| commit |
1.00 |
QA: restore masking_utils export comment with a noqa (#49209 |
|
Commit message contains explicit AI assi |
2026-09-30 |
| commit |
1.00 |
Replace datasets that no longer load with maintained uploads |
|
Commit message contains explicit AI assi |
2026-09-30 |
| commit |
1.00 |
Finish removing the MPS autocast workaround (#49157) |
|
Commit message contains explicit AI assi |
2026-09-29 |
| commit |
1.00 |
Document running the example scripts on Hugging Face Jobs (# |
|
Commit message contains explicit AI assi |
2026-09-29 |
| commit |
1.00 |
[CacheHardIntegrationTest] use dedicated safetensors repo to |
|
Commit message contains explicit AI assi |
2026-09-29 |
| commit |
1.00 |
[imagegpt/vilt/trocr] fix cache fixture tests: use hf_hub_do |
|
Commit message contains explicit AI assi |
2026-09-29 |
| commit |
1.00 |
Fix get_json_schema dropping items/enum for unions of list/d |
|
Commit message contains explicit AI assi |
2026-09-28 |
| commit |
1.00 |
[Gemma] Fix test_model_7b_fp16_static_cache expected value f |
|
Commit message contains explicit AI assi |
2026-09-28 |
| commit |
1.00 |
Restore the Unicode whitespace set in the GPT-SW3 tokenizer |
|
Commit message contains explicit AI assi |
2026-09-28 |
| commit |
1.00 |
[gemma3n] fix audio test fixture: use hf_hub_download instea |
|
Commit message contains explicit AI assi |
2026-09-27 |
| commit |
1.00 |
Remap the legacy Gemma 1 hidden_act in the config post-init |
|
Commit message contains explicit AI assi |
2026-09-25 |
| commit |
1.00 |
[NemotronH-Omni] Fix device mismatch in test tensor creation |
|
Commit message contains explicit AI assi |
2026-09-25 |
| commit |
1.00 |
Use namespaced dataset ids in docs and PyTorch examples (#49 |
|
Commit message contains explicit AI assi |
2026-09-25 |
| commit |
1.00 |
[Zamba] Fix associative scan breaking ONNX export and OOM in |
|
Commit message contains explicit AI assi |
2026-09-25 |
| commit |
1.00 |
Bump huggingface_hub upper bound to <3.0 (transformers) (#49 |
|
Commit message contains explicit AI assi |
2026-09-24 |
| commit |
1.00 |
Keep already-decoded array arguments unchanged (#49062) |
|
Commit message contains explicit AI assi |
2026-09-24 |
| commit |
1.00 |
Fix odd head_dim validation for RoPE configurations (#48524) |
|
Commit message contains explicit AI assi |
2026-09-24 |
| commit |
1.00 |
Fix startup failures: drop pull-requests: read from check_fa |
|
Commit message contains explicit AI assi |
2026-09-23 |
| commit |
1.00 |
Fix startup failures: drop permissions reusable-workflow cal |
|
Commit message contains explicit AI assi |
2026-09-23 |
| commit |
1.00 |
OpenVINO HF Exporter (#47003) |
|
Commit message contains explicit AI assi |
2026-09-22 |
| pr |
1.00 |
Fix MiniMax M3 partial 3D vision rotary embeddings |
|
PR body explicitly mentions AI collabora |
2026-09-28 |
| pr |
1.00 |
[Nemotron3Diarization] Pad the last streaming chunk by n_fft |
|
PR body explicitly mentions AI collabora |
2026-09-30 |
| pr |
1.00 |
MoE: shard the experts in every TP/EP plan, and test for it |
|
PR body explicitly mentions AI collabora |
2026-09-30 |
| pr |
1.00 |
Add a linear attention interface, starting with Mamba2's SSD |
