diff --git a/analysis/current/analysis-report-hybrid.json b/analysis/current/analysis-report-hybrid.json index cc7bded6d0a5021b5ceec7cee60ccf71f7b4ab9d..b16ba19c0541933795ddabfbeb0df3141d3cbd10 100644 --- a/analysis/current/analysis-report-hybrid.json +++ b/analysis/current/analysis-report-hybrid.json @@ -1,8 +1,8 @@ { "schema_version": "1.0", "repo": "huggingface/transformers", - "snapshot_id": "20260525T060031Z", - "generated_at": "2026-05-25T06:10:05Z", + "snapshot_id": "20260525T120039Z", + "generated_at": "2026-05-25T12:15:43Z", "evidence_quality": "full", "llm_enrichment": true, "meta_bugs": [ @@ -1756,31 +1756,277 @@ ] }, { - "cluster_id": "cluster-44002-2", + "cluster_id": "cluster-44018-2", "summary": "Cluster of 2 related pull requests linked by soft_similarity.", "status": "open", "confidence": 0.5, "canonical_issue_number": null, - "canonical_pr_number": 44002, + "canonical_pr_number": 44068, "issue_numbers": [], "pr_numbers": [ - 44002, - 44029 + 44018, + 44068 ], "evidence_types": [ "soft_similarity" ], "pr_comparisons": [ { - "left_pr_number": 44002, - "right_pr_number": 44029, - "code_similarity": 0.135, - "size_similarity": 0.43, - "file_overlap": 0.0, - "area_overlap": 0.0, - "patch_similarity": 0.329, - "shared_filenames": [], - "shared_file_areas": [] + "left_pr_number": 44018, + "right_pr_number": 44068, + "code_similarity": 0.766, + "size_similarity": 0.939, + "file_overlap": 1.0, + "area_overlap": 0.425, + "patch_similarity": 0.866, + "shared_filenames": [ + "src/transformers/models/gpt_neo/modeling_gpt_neo.py", + "tests/models/gpt_neo/test_modeling_gpt_neo.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/models/gpt_neo/modeling_gpt_neo.py", + "left_ranges": [ + [ + 26, + 46 + ], + [ + 138, + 143 + ], + [ + 182, + 187 + ], + [ + 283, + 295 + ], + [ + 327, + 341 + ], + [ + 348, + 354 + ], + [ + 360, + 369 + ], + [ + 399, + 406 + ], + [ + 411, + 419 + ], + [ + 428, + 433 + ], + [ + 472, + 492 + ], + [ + 507, + 513 + ], + [ + 519, + 528 + ], + [ + 541, + 559 + ], + [ + 562, + 567 + ], + [ + 595, + 601 + ], + [ + 607, + 614 + ], + [ + 627, + 643 + ], + [ + 685, + 690 + ], + [ + 708, + 714 + ], + [ + 720, + 727 + ], + [ + 740, + 757 + ], + [ + 761, + 766 + ], + [ + 780, + 786 + ], + [ + 791, + 798 + ], + [ + 807, + 822 + ], + [ + 840, + 845 + ] + ], + "right_ranges": [ + [ + 26, + 31 + ], + [ + 34, + 44 + ], + [ + 136, + 141 + ], + [ + 180, + 185 + ], + [ + 281, + 293 + ], + [ + 325, + 339 + ], + [ + 346, + 352 + ], + [ + 358, + 367 + ], + [ + 397, + 404 + ], + [ + 409, + 417 + ], + [ + 426, + 434 + ], + [ + 463, + 483 + ], + [ + 498, + 504 + ], + [ + 510, + 519 + ], + [ + 532, + 550 + ], + [ + 553, + 564 + ], + [ + 586, + 592 + ], + [ + 598, + 605 + ], + [ + 618, + 634 + ], + [ + 676, + 688 + ], + [ + 699, + 705 + ], + [ + 711, + 718 + ], + [ + 731, + 748 + ], + [ + 752, + 762 + ], + [ + 771, + 777 + ], + [ + 782, + 789 + ], + [ + 798, + 813 + ], + [ + 831, + 836 + ] + ] + }, + { + "filename": "tests/models/gpt_neo/test_modeling_gpt_neo.py", + "left_ranges": [ + [ + 458, + 464 + ] + ], + "right_ranges": [ + [ + 458, + 464 + ] + ] + } + ] } ] } @@ -1841,13 +2087,13 @@ "reason": "PRs in cluster-43979-11 are treated as duplicates because they converge on issue #43979 with closing_reference, shared_issue_target, soft_similarity evidence." }, { - "cluster_id": "cluster-44002-2", - "canonical_pr_number": 44002, + "cluster_id": "cluster-44018-2", + "canonical_pr_number": 44068, "duplicate_pr_numbers": [ - 44029 + 44018 ], "target_issue_number": null, - "reason": "PRs in cluster-44002-2 are treated as duplicates because they share soft_similarity evidence." + "reason": "PRs in cluster-44018-2 are treated as duplicates because they share soft_similarity evidence." }, { "cluster_id": "cluster-45081-3", diff --git a/analysis/current/analysis-report-hybrid.llm-reviews.json b/analysis/current/analysis-report-hybrid.llm-reviews.json index 511943191af01e0559b14d1df0e49b5e011bd453..8cc599bb51da47d87dc3be7d13c2807c44127c4a 100644 --- a/analysis/current/analysis-report-hybrid.llm-reviews.json +++ b/analysis/current/analysis-report-hybrid.llm-reviews.json @@ -1,8 +1,8 @@ { "schema_version": "1.0", "repo": "huggingface/transformers", - "snapshot_id": "20260525T060031Z", - "generated_at": "2026-05-25T06:10:05Z", + "snapshot_id": "20260525T120039Z", + "generated_at": "2026-05-25T12:15:43Z", "model": "gpt-5.4-mini?service_tier=flex", "reviews": [ { @@ -10,13 +10,13 @@ "component_index": 1, "component_count": 17, "review_unit_index": 1, - "review_unit_count": 49, + "review_unit_count": 50, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-427", + "cluster_id": "cluster-10000-428", "nodes": [ "issue:24643", "issue:30277", @@ -138,20 +138,20 @@ "issue:43866|issue:44863", "issue:44292|issue:46015" ], - "prepared_review_unit_hash": "968da5f4c898b0192f34d362b57e16bbea95d5141a071a16e2108502ce368b99", + "prepared_review_unit_hash": "f9259cba0923024ffa657b5f2d2363fbc928d0316cbe6b0d68e1f33f51b3e507", "component_budget": { - "node_count": 427, - "item_count": 427, - "soft_pair_count": 1576, - "serialized_chars": 816182, - "estimated_input_tokens": 204046, - "estimated_eval_tokens": 408348 + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 69, - "serialized_chars": 48404, + "serialized_chars": 48403, "estimated_input_tokens": 12101, "estimated_eval_tokens": 24458 }, @@ -160,390 +160,420 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This set is mostly heterogeneous, with only a few true duplicates around a ZeRO-3 empty-index loader failure and an AutoTokenizer v5 token-id regression. Most other soft edges are broad theme matches, not the same bug.", - "confidence": 0.64, - "canonical_issue_reason": "Issue 45920 is the broadest, most central regression report here: it captures the AutoTokenizer v5 wrong-token-id problem across multiple model families and is open, so it can serve as the umbrella issue for that subcluster.", + "summary": "Mostly heterogeneous cluster. The only clear duplicate subgroups are the DeepSpeed ZeRO-3 empty-index failures, the AutoTokenizer token-ID regression, and possibly the Qwen3_5Moe rotary-embedding device/materialization bug. Most other soft edges are different bugs or unrelated docs/issues.", + "confidence": 0.58, + "canonical_issue_reason": "No single issue covers the whole set; issue 45920 is the best canonical anchor because it is the broadest report of the tokenizer-ID regression and generalizes beyond one model family.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 45920 is the best representative issue because it describes the widest blast radius of the tokenizer regression and is phrased as a general v5 bug rather than a model-specific symptom.", + "best_issue_reason": "Issue 45920 is the strongest overall representative: broad symptom, current regression framing, and it already aggregates the tokenizer bug across multiple models.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44322", "right": "issue:45464", "accept": false, - "reason": "Different failures: missing config attribute for Qwen3.5 vs chat/completions streaming inference." + "reason": "Different failures: a config attribute error vs a streaming chat/completions bug." }, { "left": "issue:45237", "right": "issue:45685", "accept": false, - "reason": "Unrelated hardware/backend errors: AMD GPU failure vs MPS histogram not implemented." + "reason": "Different backends and code paths: AMD GPU runtime failure vs MPS histogram int issue." }, { "left": "issue:43716", "right": "issue:45237", "accept": false, - "reason": "Different models and symptoms; device mismatch in Mistral-3 is not the AMD GPU failure in GPT-OSS." + "reason": "Image-preprocessor dtype mismatch is not the same bug as the AMD GPU device failure." }, { "left": "issue:43828", "right": "issue:45237", "accept": false, - "reason": "Both are runtime errors, but the code paths are different: autocast dtype mismatch vs AMD GPU/device issue." + "reason": "Autocast dtype mismatch and AMD GPU device failure are separate problems." }, { "left": "issue:45237", "right": "issue:45507", "accept": false, - "reason": "Different bugs: AMD GPU support failure vs GraniteMoEHybrid invalid method call." + "reason": "Unrelated GPT-OSS AMD issue vs a GraniteMoE invalid-method call." }, { "left": "issue:44292", "right": "issue:45237", "accept": false, - "reason": "Different model-specific failures with no shared code-path evidence." + "reason": "Different model/back-end failures; no shared underlying bug." }, { "left": "issue:30277", "right": "issue:35545", "accept": false, - "reason": "DeepSpeed ZeRO-3 collective mismatch and ONNX export failure are unrelated." + "reason": "DeepSpeed collective mismatch and ONNX export failure are unrelated." }, { "left": "issue:35545", "right": "issue:42915", "accept": false, - "reason": "Different subsystems: export to ONNX vs FineGrainedFP8 model failure." + "reason": "ONNX export error vs FineGrainedFP8Config failure are different code paths." }, { "left": "issue:35545", "right": "issue:41720", "accept": false, - "reason": "Export bug vs runtime device-map CUDA assert; not the same issue." + "reason": "Export-time ONNX issue