| commit |
1.00 |
[Gemma4] Fix stale expected values in integration tests (#48 |
|
Commit message contains explicit AI assi |
2026-08-23 |
| commit |
1.00 |
[TableTransformer, PI0] Fix stale expected values and OOM in |
|
Commit message contains explicit AI assi |
2026-08-21 |
| commit |
1.00 |
Fix stale expected values in integration tests (cuda sm_86 / |
|
Commit message contains explicit AI assi |
2026-08-21 |
| commit |
1.00 |
Fix `gpt_oss` runs on GPU (#48118) |
|
Commit message contains explicit AI assi |
2026-08-21 |
| commit |
1.00 |
[EsmFold2] Fix stale expected distogram logit values (#48182 |
|
Commit message contains explicit AI assi |
2026-08-21 |
| commit |
1.00 |
Fix Apr 05 integration test regressions (cuda sm_86) (#48170 |
|
Commit message contains explicit AI assi |
2026-08-21 |
| commit |
1.00 |
Fix integration test expected values for cuda sm_86 (Mar 15 |
|
Commit message contains explicit AI assi |
2026-08-21 |
| commit |
1.00 |
[LLaVA] Fix pixtral integration tests for cuda sm_86 (#48166 |
|
Commit message contains explicit AI assi |
2026-08-20 |
| commit |
1.00 |
Port ESMC and ESMFold2 to Transformers (#46419) |
|
Commit message contains explicit AI assi |
2026-08-20 |
| commit |
1.00 |
[InternVL] Fix stale expected values for Llama integration t |
|
Commit message contains explicit AI assi |
2026-08-20 |
| commit |
1.00 |
Retry transient network errors (RemoteDisconnected) in githu |
|
Commit message contains explicit AI assi |
2026-08-20 |
| commit |
1.00 |
[Whisper] Fix speculative decoding: preserve cleared suppres |
|
Commit message contains explicit AI assi |
2026-08-20 |
| commit |
1.00 |
Enable mlinter findings artifact for inline PR reviews (#481 |
|
Commit message contains explicit AI assi |
2026-08-19 |
| commit |
1.00 |
Delete old mlinter review comments before posting new ones ( |
|
Commit message contains explicit AI assi |
2026-08-19 |
| commit |
1.00 |
Fix mlinter artifact path (#48088) |
|
Commit message contains explicit AI assi |
2026-08-19 |
| commit |
1.00 |
[Whisper] Fix decoder position IDs for left-padded batches i |
|
Commit message contains explicit AI assi |
2026-08-19 |
| commit |
1.00 |
[PE] Skip test_sdpa_can_dispatch_on_flash for TimmWrapper-ba |
|
Commit message contains explicit AI assi |
2026-08-18 |
| commit |
1.00 |
[Gemma3] Update integration test expected values for A10G (# |
|
Commit message contains explicit AI assi |
2026-08-18 |
| commit |
1.00 |
[CircleCI] Enable CI for private forks, no-op for public rep |
|
Commit message contains explicit AI assi |
2026-08-18 |
| commit |
1.00 |
Fix EOS for candidate generators (#47931) |
|
Commit message contains explicit AI assi |
2026-08-18 |
| commit |
1.00 |
[Gemma3n] Update integration test expected values for A10G + |
|
Commit message contains explicit AI assi |
2026-08-18 |
| commit |
1.00 |
[Florence2] Fix two integration test failures caused by torc |
|
Commit message contains explicit AI assi |
2026-08-17 |
| commit |
1.00 |
[Whisper] Fix speculative decoding: UnboundLocalError, cache |
|
Commit message contains explicit AI assi |
2026-08-16 |
| commit |
1.00 |
[Whisper] Fix integration test failures on A10G (dtype, stal |
|
Commit message contains explicit AI assi |
2026-08-16 |
| commit |
1.00 |
[DeepSeekV2] Fix integration tests OOM: use device_map=auto |
|
Commit message contains explicit AI assi |
2026-08-15 |
| commit |
1.00 |
[OLMoE] Update expected logits for A10G and add torch.no_gra |
|
Commit message contains explicit AI assi |
2026-08-15 |
| commit |
1.00 |
[OLMo] Fix OOM in logits tests by adding torch.no_grad() (#4 |
|
Commit message contains explicit AI assi |