|
PR body explicitly mentions AI collabora |
2026-09-29 |
| pr |
1.00 |
Move processing defaults to class |
|
PR body explicitly mentions AI collabora |
2026-09-24 |
| pr |
1.00 |
Apertus 1.5: add model |
|
PR body explicitly mentions AI collabora |
2026-07-31 |
| pr |
1.00 |
Fix stale _added_tokens_encoder entries in cpmant and wav2ve |
|
PR body explicitly mentions AI collabora |
2026-07-20 |
| pr |
1.00 |
Fix Trainer metrics after checkpoint resume |
|
PR body explicitly mentions AI collabora |
2026-09-30 |
| pr |
1.00 |
Composite lm head tp plan |
|
PR body explicitly mentions AI collabora |
2026-09-24 |
| pr |
1.00 |
Finegrained kernels: unified quantization kernels api (separ |
|
PR body explicitly mentions AI collabora |
2026-09-22 |
| pr |
1.00 |
[MiMo-V2-Flash] Count the query token in the sliding window |
|
PR body explicitly mentions AI collabora |
2026-09-28 |
| pr |
1.00 |
🚨 Exporters Runtimes : Export and run any Transformers model |
|
PR body explicitly mentions AI collabora |
2026-09-21 |
| pr |
1.00 |
Support quantization in Dynamo, ONNX and ET exporters |
|
PR body explicitly mentions AI collabora |
2026-08-03 |
| commit |
0.01 |
Added more usage of MemoryCleanupMixin (#49011) |
|
Terse, repository-specific commit wordin |
2026-09-22 |
| commit |
0.01 |
Keep special token ids the tokenizer does not define (#48708 |
|
Specific tokenizer terminology and incre |
2026-09-22 |
| commit |
0.01 |
Deprecate min-max pixels (#49021) |
|
Informal “lol” and terse test notes stro |
2026-09-22 |
| commit |
0.01 |
Fix assisted decoding for VLM due to dropping attn mask (#4 |
|
Very terse, project-specific fixes with |
2026-09-22 |
| commit |
0.01 |
fix noisy comments (#49013) |
|
Brief technical wording and incremental |
2026-09-22 |
| commit |
0.00 |
Dev update (#49212) |
|
— |
2026-09-30 |
| commit |
0.00 |
model: Add GTE to Transformers (#48416) |
|
— |
2026-09-30 |
| commit |
0.00 |
[Nemotron3Diarization] fix streaming last stft frame dropped |
|
— |
2026-09-30 |
| commit |
0.00 |
Fix missing router_logits in Qwen3.5-MoE and other MoE model |
|
— |
2026-09-30 |
| commit |
0.00 |
Fix MiniMax M3 partial 3D vision rotary embeddings (#49164) |
|
— |
2026-09-30 |
| commit |
0.00 |
Add Strix Halo (gfx1151) Atlas Inference Hub-kernel path for |
|
— |
2026-09-30 |
| commit |
0.00 |
Fix MPS GQA version gating (#49210) |
|
— |
2026-09-30 |
| commit |
0.00 |
Fix additional_special_tokens data loss with extra_special_t |
|
— |
2026-09-30 |
| commit |
0.00 |
Fix stale _added_tokens_encoder entries in cpmant and wav2ve |
|
— |
2026-09-30 |
| commit |
0.00 |
Fix deepstack features for mixed-input (#49177) |
|
— |
2026-09-30 |
| commit |
0.00 |
gguf user defined tokens (#49008) |
|
— |
2026-09-30 |
| commit |
0.00 |
[`Kernels`] Sync mamba version (#49205) |
|
— |
2026-09-30 |
| commit |
0.00 |
Map bare list, tuple and dict annotations to the right JSON |
|
— |
2026-09-30 |
| commit |
0.00 |
Fix SequenceBiasLogitsProcessor edge cases: token id 0 and p |
|
— |
2026-09-30 |
| commit |
0.00 |
Fail fast on eval OOM under `auto_find_batch_size` (#49198) |
|
— |
2026-09-30 |
| commit |
0.00 |