is unrelated to the Qwen3 auto-device-map CUDA assert." }, { "left": "issue:45901", "right": "issue:46015", "accept": false, - "reason": "Table QA crash and outdated RAG docs are unrelated." + "reason": "Table-QA runtime crash is unrelated to outdated RAG docs." }, { "left": "issue:44509", "right": "issue:44512", "accept": false, - "reason": "Both are v5 docs cleanup, but they reference different removed features and are not the same doc bug." + "reason": "Both are v5 docs regressions, but they concern different removed features and are not the same concrete fix." }, { "left": "issue:35545", "right": "issue:36010", "accept": false, - "reason": "ONNX export failure and GenerationMixin import error are different problems." + "reason": "ModernBERT ONNX export and GenerationMixin import failure are unrelated." }, { "left": "issue:44805", "right": "issue:45823", "accept": false, - "reason": "Mask shape mismatch and Gemma4 device-map mismatch are different failures." + "reason": "Mask/tensor shape mismatch is not the same as Gemma4 PLE device placement." }, { "left": "issue:44918", "right": "issue:46165", "accept": false, - "reason": "TRl SFT embedding unpacking failure and resume-training/trust-remote-code issue are unrelated." + "reason": "TRL SFT input-embedding unpacking failure and training-resume/trust-remote-code issue are different bugs." }, { "left": "issue:43638", "right": "issue:44805", "accept": false, - "reason": "Both are shape/index errors, but one is ZeRO-3 Bert loading and the other is a mask mismatch." - }, - { - "left": "issue:43638", - "right": "issue:46165", - "accept": false, - "reason": "Different training/loading failures with no shared underlying bug." - }, - { - "left": "issue:43596", - "right": "issue:44805", - "accept": false, - "reason": "ZeRO-3 empty-index bug vs generic mask shape mismatch are not the same issue." + "reason": "ZeRO-3 empty-index loading error is unrelated to the mask/index shape mismatch." }, { "left": "issue:44805", "right": "issue:45161", "accept": false, - "reason": "Mask/index mismatch and GPT-OSS TP-only failure are unrelated." + "reason": "Different failures: shape mismatch vs GPT-OSS tensor-parallel behavior." }, { - "left": "issue:44560", + "left": "issue:43638", "right": "issue:46165", "accept": false, - "reason": "Video StopIteration in Qwen3-vl and resume-training issue are unrelated." + "reason": "ZeRO-3 Bert loading bug is unrelated to resume-training with trust-remote-code." }, { "left": "issue:43596", - "right": "issue:46165", + "right": "issue:44805", "accept": false, - "reason": "Different bug families: ZeRO-3 model loading vs training resume/trust-remote-code." + "reason": "Bert ZeRO-3 index error and mask/tensor shape mismatch are different problems." }, { "left": "issue:44805", "right": "issue:44918", "accept": false, - "reason": "Both mention shape/loading trouble, but the failures and call paths differ." + "reason": "The tensor-mask mismatch is not the same as the Qwen3.5 input-embeddings/TRL issue." }, { "left": "issue:43572", "right": "issue:46165", "accept": false, - "reason": "Config regression vs training-resume issue; no duplicate evidence." + "reason": "StableLm config pad-token regression is unrelated to training resume behavior." }, { "left": "issue:34634", "right": "issue:35545", "accept": false, - "reason": "Processor voice_preset bug and ONNX export bug are unrelated." + "reason": "BarkProcessor voice_preset and ModernBERT ONNX export are unrelated." }, { "left": "issue:30277", "right": "issue:43638", "accept": false, - "reason": "Both involve DeepSpeed, but one is collective mismatch and the other is empty-index loading; not the same bug." + "reason": "Collective-mismatch DeepSpeed bug is different from the ZeRO-3 empty-index loading bug." }, { "left": "issue:43596", "right": "issue:43638", "accept": true, - "reason": "Same ZeRO-3 loading bug: both report index 0 out of bounds from an empty tensor/list during Bert model initialization." + "reason": "Same ZeRO-3 index-0-out-of-bounds loading failure on Bert; one is just a non-pretrained reproduction." }, { "left": "issue:43975", "right": "issue:44457", "accept": false, - "reason": "Detokenization regression and LoRA merge/save inconsistency are different problems." + "reason": "Detokenization regression and LoRA save/load mismatch are unrelated." }, { "left": "issue:43039", "right": "issue:45161", "accept": false, - "reason": "Liger Kernel cross-entropy path vs GPT-OSS tensor-parallel issue are unrelated." + "reason": "Liger cross-entropy being called and GPT-OSS tensor-parallel failure are not the same bug." }, { "left": "issue:44292", "right": "issue:44457", "accept": false, - "reason": "Qwen NVFP4 runtime error and LoRA merge/save mismatch are unrelated." + "reason": "Qwen-3 NVFP4 runtime error and LoRA merge/save mismatch are unrelated." }, { "left": "issue:34928", "right": "issue:44805", "accept": false, - "reason": "Activation-checkpointing/FSDP size mismatch and mask shape mismatch are different failures." + "reason": "Activation-checkpointing/FSDP size recomputation bug is different from the mask-shape error." }, { "left": "issue:24643", "right": "issue:30277", "accept": false, - "reason": "Both are DeepSpeed-related, but the concrete failures differ: 2-D weight error vs collective mismatch." + "reason": "DeepSpeed weight-dimension error and collective mismatch are different failures." }, { "left": "issue:45859", "right": "issue:45902", - "accept": false, - "reason": "Same model family, but different failure modes: CPU offload incompatibility vs uninitialized-memory read after materialization." + "accept": true, + "reason": "Same Qwen3_5Moe rotary-embedding state/device-transition bug; one reports CPU offload, the other meta-to-empty materialization." }, { "left": "issue:43638", "right": "issue:45161", "accept": false, - "reason": "ZeRO-3 Bert loading bug and GPT-OSS TP issue are unrelated." + "reason": "ZeRO-3 Bert loading bug is unrelated to GPT-OSS tensor-parallel behavior." }, { "left": "issue:34928", "right": "issue:36331", "accept": false, - "reason": "FSDP checkpointing mismatch and CustomTrainer compute_loss signature break are unrelated." + "reason": "FSDP activation-checkpointing bug and CustomTrainer signature breakage are different issues." }, { "left": "issue:43596", "right": "issue:45161", "accept": false, - "reason": "Different code paths: ZeRO-3 loading bug vs GPT-OSS tensor-parallel failure." + "reason": "Bert ZeRO-3 loading failure is unrelated to GPT-OSS tensor-parallel failure." }, { "left": "issue:43716", "right": "issue:45507", "accept": false, - "reason": "Mistral-3 dtype mismatch and GraniteMoEHybrid invalid method are unrelated." + "reason": "Image dtype mismatch and GraniteMoE invalid-method call are unrelated." }, { "left": "issue:45823", "right": "issue:45874", "accept": false, - "reason": "Gemma4 device mismatch and Gemma4 reverse-embedding performance regression are not the same bug." + "reason": "Gemma4 device-map mismatch and expensive reverse-embedding lookup are different forward-path bugs." }, { "left": "issue:39290", "right": "issue:45081", "accept": false, - "reason": "Gemma3 vLLM config error and Mistral tokenizer regex crash are unrelated." + "reason": "Gemma3 config/vLLM attribute error is unrelated to the Mistral regex patch crash." }, { "left": "issue:43425", "right": "issue:44292", "accept": false, - "reason": "Torch version incompatibility and Qwen NVFP4 runtime error are not duplicates." + "reason": "Torch version incompatibility and Qwen-3 NVFP4 runtime failure are not the same bug." }, { "left": "issue:34928", "right": "issue:41720", "accept": false, - "reason": "Different failures: FSDP activation checkpointing vs Qwen3 auto device-map CUDA assert." + "reason": "FSDP checkpointing issue and Qwen3 auto-device-map CUDA assert are unrelated." }, { "left": "issue:43828", "right": "issue:45161", "accept": false, - "reason": "Autocast dtype mismatch and GPT-OSS TP failure are unrelated." + "reason": "Autocast dtype mismatch and GPT-OSS tensor-parallel failure are different problems." }, { "left": "issue:44918", "right": "issue:45161", "accept": false, - "reason": "Qwen3.5 embedding/trainer bug and GPT-OSS TP bug are different issues." + "reason": "Qwen3.5/TRl input-embedding issue is unrelated to GPT-OSS tensor-parallel behavior." }, { "left": "issue:43828", "right": "issue:44292", "accept": false, - "reason": "Different model-specific runtime errors with no shared code-path." + "reason": "Autocast dtype mismatch and Qwen-3 NVFP4 runtime error are unrelated." }, { "left": "issue:43716", "right": "issue:45685", "accept": false, - "reason": "Mistral-3 dtype mismatch and MPS histogram Int error are unrelated." + "reason": "Mistral image-preprocessor dtype mismatch and MPS histogram issue are unrelated." }, { "left": "issue:43541", "right": "issue:43716", "accept": false, - "reason": "Torch dynamo grouped_mm tracing failure and Mistral-3 dtype mismatch are different issues." + "reason": "Mixtral torch-dynamo grouped_mm error and image dtype mismatch are different bugs." }, { "left": "issue:44292", "right": "issue:44918", "accept": false, - "reason": "Different