2026-08-15 |
| commit |
1.00 |
[Gemma] Update expected values for A10G (#47976) |
|
Commit message contains explicit AI assi |
2026-08-14 |
| commit |
1.00 |
[emu3] 🦮 Black Labrador is back! Fix image generation broken |
|
Commit message contains explicit AI assi |
2026-08-14 |
| commit |
1.00 |
Potential fix for code scanning alert no. 267: Artifact pois |
|
Commit message contains explicit AI assi |
2026-08-14 |
| pr |
1.00 |
[Gemma4] Investigate flaky test_generation_beyond_sliding_wi |
|
Explicitly includes a Co-Authored-By tra |
2026-08-23 |
| pr |
1.00 |
[Gemma4] Fix stale expected values in integration tests |
|
Explicit Co-Authored-By trailer credits |
2026-08-23 |
| pr |
1.00 |
Avoid shadowing pipeline function in README examples |
|
PR body explicitly mentions AI collabora |
2026-08-23 |
| pr |
1.00 |
[Zamba2] Fix block_id to use hybrid order for num_mem_blocks |
|
PR body explicitly mentions AI collabora |
2026-08-20 |
| pr |
1.00 |
Fix nested optional kernel fallback resolution |
|
PR body explicitly mentions AI collabora |
2026-08-22 |
| pr |
1.00 |
Add an opt-in per-frame pixel cap (cap_pixels_per_frame) to |
|
PR body explicitly mentions AI collabora |
2026-08-18 |
| pr |
1.00 |
Batch Rebalance Data Sampler |
|
PR body explicitly mentions AI collabora |
2026-07-15 |
| pr |
1.00 |
doc: Fix typo in SigLIP2 Flash Attention code example |
|
PR body explicitly mentions AI collabora |
2026-08-21 |
| pr |
1.00 |
fix: make auto_docstring import-time performance test robust |
|
PR body explicitly mentions AI collabora |
2026-08-20 |
| pr |
1.00 |
Add AutoModelForProteinFolding and register ESMFold2 in Auto |
|
Explicitly states the contribution was g |
2026-08-21 |
| pr |
1.00 |
Fix stale expected values in integration tests (cuda sm_86 / |
|
Explicitly states the contribution was g |
2026-08-21 |
| pr |
1.00 |
[VITS] Un-skip test_model_forward |
|
Explicitly states it was drafted with Cl |
2026-06-03 |
| pr |
1.00 |
Respect config.initializer_range for Conv1D weights in Decis |
|
Explicitly states Claude Code was used d |
2026-08-20 |
| pr |
1.00 |
Fix assisted text-to-audio pipeline initialization |
|
Explicitly acknowledges assistance from |
2026-08-19 |
| pr |
1.00 |
Fix a typo in a use of a local variable field_ in a test |
|
PR body explicitly mentions AI collabora |
2026-08-21 |
| pr |
1.00 |
Fix nested optional kernel fallback resolution |
|
PR body explicitly mentions AI collabora |
2026-08-21 |
| pr |
1.00 |
[EsmFold2] Fix stale expected distogram logit values |
|
Explicitly states the changes were gener |
2026-08-21 |
| pr |
1.00 |
Feat/laguna quantized checkpoints support |
|
Explicitly states the PR was implemented |
2026-08-20 |
| pr |
1.00 |
Add configurable Laguna router scoring |
|
Explicitly states the PR was implemented |
2026-08-19 |
| pr |
1.00 |
Add fromjson Jinja filter for chat templates |
|
Explicitly identifies Cursor as having b |
2026-08-19 |
| pr |
1.00 |
Skip caching allocator warmup when the device cannot report |
|
Explicitly discloses preparation with AI |
2026-08-19 |
| pr |
1.00 |
Fix segment-model tokenizers not returning token_type_ids by |
|
Explicitly states the patch was prepared |
2026-08-19 |
| pr |
1.00 |
Fix PyAV video sampling when stream.frames is 0 |
|
Explicitly discloses assistance by a cod |
2026-08-19 |
| pr |
1.00 |
Raise on non-zero ffmpeg exit in _ffmpeg_stream |
|
Explicitly discloses assistance by a cod |
2026-08-19 |
| pr |
1.00 |
Fix frame_sampling_rate ignored on video_processor path |
|
Explicitly discloses assistance by a cod |
2026-08-19 |
| pr |
1.00 |
Fix ASR return_timestamps=False ignored + ffmpeg silent fail |