[MPS] Let mps sdpa handle grouped query attention directly ( |
|
— |
2026-09-30 |
| commit |
0.00 |
[Parakeet] Convert NeMo's stochastic depth to layerdrop (#49 |
|
— |
2026-09-30 |
| commit |
0.00 |
Video processors - general maintenance (#48251) |
|
— |
2026-09-30 |
| commit |
0.00 |
update to torch 2.14 (#49199) |
|
— |
2026-09-30 |
| commit |
0.00 |
QA: fix noisy comment checker (#49184) |
|
— |
2026-09-30 |
| commit |
0.00 |
Fix command syntax for optimum-cli export (#46451) |
|
— |
2026-09-29 |
| commit |
0.00 |
Fix Trainer checkpoint resume crashing on CPU with multiple |
|
— |
2026-09-29 |
| commit |
0.00 |
Fix backslash handling in generate flag values in transforme |
|
— |
2026-09-29 |
| commit |
0.00 |
Document image_like_kwargs (#49180) |
|
— |
2026-09-29 |
| commit |
0.00 |
Fix lost call (#49163) |
|
— |
2026-09-29 |
| commit |
0.00 |
[CB] Fix failing tests discovered when using the B200 (#4917 |
|
— |
2026-09-29 |
| commit |
0.00 |
[PerceptionLM] Restore test_inputs_embeds overrides to fix f |
|
— |
2026-09-28 |
| commit |
0.00 |
[MPS] Remove cu_seqlens_k clone workaround for metal-flash-s |
|
— |
2026-09-28 |
| commit |
0.00 |
No more -hf repo names for ESMC (#49158) |
|
— |
2026-09-28 |
| commit |
0.00 |
[generation] Encode multimodal data only once (#45783) |
|
— |
2026-09-28 |
| commit |
0.00 |
[CI] ssh-runner: add optional cache_type input to switch bet |
|
— |
2026-09-28 |
| commit |
0.00 |
Deprecate the use_mamba_kernels config flag that no longer h |
|
— |
2026-09-28 |
| commit |
0.00 |
Pick the default flash implementation based on the current h |
|
— |
2026-09-28 |
| commit |
0.00 |
Auto generate model inits (#47829) |
|
— |
2026-09-28 |
| commit |
0.00 |
Skip flash tests that fall back to a hub kernel when kernels |
|
— |
2026-09-28 |
| commit |
0.00 |
Fix UMT5 decoder self-attention not being causal (#49135) |
|
— |
2026-09-28 |
| commit |
0.00 |
[AMD] Fix some integration tests (#49153) |
|
— |
2026-09-28 |
| commit |
0.00 |
qwen3 models map to wrong tokenizer class on the hub (#49116 |
|
— |
2026-09-28 |
| commit |
0.00 |
Fix exporters import on torch < 2.9 (is_contiguous_or_false) |
|
— |
2026-09-26 |
| commit |
0.00 |
[serge] Fix 2 integration tests for model `jamba` failing wi |
|
— |
2026-09-25 |
| commit |
0.00 |
Use grouped_mm on TPU devices under torch.compile (#49097) |
|
— |
2026-09-25 |
| commit |
0.00 |
Keep the eos ids the config declares (#49082) |
|
— |
2026-09-25 |
| commit |
0.00 |
[serge] Fix 2 integration tests for model `pvt_v2` failing w |
|
— |
2026-09-25 |
| commit |
0.00 |
[serge] Fix 2 integration tests for model `hy_v3` failing wi |
|
— |
2026-09-25 |
| commit |
0.00 |
[serge] Fix 2 integration tests for model `cvt` failing with |
|
— |
2026-09-25 |
| commit |
0.00 |
Fix MaskFormerSwin attention mask dtype to follow hidden sta |
|
— |
2026-09-25 |
| commit |
0.00 |
Fix RGB early-return skipping PNG tRNS compositing (#49005) |
|
— |
2026-09-25 |
| commit |
0.00 |
[Executorch] Add MLX recipe (#48910) |
|
— |
2026-09-25 |
| commit |
0.00 |
Add image processing tester init (#48829) |
|
— |
2026-09-24 |
| commit |
0.00 |
Add `Trainer.end` (#48875) |
|
— |
2026-09-24 |
| commit |