Qwen-related bugs: NVFP4 runtime failure vs SFT embedding unpacking." + "reason": "Qwen-3 NVFP4 runtime error and Qwen3.5/TRl input-embedding failure are unrelated." }, { "left": "issue:43716", "right": "issue:44292", "accept": false, - "reason": "Different models and failure modes." + "reason": "Image preprocessor dtype mismatch is not the same as the Qwen-3 NVFP4 issue." }, { "left": "issue:45507", "right": "issue:45685", "accept": false, - "reason": "Invalid method call in GraniteMoEHybrid is unrelated to the MPS histogram backend limitation." + "reason": "GraniteMoE invalid-method call and MPS histogram int error are unrelated." }, { "left": "issue:43039", "right": "issue:44918", "accept": false, - "reason": "Liger Kernel cross-entropy call and Qwen3.5 input-embedding failure are unrelated." + "reason": "Liger cross-entropy routing and Qwen3.5/TRl input embedding failure are different bugs." }, { "left": "issue:41762", "right": "issue:43638", "accept": true, - "reason": "Same underlying ZeRO-3 empty-index loading bug: both hit index 0 out of bounds while materializing a model with DeepSpeed ZeRO-3." + "reason": "Same ZeRO-3 empty-index loading error; Gemma3 and Bert are just different reproductions of the same root cause." }, { "left": "issue:43827", "right": "issue:44512", "accept": false, - "reason": "Both are docs regressions after v5, but they cover different removed features and are not the same issue." + "reason": "Both are docs regressions in v5, but they refer to different removed features and not one shared fix." }, { "left": "issue:43716", "right": "issue:44560", "accept": false, - "reason": "Mistral-3 dtype mismatch and Qwen3-vl video StopIteration are unrelated." + "reason": "Mistral dtype mismatch and Qwen3-vl video StopIteration are unrelated." }, { "left": "issue:43828", "right": "issue:45685", "accept": false, - "reason": "Autocast dtype mismatch and MPS Int histogram limitation are different bugs." + "reason": "Autocast dtype mismatch and MPS histogram int failure are different backend issues." }, { "left": "issue:44918", "right": "issue:45685", "accept": false, - "reason": "Qwen3.5 trainer failure and MPS backend limitation are unrelated." + "reason": "Qwen3.5/TRl input-embedding failure is unrelated to the MPS histogram issue." }, { "left": "issue:43039", "right": "issue:43828", "accept": false, - "reason": "Liger Kernel cross-entropy path and autocast dtype mismatch are not the same bug." + "reason": "Liger cross-entropy routing and autocast dtype mismatch are not the same bug." }, { "left": "issue:30277", "right": "issue:36331", "accept": false, - "reason": "DeepSpeed collective mismatch and CustomTrainer signature change are unrelated." + "reason": "DeepSpeed collective mismatch and CustomTrainer signature breakage are unrelated." }, { "left": "issue:45237", "right": "issue:46015", "accept": false, - "reason": "GPU runtime failure and outdated docs bug are unrelated." + "reason": "AMD GPU runtime failure and outdated RAG docs are unrelated." }, { "left": "issue:34928", "right": "issue:36010", "accept": false, - "reason": "Training/checkpointing bug and GenerationMixin import error are unrelated." + "reason": "FSDP checkpointing bug and GenerationMixin import error are unrelated." + }, + { + "left": "issue:43039", + "right": "issue:44805", + "accept": false, + "reason": "Liger kernel cross-entropy issue and tensor mask/index mismatch are different failures." + }, + { + "left": "issue:41762", + "right": "issue:43596", + "accept": true, + "reason": "Same ZeRO-3 index-0-out-of-bounds loading failure; both are the same empty-tensor path with different model repros." + }, + { + "left": "issue:39290", + "right": "issue:41720", + "accept": false, + "reason": "Gemma3/vLLM config attribute error and Qwen3 CUDA assert are unrelated." + }, + { + "left": "issue:43901", + "right": "issue:44512", + "accept": false, + "reason": "TextClassificationPipeline docs drift and run-command docs drift are separate documentation fixes." + }, + { + "left": "issue:45237", + "right": "issue:45901", + "accept": false, + "reason": "GPT-OSS AMD GPU failure and table-QA crash are unrelated." + }, + { + "left": "issue:44841", + "right": "issue:45823", + "accept": false, + "reason": "Voxtral processor failure and Gemma4 device-map mismatch are different issues." }, { "left": "issue:43039", "right": "issue:44560", "accept": false, - "reason": "Liger Kernel cross-entropy issue and Qwen3-vl video StopIteration are unrelated." + "reason": "Liger kernel issue and Qwen3-vl video StopIteration are unrelated." + }, + { + "left": "issue:43975", + "right": "issue:46015", + "accept": false, + "reason": "Detokenization regression and outdated RAG docs are unrelated." }, { "left": "issue:45812", "right": "issue:45920", "accept": true, - "reason": "Same AutoTokenizer v5 regression: both report wrong token IDs across affected model families." + "reason": "Same tokenizer-ID regression; 45920 is the broader report spanning multiple model families." }, { "left": "issue:44863", "right": "issue:46015", "accept": false, - "reason": "NemotronH checkpoint-loading failure and outdated RAG docs are unrelated." + "reason": "NemotronH checkpoint loading and RAG docs are unrelated." }, { "left": "issue:43866", "right": "issue:44863", "accept": false, - "reason": "Both are model loading problems, but with different models and distinct root causes." + "reason": "Checkpoint corruption and NemotronH load failure are different bugs." }, { "left": "issue:44292", "right": "issue:46015", "accept": false, - "reason": "Qwen NVFP4 runtime error and outdated docs issue are unrelated." + "reason": "Qwen-3 NVFP4 runtime error and outdated RAG docs are unrelated." } ] }, "evaluator_result": { "accept": true, - "feedback": "The summary is grounded in the packet and stays conservative overall. The accepted duplicate calls are well-supported by the titles/symptoms (notably the ZeRO-3 empty-index loader pair and the AutoTokenizer v5 wrong-token-id regression). The negative soft-edge verdicts generally avoid overclaiming duplicate evidence." + "feedback": "The analysis is grounded in the packet and stays conservative on the soft edges. The accepted duplicate groups are plausible from the titles and descriptions (ZeRO-3 empty-index loading, AutoTokenizer token-ID regression, and the Qwen3_5Moe rotary-embedding/materialization bug), and the rejections avoid overstating weak similarity. No clear mergeability or duplicate claim looks unsupported." }, "evaluator_used": true, "retried": false, @@ -556,13 +586,13 @@ "component_index": 1, "component_count": 17, "review_unit_index": 2, - "review_unit_count": 49, + "review_unit_count": 50, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-427", + "cluster_id": "cluster-10000-428", "nodes": [ "issue:36010", "issue:39186", @@ -625,8 +655,8 @@ "issue:45362|issue:45464", "issue:44292|issue:45901", "issue:39290|issue:45823", - "issue:43404|issue:43866", "issue:45464|issue:45685", + "issue:43404|issue:43866", "issue:43541|issue:45685", "issue:43541|issue:43828", "issue:44841|issue:45084", @@ -643,8 +673,8 @@ "issue:45020|issue:45800", "issue:43531|issue:45070", "issue:43541|issue:44589", - "issue:43054|issue:43994", "issue:41093|issue:43596", + "issue:43054|issue:43994", "issue:45759|issue:45902", "issue:45542|issue:45800", "issue:39186|issue:44918", @@ -654,14 +684,14 @@ "issue:43645|issue:45399", "issue:43404|issue:45005" ], - "prepared_review_unit_hash": "36e115fd6a7eb93864d734fc3b1ed767e02d8e21f2884e563e1275ac84de7b86", + "prepared_review_unit_hash": "f6acee94451fa58c36b13d3ddd3e24daa5390fd4aa877ab3195593ffbb69a12f", "component_budget": { - "node_count": 427, - "item_count": 427, - "soft_pair_count": 1576, - "serialized_chars": 816182, - "estimated_input_tokens": 204046, - "estimated_eval_tokens": 408348 + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 }, "budget": { "node_count": 48, @@ -676,253 +706,256 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The cluster is highly heterogeneous: most items are distinct model-loading, dtype, quantization, and config bugs. All soft pairs look like topical similarity rather than true duplicates, so none should be merged.", - "confidence": 0.97, - "canonical_issue_reason": null, + "summary": "This cluster is mostly heterogeneous: many issues share broad themes like loading errors, dtype mismatches, zero-3, or model-specific regressions, but they do not look like true duplicates. The only strong duplicate-like pair is the SigLIP2 output/embedding regression.", + "confidence": 0.68, + "canonical_issue_reason": "issue:43994 is the clearest concrete report in the only likely duplicate pair: it directly describes the bad SigLIP2 behavior with AutoModel/pipeline and is more actionable than the comparison-style report in 43054.", "canonical_pr_reason": null, - "best_issue_reason": null, + "best_issue_reason": "issue:43994 is the best representative of the cluster because it is the most specific, reproducible, and clearly matches the one accepted duplicate edge.