|
Explicitly states it was prepared with a |
2026-08-18 |
| pr |
1.00 |
Add Unlimited OCR |
|
PR body explicitly mentions AI collabora |
2026-06-23 |
| pr |
0.98 |
feat(utils): add ProductionDebtEvaluator and TechnicalDueDil |
|
Buzzword-heavy commercial framing strong |
2026-08-18 |
| pr |
0.68 |
Fix: add num_hidden_layers property to T5GemmaConfig and add |
|
Generic polished summary and testing cla |
2025-09-22 |
| pr |
0.68 |
fix(mluke): correct special token mapping for entity tasks |
|
Formal sectioning and generic verificati |
2026-08-23 |
| pr |
0.62 |
fix(mluke): correct special token mapping for entity tasks |
|
Highly polished, repetitive explanation |
2026-08-23 |
| pr |
0.50 |
Add support for BharatGen's Param2MoE model architecture |
|
Polished, generic model-marketing langua |
2026-06-04 |
| pr |
0.40 |
fix(pipelines): handle empty container inputs and validate b |
|
Polished, formulaic summary, but lacks c |
2026-08-22 |
| pr |
0.40 |
fix(generation): validate non-negative bounds for min_length |
|
Formulaic validation summary, though pla |
2026-08-22 |
| pr |
0.38 |
fix(trainer): pass best_model_checkpoint to sort_checkpoints |
|
Concise technical explanation is structu |
2026-08-22 |
| pr |
0.38 |
fix(tokenization): prioritize slice indexing in BatchEncodin |
|
Generic polished changelog wording witho |
2026-08-22 |
| pr |
0.35 |
fix(tokenization): preserve additional_special_tokens when e |
|
Detailed CI discussion has somewhat gene |
2026-08-21 |
| pr |
0.25 |
fix(zamba): expose use_associative_scan config flag in Zamba |
|
Detailed, domain-specific fix descriptio |
2026-08-19 |
| pr |
0.20 |
fix(nanochat): add TRUST_REMOTE_CODE guard before pickle.loa |
|
Specific security rationale and code ref |
2026-08-19 |
| pr |
0.15 |
Add native TR-HASH model support |
|
Highly polished but contains substantial |
2026-08-21 |
| pr |
0.15 |
🚨 TP dtensor API inference + training |
|
Formal wording is present, but technical |
2026-07-27 |
| pr |
0.15 |
update docs for Chameleon.md |
|
Specific documentation changes and issue |
2026-08-21 |
| pr |
0.14 |
fix(glm): validate num_attention_heads is divisible by num_k |
|
Polished but grounded in specific GLM be |
2026-08-21 |
| pr |
0.12 |
add HyperClovaX Vision |
|
Specific migration details and slightly |
2026-02-27 |
| pr |
0.12 |
Fix MPS async loading segfault when converting dtype |
|
Formal structure, but reproduction detai |
2026-08-21 |
| pr |
0.10 |
Add Ovis2.5 model support |
|
Specific model, issue, and implementatio |
2026-07-30 |
| pr |
0.09 |
Fix loss inflation under gradient accumulation for ForImageC |
|
Very polished and systematic, but ground |
2026-08-22 |
| commit |
0.08 |
[new model] step 3.7 (#46658) |
|
Terse, domain-specific commit history wi |
2026-08-16 |
| pr |
0.08 |
Let gradient checkpointing skip layers with every_n_layers |
|
Technical history, benchmarks, and conve |
2026-08-21 |
| pr |
0.08 |
Add heterogeneous model support (per-layer config and modeli |
|
Highly detailed and domain-specific, wit |
2026-04-09 |
| pr |
0.08 |
[ONNX] Skip affected models on torch 2.13 (two dynamo regres |
|
Highly specific regression details and i |
2026-08-21 |
| pr |
0.08 |
Fix AutoTokenizer returning TokenizersBackend for DeepSeek-R |
|
Specific regression history, commit refe |
2026-08-22 |
| pr |
0.08 |
Post two CI badges on a PR: CPU PR CI and GPU run-slow |
|
Polished but highly specific engineering |
2026-08-21 |
| pr |
0.08 |
Fix DataCollatorForTokenClassification numpy label handling |
|
Terse, concrete bug summary and test pla |
2026-08-19 |
| pr |
0.06 |
[Umbrella] Enable FSDP2 + expert parallelism via a 2-D (fsdp |