0.00 |
🚨 [ROCm] gpt-oss: route FA3 to aiter-flash-attn, generate RO |
|
— |
2026-09-24 |
| commit |
0.00 |
Keep `attention_mask` as `None` in OPT's causal mask creatio |
|
— |
2026-09-24 |
| commit |
0.00 |
Fix NaN in Parakeet eager attention with padded batches (#49 |
|
— |
2026-09-24 |
| commit |
0.00 |
[serge] Fix OOM in Moshi integration tests with MemoryCleanu |
|
— |
2026-09-24 |
| commit |
0.00 |
Fix assisted eos token condition (#49075) |
|
— |
2026-09-24 |
| commit |
0.00 |
Fix XPU Docker image build (#49073) |
|
— |
2026-09-24 |
| commit |
0.00 |
Add support for Nemotron Omni (#46509) |
|
— |
2026-09-24 |
| commit |
0.00 |
Honor `config.output_router_logits` in the MoE VLM wrappers |
|
— |
2026-09-23 |
| commit |
0.00 |
fix(ci): harden GitHub Actions workflows (#49057) (#49059) |
|
— |
2026-09-23 |
| commit |
0.00 |
[Nemotron3Diarization] nit: hub pr merged to main (#49065) |
|
— |
2026-09-23 |
| commit |
0.00 |
[docs] Loading behavior (#49023) |
|
— |
2026-09-23 |
| commit |
0.00 |
Add MPS maintainer (#49052) |
|
— |
2026-09-23 |
| commit |
0.00 |
Fix assistant masks for processors (#48793) |
|
— |
2026-09-23 |
| commit |
0.00 |
Add Nemotron3Diarization (#49056) |
|
— |
2026-09-23 |
| commit |
0.00 |
Add workflow for building XPU CI Docker images (#49039) |
|
— |
2026-09-23 |
| commit |
0.00 |
Register the remaining mamba-ssm kernel layers on XPU (#4903 |
|
— |
2026-09-23 |
| commit |
0.00 |
QA: Fix llama4 leak (#49042) |
|
— |
2026-09-23 |
| commit |
0.00 |
Only seed numpy in BigBird block-sparse attention during tra |
|
— |
2026-09-23 |
| commit |
0.00 |
Pin GitHub Actions to commit SHAs (#49049) |
|
— |
2026-09-23 |
| commit |
0.00 |
[Chat] Loading GGUF models served with the Chat CLI (#49031) |
|
— |
2026-09-23 |
| commit |
0.00 |
Fix mask creation not being skipped under `torch.compile` (# |
|
— |
2026-09-23 |
| commit |
0.00 |
Scope GITHUB_TOKEN permissions per job (#49046) |
|
— |
2026-09-23 |
| commit |
0.00 |
fix(zamba): add use_associative_scan config flag to avoid t |
|
— |
2026-09-23 |
| commit |
0.00 |
PEFT x Dtensor-based TP integration (#48485) |
|
— |
2026-09-23 |
| commit |
0.00 |
Shorten noisy OpenVINO SDPA comment (#49033) |
|
— |
2026-09-23 |
| pr |
0.00 |
[`distributed`] Add expert-parallel token dispatch, default |
|
— |
2026-09-16 |
| pr |
0.00 |
[serge] Fix 5 integration tests for model `utils` failing wi |
|
— |
2026-09-30 |
| pr |
0.00 |
[`distributed`] Default MoE `ep_plans` to token dispatch |
|
— |
2026-09-28 |
| pr |
0.00 |
[`distributed`]: adapt Trainer to work Expert Parallel |
|
— |
2026-09-16 |
| pr |
0.00 |
fix(utils): support tuple and set typed fields in HfArgument |
|
— |
2026-08-30 |
| pr |
0.00 |
[`distributed`] decoupled tp_plan and ep_plan |
|
— |
2026-09-16 |
| pr |
0.00 |
[`distributed`] remaining ep_plan to update |
|
— |
2026-09-28 |
| pr |
0.00 |
[`distributed`]: Fix DeepSeek-V4 tied embeddings with combin |
|
— |
2026-09-15 |
| pr |
0.00 |
[`distributed`] Add dense and expert device mesh |
|
— |
2026-09-16 |
| pr |
0.00 |
feat: update safetensors with arbitrary slice in prefetch pl |
|
— |
2026-09-30 |
| pr |
0.00 |
feat: use safetensors prefetch api |
|