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:39290", "right": "issue:43531", "accept": false, - "reason": "Both mention model config behavior, but one is a Gemma3/vLLM attribute error and the other is a Qwen3-MoE sliding_window bug; different models and failure modes." + "reason": "Both mention model config/windowing regressions, but they are different models and different failure modes; too broad to treat as the same bug." }, { "left": "issue:43854", "right": "issue:43866", "accept": false, - "reason": "PaddleOCR-VL load failure and Ovis2 checkpoint corruption are unrelated problems; no shared code path is evident." + "reason": "One is a unit-test model load failure, the other is a corrupted checkpoint report; not the same underlying issue." }, { "left": "issue:43792", "right": "issue:45901", "accept": false, - "reason": "Whisper inference failure and table-question-answering crashes are different tasks and likely different stacks." + "reason": "Whisper loading failure and table-question-answering crash are unrelated code paths and symptoms." }, { "left": "issue:41093", "right": "issue:43638", "accept": false, - "reason": "Both are DeepSpeed zero3/index errors, but one is a mask-shape mismatch and the other is an out-of-bounds index during Bert model init; not clearly the same bug." + "reason": "Both involve zero-3/index errors, but the triggers and stack-level symptoms differ enough that this does not look like one concrete bug." }, { "left": "issue:43425", "right": "issue:45070", "accept": false, - "reason": "Torch version incompatibility and a PretrainedConfig/pydantic regression are unrelated compatibility issues." + "reason": "Torch version incompatibility and a pydantic PretrainedConfig-field regression are unrelated." }, { "left": "issue:43596", "right": "issue:45800", "accept": false, - "reason": "Both involve zero3/runtime loading, but one is an index error on BertModel and the other is a specific torch/transformers version incompatibility." + "reason": "These are different issues in different areas: zero-3/Bert loading vs a torch-transformers version incompatibility." }, { "left": "issue:43975", "right": "issue:45901", "accept": false, - "reason": "Incorrect detokenization for DeepSeek Coder is unrelated to table-question-answering crashes." + "reason": "Incorrect detokenization and a table QA crash are not the same bug." }, { "left": "issue:43792", "right": "issue:45070", "accept": false, - "reason": "Whisper runtime failure and a pydantic/config regression are different issues with no shared failure path." + "reason": "Whisper cannot run and pydantic config handling are unrelated." }, { "left": "issue:45362", "right": "issue:45464", "accept": false, - "reason": "Both involve Qwen3.5, but one is a generic chat crash and the other is a streaming inference API failure; too different to collapse." + "reason": "Both mention Qwen3.5, but one is a generic chat crash and the other is a streaming inference API failure; not enough to merge as one bug." }, { "left": "issue:44292", "right": "issue:45901", "accept": false, - "reason": "Qwen-NVFP4 runtime error and table QA crashes are separate model/task failures." + "reason": "Qwen-3-8B-NVFP4 loading failure and table QA crash are unrelated." }, { "left": "issue:39290", "right": "issue:45823", "accept": false, - "reason": "Gemma3/vLLM sliding_window attribute error and Gemma4 device mismatch are different Gemma-family bugs with different causes." + "reason": "Both are model-specific regressions, but Gemma3 sliding-window behavior and Gemma4 device mismatch are different problems." }, { - "left": "issue:43404", - "right": "issue:43866", + "left": "issue:45464", + "right": "issue:45685", "accept": false, - "reason": "Unbound tied weights in Mistral3 and a corrupted Ovis2 checkpoint are unrelated." + "reason": "Streaming chat failure and an MPS histogram/int issue are completely different code paths." }, { - "left": "issue:45464", - "right": "issue:45685", + "left": "issue:43404", + "right": "issue:43866", "accept": false, - "reason": "Qwen3.5 streaming inference and an MPS histogram/int error in moe are different code paths and platforms." + "reason": "Tied lm_head weights and a corrupted checkpoint are unrelated." }, { "left": "issue:43541", "right": "issue:45685", "accept": false, - "reason": "Torch dynamo grouped_mm tracing failure and MPS histogram kernel failure are unrelated execution environments and bugs." + "reason": "Mixtral tracing/grouped_mm and MPS histogram failures are unrelated." }, { "left": "issue:43541", "right": "issue:43828", "accept": false, - "reason": "Both mention dtype-related failures, but one is torch dynamo tracing on Mixtral and the other is autocast on Phi-tiny-MoE; not the same concrete defect." + "reason": "Both are numeric/dtype-adjacent, but the specific model paths and failure mechanisms differ." }, { "left": "issue:44841", "right": "issue:45084", "accept": false, - "reason": "Voxtral processor failure and a compiler/template-node TypeError are unrelated." + "reason": "Voxtral processor failure and a non-template-node compile error are unrelated." }, { "left": "issue:36010", "right": "issue:39290", "accept": false, - "reason": "ImportError for GenerationMixin and Gemma3 sliding_window attribute error are different API breakages." + "reason": "Importing GenerationMixin and Gemma3 sliding_window are unrelated issues." }, { "left": "issue:45800", "right": "issue:45823", "accept": false, - "reason": "Torch/transformers version incompatibility and Gemma4 device_map forward mismatch are separate issues." + "reason": "A torch/transformers compatibility complaint and a Gemma4 device-map bug are different failures." }, { "left": "issue:41093", "right": "issue:41762", "accept": false, - "reason": "Both are index errors, but one is a mask-length mismatch and the other is zero-sized tensor access when loading Gemma3 with ZeRO-3; different triggers and models." + "reason": "Both are zero-3-related index errors, but one is a mask-shape mismatch and the other is a model-loading out-of-bounds error; too specific to unify." }, { "left": "issue:43296", "right": "issue:43366", "accept": false, - "reason": "PaddleOCR-VL load failure and GGUF gpt-oss support are distinct model-support requests/failures." + "reason": "PaddleOCR-VL loading in vLLM and GGUF gpt-oss support are unrelated." }, { "left": "issue:43572", "right": "issue:45070", "accept": false, - "reason": "Missing pad_token_idx in StableLmConfig and a pydantic PretrainedConfig regression are not the same bug." + "reason": "StableLm pad_token_idx regression and pydantic PretrainedConfig changes are not the same bug." }, { "left": "issue:43541", "right": "issue:45161", "accept": false, - "reason": "Torch dynamo tracing on Mixtral and GPT-OSS MoE tensor-parallel failure are different MoE-related problems, not the same code path." + "reason": "Mixtral grouped_mm tracing and GPT-OSS MoE tensor-parallel support are different model/runtime problems." }, { "left": "issue:44387", "right": "issue:45005", "accept": false, - "reason": "Int4 OOM behavior and tied-weight translation-model issues are unrelated." + "reason": "CUDA memory growth under int4 quantization and tied weights on translation models are unrelated." }, { "left": "issue:43828", "right": "issue:45464", "accept": false, - "reason": "Autocast dtype mismatch on Phi-tiny-MoE and Qwen3.5 streaming API failure are different models and symptoms." + "reason": "An autocast dtype mismatch and a Qwen3.5 streaming API crash are unrelated." }, { "left": "issue:44291", "right": "issue:45070", "accept": false, - "reason": "init_empty_weights/_is_hf_initialized TypeError and the v5.4 pydantic field break are distinct initialization regressions." + "reason": "init_empty_weights argument handling and a pydantic config-field regression do not match." }, { "left": "issue:43366", "right": "issue:45464", "accept": false, - "reason": "GGUF GPT-OSS support and Qwen3.5 streaming inference do not share the same underlying defect." + "reason": "GPT-OSS GGUF support and Qwen3.5 streaming chat failures are unrelated." }, { "left": "issue:45020", "right": "issue:45800", "accept": false, - "reason": "Remote_code breakage and torch/transformers compatibility are both broad version issues, but they are not the same bug." + "reason": "Remote-code model breakage and a torch/transformers version incompatibility are too broad and not the same concrete bug." }, { "left": "issue:43531", "right": "issue:45070", "accept": false, - "reason": "Qwen3-MoE sliding_window handling and PretrainedConfig pydantic-field breakage are unrelated." + "reason": "Qwen3-MoE sliding_window behavior and pydantic config handling are unrelated." }, { "left": "issue:43541", "right": "issue:44589", "accept": false, - "reason": "Grouped_mm tracing failure and Float8 storage lookup failure are different runtime errors." - }, - { - "left": "issue:43054", - "right": "issue:43994", - "accept": false, - "reason": "Siglip2 embedding quality regression and siglip2 nonsensical pipeline results are related by model family, but not clearly the same failure or fix." + "reason": "Mixtral tracing errors and Float8 storage lookup failures are different runtime