|
Concrete benchmarks, implementation deta |
2026-08-22 |
| pr |
0.06 |
Fix InternVL NumPy cumsum tensor conversion |
|
Specific technical details and concise v |
2026-08-23 |
| pr |
0.06 |
Skip embeddings without `.weight` in `_init_weights` |
|
Concrete failure mode, implementation fi |
2026-08-19 |
| pr |
0.06 |
Add multimodal sequence classification support for Mistral 3 |
|
Specific technical scope and tests are c |
2026-08-21 |
| pr |
0.05 |
fix: parse dictionary dataclass arguments as JSON |
|
Domain-specific changes and concrete loc |
2026-08-23 |
| pr |
0.05 |
Warn once when a hub-kernel function falls back to its refer |
|
Highly technical and specific; no clear |
2026-08-21 |
| pr |
0.05 |
Honor per-param `_is_hf_initialized` flags for built-in mode |
|
Highly specific FSDP/ZeRO details and me |
2026-08-20 |
| commit |
0.04 |
Support `BatchFeature` in length-grouped samplers (#48034) |
|
Short technical title is plausible as a |
2026-08-18 |
| commit |
0.04 |
Let the GPU verify caller turn on the memory probe (#48001) |
|
Detailed, domain-specific explanation la |
2026-08-17 |
| commit |
0.04 |
[Whisper] Fix batch decode_with_timestamps in WhisperTokeniz |
|
Concise technical commit message; no AI- |
2026-08-16 |
| commit |
0.04 |
Fix CLIP _init_weights when a child module carries quantized |
|
Detailed technical explanation is polish |
2026-08-14 |
| pr |
0.04 |
[generate] Fix CUDAGraphs tensor overwrite for encoder-decod |
|
Detailed root-cause analysis and impleme |
2026-08-21 |
| pr |
0.04 |
Add WeatherNext 2 weather forecasting model |
|
Domain-specific detail, candid uncertain |
2026-08-10 |
| pr |
0.04 |
Fix nested optional kernel silent fallback |
|
Brief, specific bug description without |
2026-08-22 |
| pr |
0.04 |
[TableTransformer, PI0] Fix stale expected values and OOM in |
|
Specific GPU, optimizer, dtype, and test |
2026-08-21 |
| pr |
0.04 |
fix(processing): convert tensor to numpy array before np.cum |
|
Specific processor names and NumPy behav |
2026-08-21 |
| commit |
0.03 |
🚨[wav2vec2] Support attn_implementation=sdpa dispatch (#4619 |
|
Technical incremental commit notes and d |
2026-08-20 |
| commit |
0.03 |
Always tie embeddings for LongT5 and Pop2Piano (#47620) |
|
Detailed, technically specific debugging |
2026-08-20 |
| commit |
0.03 |
Improve group images by shape (#47964) |
|
Very terse wording and a typo are strong |
2026-08-19 |
| commit |
0.03 |
fix(pipeline): preserve model.generation_config precedence o |
|
Concise technical fix with review feedba |
2026-08-18 |
| commit |
0.03 |
Fix GatedDeltaNet A_log dtype to prevent -inf under bfloat16 |
|
Includes concrete dtype behavior and rev |
2026-08-14 |
| pr |
0.03 |
Init the process group with a load-scaled timeout for sharde |
|
Terse, concrete operational details and |
2026-08-23 |
| pr |
0.03 |
Warm the page cache before sharded loading (opt-in via `HF_S |
|
Specific benchmarks, filesystem details, |
2026-08-23 |
| pr |
0.03 |
Support per-layer cache configuration and attention-mask sel |
|
Concise, repository-specific explanation |
2026-08-21 |
| pr |
0.03 |
[mLUKE] Un-skip MLukeTokenizerTest |
|
Highly specific measurements, commands, |
2026-08-22 |
| pr |
0.03 |
Fix LlamaTokenizer overwriting pre_tokenizer and decoder fro |
|
Detailed root-cause analysis uses precis |
2026-08-22 |
| pr |
0.03 |
Warn when the RoPE rotary dimension is odd |
|
Nuanced technical reasoning and a casual |
2026-08-22 |
| pr |
0.03 |
fix(data_collator): align TokenClassification numpy_call wit |
|
Specific behavioral differences and test |
2026-08-22 |
| pr |
0.03 |
Fix expert parallelism through Trainer |