— |
2026-09-18 |
| pr |
0.00 |
Fix SAM3 video reconditioning masks |
|
— |
2026-09-30 |
| pr |
0.00 |
[Nemotron3Diarization] fix streaming last stft frame dropped |
|
— |
2026-09-28 |
| pr |
0.00 |
correct final metrics due to resuming from checkpoint |
|
— |
2026-09-30 |
| pr |
0.00 |
[i18n-TR] Translate model_sharing.md to Turkish |
|
— |
2026-09-13 |
| pr |
0.00 |
[ESM/Persimmon/NllbMoe/ESMFold] use safetensors repos to fix |
|
— |
2026-09-29 |
| pr |
0.00 |
Why override when you can fix in `Mixin` |
|
— |
2026-09-25 |
| pr |
0.00 |
🚨 Attention 🚨 Deprecate "paged|" prefix for SDPA and flash |
|
— |
2026-09-25 |
| pr |
0.00 |
model: Add GTE to Transformers |
|
— |
2026-08-30 |
| pr |
0.00 |
Processing Utils: honor pre-built sub-processor kwargs in fr |
|
— |
2026-04-24 |
| pr |
0.00 |
[Fix] Fix CSM codebook embedding tying |
|
— |
2026-09-30 |
| pr |
0.00 |
Dev update |
|
— |
2026-09-30 |
| pr |
0.00 |
🚨🚧 FeatureExtractor → AudioProcessor |
|
— |
2026-03-02 |
| pr |
0.00 |
Support per-layer cache configuration and attention-mask sel |
|
— |
2026-08-21 |
| pr |
0.00 |
Add chat input sanitization |
|
— |
2026-07-17 |
| pr |
0.00 |
Add TDT loss kernel |
|
— |
2026-05-19 |
| pr |
0.00 |
Fix missing router_logits in Qwen3.5-MoE and other MoE model |
|
— |
2026-09-29 |
| pr |
0.00 |
Add Strix Halo (gfx1151) Atlas Inference Hub-kernel path for |
|
— |
2026-09-26 |
| pr |
0.00 |
Fix MPS GQA version gating |
|
— |
2026-09-30 |
| pr |
0.00 |
Fix additional_special_tokens data loss with extra_special_t |
|
— |
2026-08-09 |
| pr |
0.00 |
handle empty precompiled charsmap in spm converter |
|
— |
2026-09-30 |
| pr |
0.00 |
[Bugfix] [Modeling] Fix DSV4 norm |
|
— |
2026-07-22 |
| pr |
0.00 |
Add Molmo2 |
|
— |
2026-01-23 |
| pr |
0.00 |
Adding Omnilingual ASR models |
|
— |
2026-01-13 |
| pr |
0.00 |
Fix T5 SentencePiece load when precompiled_charsmap is empty |
|
— |
2026-09-20 |
| pr |
0.00 |
Return router logits from every MoE model that computes them |
|
— |
2026-09-18 |
| pr |
0.00 |
Fix deepstack features for mixed-input |
|
— |
2026-09-29 |
| pr |
0.00 |
Enable model test on upcoming `torch_tpu` backend |
|
— |
2026-09-30 |
| pr |
0.00 |
QA: restore masking_utils export comment with a noqa |
|
— |
2026-09-30 |
| pr |
0.00 |
Fix AutoTokenizer backend selection for CohereLabs tiny-aya |
|
— |
2026-09-29 |
| pr |
0.00 |
gguf user defined tokens |
|
— |
2026-09-22 |
| pr |
0.00 |
[MPS] Remove cu_seqlens_k clone workaround for metal-flash-s |
|
— |
2026-09-24 |
| pr |
0.00 |
[MPS] Let mps sdpa handle grouped query attention directly |
|
— |
2026-09-29 |
| pr |
0.00 |
QA: fix noisy comment checker |
|
— |
2026-09-29 |
| pr |
0.00 |
This PR fixes Gemma3n fp16 loading when missing weights need |
|
— |
2026-09-28 |
| pr |
0.00 |
[`Kernels`] Sync mamba version |
|
— |
2026-09-30 |
| pr |
0.00 |
Perf MoE: Use a GEMM for the expert-bias gather backward in |
|
— |
2026-09-30 |
| pr |
0.00 |
Map bare list, tuple and dict annotations to the right JSON |
|
— |
2026-09-27 |
| pr |
0.00 |
Fix SequenceBiasLogitsProcessor edge cases: token id 0 and p |
|
— |
2026-09-25 |
| pr |
0.00 |
Fail fast on eval OOM under `auto_find_batch_size` |
|
— |