issues." }, { "left": "issue:41093", "right": "issue:43596", "accept": false, - "reason": "Both are zero3 index errors, but one is a mask shape mismatch and the other is size-0 access on BertModel; too different to merge." + "reason": "Both mention zero-3 IndexErrors, but the exact failure surfaces and models differ enough to reject as duplicates." + }, + { + "left": "issue:43054", + "right": "issue:43994", + "accept": true, + "reason": "Same SigLIP2 regression: both reports describe bad/nonsensical SigLIP2 text embeddings or outputs compared with expected behavior." }, { "left": "issue:45759", "right": "issue:45902", "accept": false, - "reason": "Composite Qwen config unwrapping and Qwen3.5Moe rotary embedding uninitialized memory are separate model-initialization bugs." + "reason": "Composite Qwen config unwrapping and uninitialized rotary embedding memory are different bugs." }, { "left": "issue:45542", "right": "issue:45800", "accept": false, - "reason": "TensorBoard-without-TensorFlow backend error and a torch/transformers incompatibility are unrelated environment issues." + "reason": "TensorBoard/TensorFlow backend handling and torch/transformers version incompatibility are unrelated." }, { "left": "issue:39186", "right": "issue:44918", "accept": false, - "reason": "FSDP weight dimensionality and Qwen3.5 embedding unpacking are different training/runtime problems." + "reason": "FSDP 2-D weight error and Qwen3.5 SFT embedding unpacking are different problems." }, { "left": "issue:40990", "right": "issue:44960", "accept": false, - "reason": "GPT-OSS perplexity regression and GLM5 are different model behavior reports with no clear shared defect." + "reason": "High perplexity on GPT-OSS and a GLM5 issue are unrelated." }, { "left": "issue:43716", "right": "issue:44918", "accept": false, - "reason": "Gemma-3 image preprocessor dtype mismatch and Qwen3.5 embedding unpacking are different multimodal/training bugs." + "reason": "Mistral-3 image-preprocessor dtype mismatch and Qwen3.5 SFT embedding unpacking are unrelated." }, { "left": "issue:45632", "right": "issue:45698", "accept": false, - "reason": "Cache-path collision for remote code and wrong custom module loaded after save/load are both custom-code issues, but they are different mechanisms." + "reason": "Both involve custom model loading, but one is a cache-path collision and the other is wrong module selection after save/load; not clearly the same concrete bug." }, { "left": "issue:43645", "right": "issue:45399", "accept": false, - "reason": "Notebook custom-model initialization regression and flash-attn2 fallback blocking are unrelated regression types." + "reason": "Notebook custom-model initialization and flash-attn2 fallback gating are unrelated." }, { "left": "issue:43404", "right": "issue:45005", "accept": false, - "reason": "Mistral3 tied-weight bug and translation tied-weight issue are both about weights, but they affect different models and code paths." + "reason": "Both discuss tied weights, but they affect different model families and likely different code paths; too broad to merge." } ] }, - "evaluator_result": null, - "evaluator_used": false, + "evaluator_result": { + "accept": true, + "feedback": "The writeup is grounded in the packet and stays conservative on the rejects. The one accepted soft pair, issue:43054 and issue:43994, is reasonably supported by the titles alone as the same SigLIP2 output-quality regression. No mergeability problems apply here." + }, + "evaluator_used": true, "retried": false, - "accepted_nontrivial_soft_edge": false, + "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null }, @@ -931,13 +964,13 @@ "component_index": 1, "component_count": 17, "review_unit_index": 3, - "review_unit_count": 49, + "review_unit_count": 50, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-427", + "cluster_id": "cluster-10000-428", "nodes": [ "issue:36331", "issue:38175", @@ -999,8 +1032,8 @@ "issue:43975|issue:44292", "issue:44292|issue:45507", "issue:43950|issue:44534", - "issue:39692|issue:43054", "issue:43644|issue:43950", + "issue:39692|issue:43054", "issue:43828|issue:44928", "issue:36331|issue:41762", "issue:41628|issue:45800", @@ -1018,256 +1051,250 @@ "issue:44403|issue:44451", "issue:43975|issue:45237", "issue:43643|issue:45020", - "issue:42947|issue:43856", "issue:44403|issue:44488", + "issue:42947|issue:43856", "issue:41762|issue:42915", "issue:43881|issue:43883", "issue:43782|issue:44960" ], - "prepared_review_unit_hash": "a6ecc7d084e25fdc6b53780b75e6877581edf65f8a7ea7687ed0d49f653f50a4", + "prepared_review_unit_hash": "28ff74252fd5eb7c48632c059f5dc1887acb1b7e8491ed2c4e1aba2bd7c987fc", "component_budget": { - "node_count": 427, - "item_count": 427, - "soft_pair_count": 1576, - "serialized_chars": 816182, - "estimated_input_tokens": 204046, - "estimated_eval_tokens": 408348 + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 34, - "serialized_chars": 37508, - "estimated_input_tokens": 9377, - "estimated_eval_tokens": 19010 + "serialized_chars": 37510, + "estimated_input_tokens": 9378, + "estimated_eval_tokens": 19012 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This set is mostly heterogeneous. Only two soft pairs look like true duplicates: the Transformers non-persistent-buffer regression and the SigLIP2 implementation discrepancy; the rest are same-subsystem or same-model but not the same concrete bug.", - "confidence": 0.54, - "canonical_issue_reason": "No single issue cleanly represents the whole mixed set, but issue:43950 is the strongest canonical candidate for the only clear duplicate cluster because it states the broad non-persistent-buffer regression and has the closest matching duplicate candidate.", + "summary": "This cluster is mostly heterogeneous, with one clear duplicate subgroup around the Transformers 5.x non-persistent-buffer corruption regression. The rest of the pairs are only broadly similar by subsystem or model family, not the same concrete bug.", + "confidence": 0.9, + "canonical_issue_reason": "issue:43644 is the earliest and broadest statement of the shared regression (\u201cfills non-persistent buffers with junk\u201d), making it the best anchor for the duplicate set; later reports like issue:43950 and issue:44534 look like narrower manifestations of the same loading bug.", "canonical_pr_reason": null, - "best_issue_reason": "Issue:43950 is the best representative of a concrete, reproducible regression with a strong duplicate nearby; it is more canonical than the other items that are either broader, more symptom-level, or clearly unrelated.", + "best_issue_reason": "issue:43950 is the most actionable standalone report because it names the failure mode precisely (`from_pretrained()` corrupting `register_buffer(persistent=False)` state) and aligns cleanly with the duplicate reports.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:42915", "right": "issue:45399", "accept": false, - "reason": "Both involve model/runtime problems, but one is a Qwen3MoE FP8 failure and the other is a flash-attn2 fallback gate; not the same bug." + "reason": "Different problems: Qwen3Moe FP8 failure vs a flash-attn2 fallback gating issue." }, { "left": "issue:44530", "right": "issue:45005", "accept": false, - "reason": "PagedAttentionCache crash on Qwen3.5 is unrelated to tied-weights handling on translation models." + "reason": "Different bugs: Qwen3.5 PagedAttention cache crash vs tied-weights handling in translation models." }, { "left": "issue:43873", "right": "issue:45005", "accept": false, - "reason": "Quantization/offloading behavior is a different code path from tied-weights issues." + "reason": "Quantization/offloading behavior is unrelated to tied-weights regressions." }, { "left": "issue:44863", "right": "issue:45005", "accept": false, - "reason": "NemotronH checkpoint loading failure is not the same as the v5 tied-weights regression." + "reason": "NemotronH checkpoint loading is a separate model-compatibility issue, not tied-weights handling." }, { "left": "issue:44360", "right": "issue:44512", "accept": false, - "reason": "DSA indexer ReLU discussion is unrelated to a docs typo about the removed run command." + "reason": "DSA indexer architecture discussion is unrelated to docs mentioning a removed CLI command." }, { "left": "issue:44918", "right": "issue:45507", "accept": false, - "reason": "TRL unpacking failure for Qwen3.5 embeddings is unrelated to GraniteMoEHybrid calling an invalid method." + "reason": "TRl SFT embedding unpacking and GraniteMoE invalid-method calls are different model/path failures." }, { "left": "issue:44492", "right": "issue:44512", "accept": false, - "reason": "Cache-strategies typo and removed run-command docs are different documentation issues." + "reason": "A typo in cache strategy docs is not the same as stale docs about the removed run command." }, { "left": "issue:43975", "right": "issue:44292", "accept": false, - "reason": "DeepSeek detokenization regression and Qwen-3 NVFP4 runtime error are different model bugs." + "reason": "Different model/output issues: detokenization regression vs NVFP4 runtime failure." }, { "left": "issue:44292", "right": "issue:45507", "accept": false, - "reason": "These