|
Informal, highly technical failure analy |
2026-08-22 |
| pr |
0.03 |
Skip the expert-parallel sentinel masking when expert parall |
|
Concise domain-specific performance anal |
2026-08-21 |
| pr |
0.03 |
[Optimization] Avoid redundant token length initialization l |
|
Terse optimization description with prec |
2026-08-21 |
| pr |
0.03 |
Refactor GGUF to speed-up inference |
|
Specific implementation details and info |
2026-08-04 |
| pr |
0.03 |
Fix build_2d_sinusoidal_position_embedding on MPS |
|
Concise, issue-specific explanation with |
2026-08-11 |
| pr |
0.03 |
Fix nemotron_h save_pretrained emitting singular backbone.em |
|
Detailed checkpoint-specific diagnosis r |
2026-08-19 |
| pr |
0.03 |
Assign a reviewer even when a codeowner has left, and route |
|
Specific workflow details and audit find |
2026-08-19 |
| pr |
0.03 |
Pipeline parallel naive inference |
|
Informal wording, typos, and terse plann |
2026-07-13 |
| pr |
0.03 |
Fix VoxtralRealtime rejecting non-static cache implementatio |
|
Focused reproduction and framework-speci |
2026-08-19 |
| pr |
0.03 |
Validate even rotary head dimensions |
|
Brief, issue-linked, domain-specific fix |
2026-08-19 |
| commit |
0.02 |
Post two CI badges on a PR: CPU PR CI and GPU run-slow (#481 |
|
Specific CI behavior and issue reference |
2026-08-21 |
| commit |
0.02 |
[Qwen2.5-Omni] Update stale expected values for cuda sm_86 ( |
|
Concise, domain-specific commit text wit |
2026-08-20 |
| commit |
0.02 |
[Mistral3] Fix batched integration tests: padding_side=left |
|
Specific test and hardware details indic |
2026-08-20 |
| commit |
0.02 |
Use generator with seed for LengthGroupedSampler in Trainer. |
|
Specific implementation details and ters |
2026-08-19 |
| commit |
0.02 |
Fix Qwen2.5-Omni / Qwen3-Omni-MoE generation with a compilab |
|
Model-specific implementation details an |
2026-08-19 |
| commit |
0.02 |
feat: add nvfp4 quantization (#47883) |
|
Specific feature and documentation commi |
2026-08-18 |
| commit |
0.02 |
[Fix] Small FA-related test failures in CB (#47341) |
|
Extremely terse, informal fix message la |
2026-08-18 |
| commit |
0.02 |
fix: honor empty processor_kwargs={} in multimodal pipelines |
|
Specific bug mechanism, variable names, |
2026-08-18 |
| commit |
0.02 |
[docs] Muse Glimmer (#47882) |
|
Extremely brief, informal entries are mo |
2026-08-17 |
| commit |
0.02 |
[CI] Fix startup failure in pr_build_doc_with_comment workfl |
|
Specific CI workflow and dependency deta |
2026-08-17 |
| commit |
0.02 |
Bump default flash-attn2 hub kernel version to v3 (#47863) |
|
Detailed kernel, Torch, and XPU referenc |
2026-08-17 |
| commit |
0.02 |
[GPTNeoX] Fix post_processor not overridden when loading fro |
|
Highly specific tokenizer regression ana |
2026-08-15 |
| commit |
0.02 |
Remove stale (None, None) fallback in qwen2_5_vl batch_diffe |
|
Specific stale-value and hardware detail |
2026-08-14 |
| commit |
0.02 |
Scan a diff in trufflehog, not the whole repo history (#4794 |
|
Highly specific debugging details and me |
2026-08-13 |
| pr |
0.02 |
Handle empty FSDP shards in sharded loading: crash and hang |
|
Detailed reproduction steps and precise |
2026-08-23 |
| pr |
0.02 |
Fix Mllama vision/cross-attention being treated as causal un |
|
Specific code paths, issue references, a |
2026-08-22 |
| pr |
0.02 |
Fix expert-parallel training: NaN gradients and missing grad |
|
Dense domain-specific debugging evidence |
2026-08-22 |
| pr |
0.02 |
[WIP] Support custom kernels for processing ops |
|
Casual phrasing, technical specifics, an |
2026-06-19 |
| pr |
0.02 |
Simplify group images by shape |
|