2026-09-30 |
| pr |
0.00 |
Prevent Whisper long-form crashes after Trainer.train() |
|
— |
2026-05-12 |
| pr |
0.00 |
Add FastIntegrationTestMixin and fast integration test class |
|
— |
2026-09-25 |
| pr |
0.00 |
:rotating_light: Infer config class default fields without i |
|
— |
2026-08-25 |
| pr |
0.00 |
Replace datasets that no longer load with maintained uploads |
|
— |
2026-09-22 |
| pr |
0.00 |
[Parakeet] Convert NeMo's stochastic depth to layerdrop |
|
— |
2026-09-29 |
| pr |
0.00 |
Carry weight_global_scale when decompressing NVFP4 MoE exper |
|
— |
2026-08-27 |
| pr |
0.00 |
Video processors - general maintenance |
|
— |
2026-08-24 |
| pr |
0.00 |
update to torch 2.14 |
|
— |
2026-09-30 |
| pr |
0.00 |
QA: Check for a leaked class fixture at the class boundary |
|
— |
2026-09-21 |
| pr |
0.00 |
Restore BLT's 512-byte sliding windows (entropy patcher + lo |
|
— |
2026-09-29 |
| pr |
0.00 |
Select the OWLv2 image query by objectness |
|
— |
2026-09-30 |
| pr |
0.00 |
Register kimi_k2 as a native deepseek_v3 alias |
|
— |
2026-08-27 |
| pr |
0.00 |
Let continuous batching page the call, not the model |
|
— |
2026-09-12 |
| pr |
0.00 |
Give each tensor parallel model in a process its own device |
|
— |
2026-08-28 |
| pr |
0.00 |
Expert-parallel token dispatch: each rank trains on its own |
|
— |
2026-09-04 |
| pr |
0.00 |
Add heterogeneous model support (per-layer config and modeli |
|
— |
2026-04-09 |
| pr |
0.00 |
FSDP x TP Saving / Loading |
|
— |
2026-09-14 |
| pr |
0.00 |
Add WeatherNext 2 weather forecasting model |
|
— |
2026-08-10 |
| pr |
0.00 |
Fix backslash handling in generate flag values in transforme |
|
— |
2026-09-11 |
| pr |
0.00 |
Serialize MPS transfers in async model loading (fix #48029) |
|
— |
2026-09-28 |
| pr |
0.00 |
Fix Trainer checkpoint resume crashing on CPU with multiple |
|
— |
2026-09-26 |
| pr |
0.00 |
Fix command syntax for optimum-cli export |
|
— |
2026-06-05 |
| pr |
0.00 |
[docs] TP for PEFT adapters |
|
— |
2026-09-23 |
| pr |
0.00 |
Finish removing the MPS autocast workaround |
|
— |
2026-09-28 |
| pr |
0.00 |
docs: percent-encode parentheses in DETR notebook links |
|
— |
2026-09-29 |
| pr |
0.00 |
Add MOSS-Transcribe-Diarize model |
|
— |
2026-08-28 |
| pr |
0.00 |
Clip grad norm dtensor |
|
— |
2026-09-29 |
| pr |
0.00 |
Fix parse_generate_flags to properly handle backslashes in s |
|
— |
2026-09-25 |
| pr |
0.00 |
Fix backslash handling in chat generate flags |
|
— |
2026-09-10 |
| pr |
0.00 |
fix: preserve backslashes in generate flag values |
|
— |
2026-09-10 |
| pr |
0.00 |
[docs] Add warning about remove_unused_columns silently drop |
|
— |
2026-09-27 |
| pr |
0.00 |
Preserve full request history for continuous batching logits |
|
— |
2026-09-01 |
| pr |
0.00 |
fix bug for DebertaV2ForMaskedLM non-legacy load issue |
|
— |
2026-09-28 |
| pr |
0.00 |
Document running the example scripts on Hugging Face Jobs |
|
— |
2026-09-23 |
| pr |
0.00 |
[CacheHardIntegrationTest] use dedicated safetensors repo to |
|
— |
2026-09-29 |
| pr |
0.00 |
[generation] Encode multimodal data only once |
|
— |
2026-05-05 |
| pr |
0.00 |
Document image_like_kwargs |
|
— |
2026-09-29 |