are unrelated model failures with no shared concrete code-path." + "reason": "Qwen NVFP4 runtime failure is unrelated to GraniteMoE calling an invalid method." }, { "left": "issue:43950", "right": "issue:44534", "accept": true, - "reason": "Both describe the same non-persistent-buffer regression in v5: buffers that should be preserved are coming back as junk/corrupted." - }, - { - "left": "issue:39692", - "right": "issue:43054", - "accept": false, - "reason": "SigLIP2 docs example errors are not the same as degraded text embeddings." + "reason": "Both report the same Transformers 5.x regression where non-persistent buffers are materialized with junk values." }, { "left": "issue:43644", "right": "issue:43950", + "accept": true, + "reason": "Same underlying bug: Transformers 5.x corrupts non-persistent buffers during loading/materialization." + }, + { + "left": "issue:39692", + "right": "issue:43054", "accept": false, - "reason": "Missing fields in AutoConfig with trust_remote_code is a different bug from non-persistent buffer corruption." + "reason": "Both involve SigLIP2, but one is a docs/example mismatch and quantization failure while the other is degraded text embeddings." }, { "left": "issue:43828", "right": "issue:44928", "accept": false, - "reason": "Autocast dtype mismatch and RLHF NaN from 3D position_ids are distinct failures." + "reason": "Both mention Qwen/positioning problems, but the concrete failures differ: dtype mismatch under autocast vs RLHF gradient explosion." }, { "left": "issue:36331", "right": "issue:41762", "accept": false, - "reason": "Trainer keyword-argument incompatibility is unrelated to the Gemma3 ZeRO-3 index error." + "reason": "CustomTrainer signature mismatch and Gemma3 ZeRO-3 loading error are unrelated runtime failures." }, { "left": "issue:41628", "right": "issue:45800", "accept": false, - "reason": "Import failure for AutoImageProcessor is not the same as a torch/transformers version incompatibility report." + "reason": "AutoImageProcessor import/export breakage is not the same as a torch-version compatibility issue." }, { "left": "issue:43950", "right": "issue:45902", "accept": false, - "reason": "Both mention buffer initialization, but 45902 is a specific Qwen rotary-embedding materialization issue, not the same from_pretrained regression." + "reason": "Related buffer/materialization theme, but 45902 is a specific Qwen3.5 rotary-embedding uninitialized-memory bug, not the same generic from_pretrained corruption report." }, { "left": "issue:43872", "right": "issue:44589", "accept": false, - "reason": "bitsandbytes kwarg incompatibility and Float8 storage lookup failure are different quantization/storage bugs." + "reason": "bitsandbytes parameter-construction incompatibility is distinct from missing Float8 storage support." }, { "left": "issue:43866", "right": "issue:44403", "accept": false, - "reason": "Corrupted checkpoint and loading noise are not the same underlying issue." + "reason": "Checkpoint corruption and noisy loading are different failures." }, { "left": "issue:43054", "right": "issue:45701", "accept": false, - "reason": "SigLIP2 embedding quality is unrelated to tokenization changes in transformers versions." + "reason": "Both mention tokenization/model output quality, but one is SigLIP2 text embedding degradation and the other is a general tokenization change regression." }, { "left": "issue:41553", "right": "issue:43883", "accept": false, - "reason": "Bad AutoTokenizer error for Voxtral is unrelated to Molmo missing all_tied_weights_keys." + "reason": "Bad AutoTokenizer error handling for Voxtral is unrelated to Molmo missing `all_tied_weights_keys`." }, { "left": "issue:38175", "right": "issue:43054", "accept": false, - "reason": "Unexpected zero probabilities and worse SigLIP2 embeddings are similar in theme but not the same concrete bug." + "reason": "Zero probabilities in siglip2-base and worse SigLIP2 text embeddings are different symptoms with different likely causes." }, { "left": "issue:44910", "right": "issue:45859", "accept": false, - "reason": "Flash-Attention illegal-memory access from 3D position_ids is different from CPU-offload incompatibility in Qwen3_5MoeTextRotaryEmbedding." + "reason": "Qwen3.5 flash-attention crash from 3D position_ids is not the same as CPU-offload incompatibility in rotary embedding." }, { "left": "issue:45663", "right": "issue:45902", "accept": false, - "reason": "Gemma4 FSDP2 shared_kv_states bug is unrelated to Qwen rotary inv_freq uninitialized memory." + "reason": "Gemma4 FSDP2 shared_kv_states bug is unrelated to Qwen3.5 rotary-embedding uninitialized memory." }, { "left": "issue:41553", "right": "issue:43881", "accept": false, - "reason": "Voxtral tokenizer loading error and glm-4v loading failure are different model-specific problems." + "reason": "Bad AutoTokenizer error messaging for Voxtral and glm-4v-9b loading failure are separate model-loading issues." }, { "left": "issue:42947", "right": "issue:44387", "accept": false, - "reason": "LoRA gradient-checkpointing ineffectiveness is not the same as int4 memory-reserved/OOM behavior." + "reason": "LoRA gradient-checkpointing ineffectiveness and int4 CUDA memory growth are different training/runtime regressions." }, { "left": "issue:43054", "right": "issue:43493", - "accept": true, - "reason": "Both point to the same SigLIP2 implementation discrepancy versus the original JAX behavior, with degraded text embeddings as a manifestation." + "accept": false, + "reason": "Both are SigLIP2-related, but one is text embedding quality while the other is HF-vs-JAX implementation discrepancy." }, { "left": "issue:44403", "right": "issue:44451", "accept": false, - "reason": "Generic loading noise and a specific model load failure are not the same issue." + "reason": "General loading noise and inability to load a specific model are not the same bug." }, { "left": "issue:43975", "right": "issue:45237", "accept": false, - "reason": "Detokenization regression and AMD GPU support failure are unrelated." + "reason": "Detokenization regression and AMD GPU runtime incompatibility are unrelated." }, { "left": "issue:43643", "right": "issue:45020", "accept": false, - "reason": "Remote_code breakage is too broad here; 43643 is specifically about missing fields from AutoConfig, which may be only one symptom of 45020." + "reason": "Missing fields from AutoConfig with trust_remote_code is a different issue than models broadly breaking under recent transformers versions." }, { - "left": "issue:42947", - "right": "issue:43856", + "left": "issue:44403", + "right": "issue:44488", "accept": false, - "reason": "Gradient-checkpointing ineffectiveness and Qwen3 MoE memory usage are different performance problems." + "reason": "Both are loading-related, but they target different models and failure modes." }, { - "left": "issue:44403", - "right": "issue:44488", + "left": "issue:42947", + "right": "issue:43856", "accept": false, - "reason": "Two separate model-loading reports, not one underlying bug." + "reason": "LoRA checkpointing behavior and Qwen3 MoE memory usage are different optimization/training issues." }, { "left": "issue:41762", "right": "issue:42915", "accept": false, - "reason": "Gemma3 ZeRO-3 loading failure is unrelated to Qwen3MoE FP8 failure." + "reason": "Gemma3 ZeRO-3 loading failure and Qwen3Moe FP8 failure are unrelated model-loading bugs." }, { "left": "issue:43881", "right": "issue:43883", "accept": false, - "reason": "glm-4v load failure and Molmo tied-weights attribute error are unrelated." + "reason": "Different model load failures: glm-4v-9b loading vs Molmo tied-weights attribute error." }, { "left": "issue:43782", "right": "issue:44960", "accept": false, - "reason": "Qwen3VL weight_only error is unrelated to GLM5." - }, - { - "left": "issue:44360", - "right": "issue:44512", - "accept": false, - "reason": "Duplicate of the docs typo check: still unrelated to the DSA/ReLU discussion." + "reason": "Qwen3VL weight-only loading error and GLM5 are separate model-specific failures." } ] }, "evaluator_result": { "accept": true, - "feedback": "The summary is grounded in the packet and the soft-edge calls are mostly conservative. The two accepted duplicate pairs are reasonably supported by the issue titles, and the rejected pairs generally distinguish different bug classes rather than overclaiming duplicates." + "feedback": "The analysis is grounded in the packet and stays conservative overall. The canonical cluster anchor around issue:43644 / issue:43950 / issue:44534 is well-supported by the titles, timestamps, and the shared non-persistent-buffer corruption wording. The remaining soft-edge rejections are cautious and consistent with the available evidence, and the accepted pairings are the strongest clear duplicate links in the set." }, "evaluator_used": true, "retried": false, @@ -1280,13 +1307,13 @@ "component_index": 1, "component_count": 17, "review_unit_index": 4, - "review_unit_count": 49, + "review_unit_count": 50, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-427", + "cluster_id": "cluster-10000-428", "nodes": [ "issue:24643", "issue:30064", @@ -1339,8 +1366,8 @@ ], "soft_pairs": [ "issue:45685|issue:45823", - "issue:44991|issue:45399", "issue:42491|issue:44368", + "issue:44991|issue:45399", "issue:43824|issue:43872", "issue:45020|issue:45823", "issue:44661|issue:45020", @@ -1352,13 +1379,13 @@ "issue:43404|issue:43950", "issue:43638|issue:45823", "issue:42915|issue:45020", - "issue:43257|issue:43866", "issue:43792|issue:44220", + "issue:43257|issue:43866", "issue:43872|issue:44291", "issue:42915|issue:43872", "issue:41553|issue:44991", - "issue:43257|issue:45305", "issue:36331|issue:41093", + "issue:43257|issue:45305", "issue:42947|issue:45923", "issue:43940|issue:45313", "issue:43854|issue:44863", @@ -1372,238 +1399,238 @@ "issue:45325|issue:45902", "issue:43366|issue:43828" ], - "prepared_review_unit_hash": "1f404cad8d1ddc9392b593d0f595ed49d7aebde4748484653d8b73764eb884d9", + "prepared_review_unit_hash": "faabc12376afaeba77f54a8cd36a21250ffeb99d184495018aa61abf6f726e8a", "component_budget": { - "node_count": 427, - "item_count": 427, - "soft_pair_count": 1576, - "serialized_chars": 816182, - "estimated_input_tokens": 204046, - "estimated_eval_tokens": 408348 + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 33, - "serialized_chars": 38067, - "estimated_input_tokens": 9517, - "estimated_eval_tokens": 19290 + "serialized_chars": 38072, + "estimated_input_tokens": 9518, + "estimated_eval_tokens": 19292 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "Mostly unrelated issue reports. I accepted only the clearly duplicate docs cleanup pair, the exact `_is_hf_initialized` loading regression pair, and the near-identical mask-shape regression pair; the rest are too model- or context-specific to merge confidently.", - "confidence": 0.88, - "canonical_issue_reason": "Issue 44291 is the clearest root-cause report for the `_is_hf_initialized` / `init_empty_weights` regression and is a better canonical anchor than the more symptom-specific duplicate report.", + "summary": "Mostly a heterogeneous cluster of unrelated bugs. I only see a few true duplicate-style matches: two docs issues about stale v5 pipeline references, two mask-shape mismatch reports, and one `_is_hf_initialized` initialization regression. The rest are similar-at-a-glance but not the same underlying problem.", + "confidence": 0.74, + "canonical_issue_reason": "There is no single canonical issue for the whole cluster; 44509 is the cleanest anchor only for the docs/pipeline-removal duplicate pair because it is the broader, more explicit report.", "canonical_pr_reason": null, - "best_issue_reason": "44291 is the strongest standalone representative because it states the concrete failure mode and underlying loading-path regression most precisely.", + "best_issue_reason": "44509 is the strongest overall issue candidate among the confirmed duplicates: it clearly states the v5 docs regression and subsumes 43827.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:45685", "right": "issue:45823", "accept": false, - "reason": "Different failure modes: MPS int histogram error vs Gemma4 device placement mismatch." + "reason": "Different models and failures: MPS/int histogram error vs Gemma4 PLE device mismatch." }, { - "left": "issue:44991", - "right": "issue:45399", + "left": "issue:42491", + "right": "issue:44368", "accept": false, - "reason": "Tokenizer loading for one model vs flash-attn fallback gating; same broad area only." + "reason": "Both involve Qwen3 LoRA/configuration, but one is a cross-version compatibility break and the other is a tie_word_embeddings warning; not the same bug." }, { - "left": "issue:42491", - "right": "issue:44368", + "left": "issue:44991", + "right": "issue:45399", "accept": false, - "reason": "Both mention Qwen/LoRA, but one is a version compatibility break and the other is only a warning about tie_word_embeddings." + "reason": "Tokenizer loading regression vs flash-attn fallback gating; different code paths and symptoms." }, { "left": "issue:43824", "right": "issue:43872", "accept": false, - "reason": "Import error for Qwen2.5-VL class vs bitsandbytes `_is_hf_initialized` constructor issue; unrelated code paths." + "reason": "Import error for a model class vs bitsandbytes `_is_hf_initialized` incompatibility; unrelated failures." }, { "left": "issue:45020", "right": "issue:45823", "accept": false, - "reason": "Remote_code regressions and a Gemma4 device-mapping bug are different underlying problems." + "reason": "Broad remote_code breakage vs a specific device-map forward mismatch; not the same underlying issue." }, { "left": "issue:44661", "right": "issue:45020", "accept": false, - "reason": "Tokenizer mapping bug vs remote_code breakage; same loader area but not the same defect." + "reason": "A tooling failure in add-new-model-like vs a runtime remote_code regression; different problems." }, { "left": "issue:43828", "right": "issue:45823", "accept": false, - "reason": "Autocast dtype mismatch in Phi-tiny-MoE is unrelated to Gemma4 device mismatch." + "reason": "Autocast dtype mismatch on Phi-tiny-MoE is separate from Gemma4 device placement." }, { "left": "issue:45081", "right": "issue:45399", "accept": false, - "reason": "Mistral regex patch crash vs flash-attn fallback checks; different code paths." + "reason": "Mistral tokenizer regex patch crash is unrelated to flash-attn fallback logic." }, { "left": "issue:45020", "right": "issue:45356", "accept": false, - "reason": "General remote_code regression vs Kimi-K2.5 tokenizer codec regression; not the same bug." + "reason": "Different tokenizer regressions with different failure modes; not one concrete bug." }, { "left": "issue:43742", "right": "issue:45399", "accept": false, - "reason": "MobileLLM load failure vs flash-attn fallback gating; no concrete shared defect." + "reason": "Generic model load failure vs flash-attn fallback checks; no shared code-path evidence." }, { "left": "issue:43827", "right": "issue:44509", "accept": true, - "reason": "Both are the same docs cleanup: stale references to removed v5 pipeline tasks." + "reason": "Same underlying docs regression: both report stale references to pipeline tasks removed in v5; 44509 is the broader formulation." }, { "left": "issue:43404", "right": "issue:43950", "accept": false, - "reason": "Weight tying bug in a model class vs silent non-persistent buffer corruption in `from_pretrained()`; different fixes." + "reason": "lm_head tying bug vs non-persistent buffer corruption; different state-handling issues." }, { "left": "issue:43638", "right": "issue:45823", "accept": false, - "reason": "DeepSpeed ZeRO-3 index error in Bert training vs Gemma4 forward device mismatch." + "reason": "DeepSpeed ZeRO-3 index error on a Bert model vs Gemma4 device mismatch; unrelated." }, { "left": "issue:42915", "right": "issue:45020", "accept": false, - "reason": "MoE weight conversion under accelerate+deepspeed is not the same as remote_code loader breakage." + "reason": "Qwen3 MoE + FineGrainedFP8Config vs broad remote_code breakage; too different." }, { - "left": "issue:43257", - "right": "issue:43866", + "left": "issue:43792", + "right": "issue:44220", "accept": false, - "reason": "Qwen3 MoE weight loading bug vs Ovis2 checkpoint corruption; unrelated." + "reason": "Whisper failing to run and a fbank helper issue may be related, but the evidence is too thin to treat them as the same bug." }, { - "left": "issue:43792", - "right": "issue:44220", + "left": "issue:43257", + "right": "issue:43866", "accept": false, - "reason": "Whisper runtime failure and `_torch_extract_fbank_features()` may be related, but the reports are too broad to treat as the same bug." + "reason": "Qwen3 MoE weight-conversion/loading bug vs corrupted Ovis2 checkpoint; different root causes." }, { "left": "issue:43872", "right": "issue:44291", "accept": true, - "reason": "Same concrete loading regression: `_is_hf_initialized` unexpectedly reaches `Int8Params` / `init_empty_weights`." + "reason": "Same `_is_hf_initialized` initialization regression showing up in bitsandbytes/Int8Params and init_empty_weights loading paths." }, { "left": "issue:42915", "right": "issue:43872", "accept": false, - "reason": "Different DeepSpeed/load-time failures: weight conversion vs bitsandbytes constructor incompatibility." + "reason": "Different Qwen3/MoE loading and FP8-related problems; no shared concrete failure." }, { "left": "issue:41553", "right": "issue:44991", "accept": false, - "reason": "Bad Voxtral tokenizer error message vs tokenizer loading failure for est-roberta; different issues." + "reason": "Bad error message for Voxtral tokenizer loading vs est-roberta tokenizer load failure; different issues." }, { - "left": "issue:43257", - "right": "issue:45305", + "left": "issue:36331", + "right": "issue:41093", "accept": false, - "reason": "Weight conversion failure vs gradient averaging bug under GAS; different code paths." + "reason": "Trainer API signature break vs tensor/mask shape mismatch; unrelated." }, { - "left": "issue:36331", - "right": "issue:41093", + "left": "issue:43257", + "right": "issue:45305", "accept": false, - "reason": "Unexpected `compute_loss` kwarg vs tensor/mask shape mismatch; unrelated training