Terse wording and minor grammar issues s |
2026-08-21 |
| pr |
0.02 |
🚨[wav2vec2] Support attn_implementation=sdpa dispatch |
|
Brief technical summary with domain-spec |
2026-05-25 |
| pr |
0.02 |
Simplify group_images_by_shape |
|
Terse wording and minor grammar issues s |
2026-08-21 |
| pr |
0.02 |
Apply context parallelism to the evaluation path |
|
Detailed reproduction and implementation |
2026-08-20 |
| pr |
0.02 |
Ignore causal masks a bit earlier for sdpa |
|
Concise, technical description with info |
2026-08-13 |
| pr |
0.02 |
Fix `gpt_oss` runs on GPU |
|
Concrete memory measurements and CI deta |
2026-08-19 |
| commit |
0.01 |
Assign a reviewer even when a codeowner has left, and route |
|
Highly domain-specific details, metrics, |
2026-08-21 |
| commit |
0.01 |
Fix ASR pipeline mono conversion for channels-last audio (fi |
|
Detailed domain-specific debugging and r |
2026-08-19 |
| commit |
0.01 |
Fix CpmAnt loading: size lm_head to vocab_size (#48012) |
|
Highly specific checkpoint and tensor-sh |
2026-08-19 |
| commit |
0.01 |
[Video] Warn and return all frames when num_frames exceeds t |
|
Repetitive review notes and CI adjustmen |
2026-08-19 |
| commit |
0.01 |
Accept artifact dir as argument in post_mlinter_review.py (# |
|
Concise repository-specific follow-up wi |
2026-08-19 |
| commit |
0.01 |
[Video] Fix convert_to_rgb channel slicing and alpha blendin |
|
Specific review feedback and enum/testin |
2026-08-19 |
| commit |
0.01 |
Fallback from 'lanczos' to 'bicubic' when on cuda (#48026) |
|
Terse commits, typo, and informal wordin |
2026-08-18 |
| commit |
0.01 |
[GPT2] Fix encoder_attention_mask being silently discarded i |
|
Detailed domain-specific diagnosis and c |
2026-08-18 |
| commit |
0.01 |
fix failed test cases for muse_glimmer (#48011) |
|
Brief test-fix commits include specific |
2026-08-18 |
| commit |
0.01 |
Cohere compass tests (#47895) |
|
Terse exploratory test notes are typical |
2026-08-18 |
| commit |
0.01 |
docs: use relative paths for README language menus and add f |
|
Terse, specific commit-style wording wit |
2026-08-18 |
| commit |
0.01 |
Moving mlinter to 0.1.4 (#47918) |
|
Informal commit history and standard hum |
2026-08-18 |
| commit |
0.01 |
Fix sliding window cache index off-by-one on wraparound (#47 |
|
Concise technical fix description with a |
2026-08-18 |
| commit |
0.01 |
[MoE] Fix Blackwell GPU crash with torch._grouped_mm on torc |
|
Specific GPU versions and architecture d |
2026-08-17 |
| commit |
0.01 |
Proper separation of tests (#47943) |
|
Terse iterative notes and a revert refer |
2026-08-17 |
| commit |
0.01 |
unpin `pytest` in the `examples_torch` deps (#48023) |
|
Informal typo-filled commit notes strong |
2026-08-17 |
| commit |
0.01 |
Fix MTP config when mlp_layer_types is absent (#48015) |
|
Brief fix note with a simple follow-up i |
2026-08-17 |
| commit |
0.01 |
Remove duplicate block_sparse_moe assignment in GraniteMoeDe |
|
A concise, repository-specific change ti |
2026-08-17 |
| commit |
0.01 |
Align logit distributions for CandidateGenerators using samp |
|
Short informal iterative notes are typic |
2026-08-17 |
| commit |
0.01 |
[ModernVBERT] Fix integration test checkpoint (404 since Apr |
|
Terse, domain-specific fix description w |
2026-08-17 |
| commit |
0.01 |
Fix DFlash candidate token device mismatch with device_map=" |
|
Technical commit series with informal up |
2026-08-17 |
| commit |
0.01 |
:red_circle: Allow tokenizers 0.23.1 (#46381) |
|
Detailed, repository-specific dependency |
2026-08-17 |
| commit |
0.01 |
[AXK1] Fix expected logits for CUDA A10G (#47980) |
|
Specific hardware and test details indic |
