
0.00 |
fix: resolve false-positive regex warning for non-mistral mo |
|
— |
2026-03-16 |
| pr |
0.00 |
Fix: propagate interpolate_pos_encoding through PixioEmbeddi |
|
— |
2026-03-15 |
| pr |
0.00 |
feat(integration): Add KubeflowCallback to enable automatic |
|
— |
2026-03-06 |
| pr |
0.00 |
Add AudioFlamingoNext model |
|
— |
2026-03-18 |
| pr |
0.00 |
fix series of failed test case for janus model |
|
— |
2026-03-16 |
| pr |
0.00 |
Add GGUF support for MiniMax-M2.1 model |
|
— |
2026-03-08 |