2026-08-14 |
| commit |
0.01 |
[docs] torchcodec + vision processor features (#47935) |
|
Terse commit summaries with no AI-style |
2026-08-14 |
| commit |
0.01 |
Fix gemma4 video to device (#47896) |
|
Brief, repository-specific fix descripti |
2026-08-14 |
| commit |
0.01 |
Fix Jinja block endings in CHAT WITH MODELS' Writing a chat |
|
Short, specific documentation fix with n |
2026-08-14 |
| commit |
0.01 |
[tests] Fix expected output for Qwen2.5-VL batch_wo_image on |
|
Concrete CUDA output changes and test na |
2026-08-14 |
| commit |
0.01 |
Declare sdpa support in `TimmWrapper` (#47939) |
|
Terse, specific implementation summary i |
2026-08-14 |
| pr |
0.01 |
Fix T5 tied weights order for lm_head |
|
Very terse wording and typographical esc |
2026-08-23 |
| pr |
0.01 |
Processing tests [part 2] |
|
Informal phrasing, typos, and personal p |
2026-08-12 |
| pr |
0.01 |
Add Molmo2 |
|
Almost entirely template content, with n |
2026-01-23 |
| pr |
0.01 |
deterministic layer_types buffer registration in multiple mo |
|
Typos, informal phrasing, and detailed d |
2026-08-20 |
| commit |
0.00 |
[VITS] Un-skip test_model_forward (#46375) |
|
Terse commit message with a human co-aut |
2026-08-21 |
| commit |
0.00 |
Apply context parallelism to the evaluation path (#48167) |
|
Brief, technical commit message with no |
2026-08-21 |
| commit |
0.00 |
🚨 TP dtensor API inference + training (#47579) |
|
Long series of terse commits, typos, and |
2026-08-21 |
| commit |
0.00 |
Add MLU support to is_flash_linear_attention_available (#469 |
|
Concise domain-specific commit title wit |
2026-08-21 |
| commit |
0.00 |
Fix ROCm SDPA-flash skip guard that crashes on RDNA GPUs (#4 |
|
Terse technical commit message with an i |
2026-08-21 |
| commit |
0.00 |
Fix DynamicCache reconstruction during ExecuTorch export (#4 |
|
Iterative technical commits and human co |
2026-08-21 |
| commit |
0.00 |
Revert "Support per-layer cache configuration and attention- |
|
Standard revert message with direct comm |
2026-08-21 |
| commit |
0.00 |
Support per-layer cache configuration and attention-mask sel |
|
Brief technical commit series with a hum |
2026-08-21 |
| commit |
0.00 |
Fix DeepSeek V2 default vocab size (#48159) |
|
Terse conventional fix message with a hu |
2026-08-20 |
| commit |
0.00 |
Revert "[Quantization]: Refactor is_quantization_compressed |
|
Standard terse revert message. |
2026-08-20 |
| commit |
0.00 |
fix bugs for clvp model (#47127) |
|
Terse commit series with repeated human |
2026-08-20 |
| commit |
0.00 |
[serge] Fix 2 integration tests for model `generation` faili |
|
— |
2026-08-18 |
| commit |
0.00 |
[serge] Fix 2 integration tests regressed by commit b9090ae5 |
|
— |
2026-08-18 |
| commit |
0.00 |
:rotating_light: Leftover processors (#47924) |
|
Fragmentary informal notes, typos, and e |
2026-08-18 |
| commit |
0.00 |
Fix cropping (#48006) |
|
Extremely terse and informal wording str |
2026-08-17 |
| commit |
0.00 |
[serge] Fix 2 integration tests for model `got_ocr2` failing |
|
— |
2026-08-14 |
| commit |
0.00 |
[serge] Fix 2 integration tests for model `opt` failing with |
|
— |
2026-08-14 |
| commit |
0.00 |
[serge] Fix 2 integration tests for model `vivit` failing wi |
|
— |
2026-08-13 |
| pr |
0.00 |
[serge] Fix 1 integration test for model `moshi` failing wit |
|
— |
2026-08-22 |
| pr |
0.00 |
[serge] Fix 2 integration tests regressed by commit 6034e90c |
|
— |
2026-08-22 |
| pr |
0.00 |
Bump the actions group across 1 directory with 14 updates |
|
— |
2026-08-10 |
| pr |
0.00 |
[serge] Fix 4 integration tests for model `generation` faili |
|
— |
2026-08-19 |