errors." + "reason": "Qwen3 MoE weight loading under accelerate+deepspeed vs gradient averaging with GAS; different training bugs." }, { "left": "issue:42947", "right": "issue:45923", "accept": false, - "reason": "Gradient checkpointing ineffective with PEFT LoRA is broader and different from a missing `supports_gradient_checkpointing` flag." + "reason": "General ineffective gradient checkpointing under PEFT LoRA vs a specific missing supports_gradient_checkpointing flag; not enough evidence they are the same bug." }, { "left": "issue:43940", "right": "issue:45313", "accept": false, - "reason": "Both are Qwen DeepSpeed ZeRO-3 load failures, but different model families and different missing-weight symptoms." + "reason": "Both are ZeRO-3 weight-loading failures on Qwen variants, but the affected model paths and symptoms differ enough that I would not merge them." }, { "left": "issue:43854", "right": "issue:44863", "accept": false, - "reason": "GLM-4.7-Flash test load failure vs NemotronH checkpoint loading failure; unrelated model-specific bugs." + "reason": "Different model-loading failures for different architectures; no concrete match." }, { "left": "issue:44928", "right": "issue:45588", "accept": false, - "reason": "RLHF gradient explosion from 3D position_ids is not the same as a flash_attention `s_aux=None` crash." + "reason": "RLHF NaN/SDPA fallback problem vs flash_attention `s_aux=None` crash; separate issues." }, { "left": "issue:43872", "right": "issue:45538", "accept": false, - "reason": "Bitsandbytes `_is_hf_initialized` incompatibility vs CLIPTokenizer max-length constant bug." + "reason": "Bitsandbytes `_is_hf_initialized` incompatibility vs CLIPTokenizer max length bug; unrelated." }, { "left": "issue:41093", "right": "issue:44805", "accept": true, - "reason": "Near-identical mask/tensor length mismatch error; same underlying off-by-one indexing bug." + "reason": "Same mask/tensor shape mismatch class with essentially the same IndexError, just reported with different concrete sizes." }, { "left": "issue:41553", "right": "issue:44488", "accept": false, - "reason": "Voxtral tokenizer error-message issue vs inability to load cjvt/sleng-bert; different model loading problems." + "reason": "Bad Voxtral error message vs failure to load a different model; no shared bug." }, { "left": "issue:44479", "right": "issue:44560", "accept": false, - "reason": "Video-input regression for several VL models vs Qwen3-VL embedding StopIteration; related area but not the same defect." + "reason": "Separate video-input regressions affecting different model families and error paths." }, { "left": "issue:44530", "right": "issue:44863", "accept": false, - "reason": "PagedAttentionCache crash on Qwen3.5 vs NemotronH checkpoint loading failure; different underlying bugs." + "reason": "PagedAttentionCache/linear_attention crash vs checkpoint-loading failure; unrelated." }, { "left": "issue:24643", "right": "issue:30064", "accept": false, - "reason": "DeepSpeed training weight-shape error vs image processor void-segmentation-map bug; unrelated." + "reason": "DeepSpeed weight shape error vs void segmentation-map processing; completely different subsystems." }, { "left": "issue:45325", "right": "issue:45902", "accept": false, - "reason": "Qwen2.5-VL rope scaling bug vs Qwen3.5 MoE `inv_freq` uninitialized-memory bug; different mechanisms." + "reason": "Qwen2.5-VL rope scaling bug vs Qwen3.5 MoE rotary embedding memory/materialization bug; different code paths." }, { "left": "issue:43366", "right": "issue:43828", "accept": false, - "reason": "GGUF/gpt-oss architecture support vs autocast dtype mismatch in Phi-tiny-MoE; unrelated." + "reason": "GGUF gpt-oss support request vs Phi-tiny-MoE autocast dtype mismatch; not the same bug." } ] }, "evaluator_result": { "accept": true, - "feedback": "The summary is grounded in the packet and stays mostly conservative. The three accepted pairs are well-supported by their titles and issue descriptions: the docs cleanup pair, the `_is_hf_initialized` / `init_empty_weights` loading regression pair, and the mask-shape mismatch pair. The rejected pairs are framed cautiously and do not overclaim equivalence. The canonical/best-issue rationale is also supported by the packet." + "feedback": "The summary is grounded in the packet and stays appropriately conservative. The three accepted duplicate-style pairs are well supported by the issue titles/bodies, and the rest are cautiously rejected without overclaiming shared root cause. No mergeability concerns apply here since these are issue pairs, not PR pairs." }, "evaluator_used": true, "retried": false, @@ -1616,13 +1643,13 @@ "component_index": 1, "component_count": 17, "review_unit_index": 5, - "review_unit_count": 49, + "review_unit_count": 50, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-427", + "cluster_id": "cluster-10000-428", "nodes": [ "issue:36010", "issue:38175", @@ -1683,8 +1710,8 @@ "issue:43645|issue:43646", "issue:42915|issue:43957", "issue:43366|issue:44589", - "issue:42617|issue:43366", "issue:44560|issue:44918", + "issue:42617|issue:43366", "issue:36010|issue:42915", "issue:44661|issue:45823", "issue:44560|issue:44805", @@ -1704,223 +1731,234 @@ "issue:45313|issue:45357", "issue:42915|issue:44756", "issue:42491|issue:45276", - "issue:43708|issue:44075", - "issue:39692|issue:43994" + "issue:43708|issue:44075" ], - "prepared_review_unit_hash": "30a20aff55deb35733896a2633689f3372ab36354ff48760aec213238f323bb1", + "prepared_review_unit_hash": "944e43036bcbde4e806bcc5895c616fc0b7dd3e2319d6e9543a46d4dfc552d99", "component_budget": { - "node_count": 427, - "item_count": 427, - "soft_pair_count": 1576, - "serialized_chars": 816182, - "estimated_input_tokens": 204046, - "estimated_eval_tokens": 408348 + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 }, "budget": { "node_count": 47, "item_count": 47, - "soft_pair_count": 33, - "serialized_chars": 36926, - "estimated_input_tokens": 9232, - "estimated_eval_tokens": 18720 + "soft_pair_count": 32, + "serialized_chars": 36619, + "estimated_input_tokens": 9155, + "estimated_eval_tokens": 18566 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "Mostly a loose collection of unrelated Transformers issues, with a few clear duplicate pairs around the same regression topics (non-persistent buffers in v5, SigLIP2 bad outputs, custom model init, and BERT model-load failures).", - "confidence": 0.86, - "canonical_issue_reason": "issue:43644 is the clearest canonical pick: it states the non-persistent buffer corruption regression directly and has an exact near-duplicate in 44534.", + "summary": "This cluster is mostly a mix of unrelated transformer bugs; only a few soft pairs look like true duplicates of the same concrete regression. The clearest duplicate subcluster is the Transformers v5 non-persistent-buffer corruption report.", + "confidence": 0.78, + "canonical_issue_reason": "No single canonical issue fits the whole cluster because most items describe different bugs. If one representative is needed, issue 43644 is the best anchor for the clear duplicate pair with 44534 about non-persistent buffers being filled with junk in Transformers v5.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43644 is the strongest representative issue in this set because it is specific, clearly reproducible, and has the cleanest duplicate linkage.", + "best_issue_reason": "Issue 43644 is the strongest representative because it predates 44534 and describes the core regression most directly: Transformers v5 corrupting non-persistent buffers during load.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:45663", "right": "issue:45823", "accept": false, - "reason": "Different models and failure modes: Gemma4 FSDP2 KeyError vs Gemma4 PLE device mismatch on auto device_map." + "reason": "Different Gemma4 failures: FSDP2 shared_kv_states KeyError vs device_map='auto' PLE device mismatch." }, { "left": "issue:44155", "right": "issue:45584", "accept": false, - "reason": "Different code paths: AudioFlamingo3 track leakage vs Whisper empty-transcription handling after align_special_tokens." + "reason": "Different models and failure modes: AudioFlamingo3 batched leakage vs Whisper empty-transcription generation." }, { "left": "issue:43366", "right": "issue:43531", "accept": false, - "reason": "Unrelated: GGUF gpt-oss support request vs Qwen3-MoE sliding_window bug." + "reason": "Unrelated Qwen3-MoE sliding_window bug vs GGUF gpt-oss architecture support request." }, { "left": "issue:43950", "right": "issue:45834", "accept": false, - "reason": "Different regressions: non-persistent buffer corruption vs Kosmos2.5 long-OCR index error." + "reason": "Non-persistent buffer regression vs Kosmos2.5 long-ocr index error are different bugs." }, { "left": "issue:43824", "right": "issue:43957", "accept": false, - "reason": "Different issues: missing export for Qwen2_5_VLForConditionalGeneration vs meta-device loading problems across some models." + "reason": "Different loading problems: missing Qwen2_5_VL export vs meta-device loading issues across models." }, { "left": "issue:44560", "right": "issue:45685", "accept": false, - "reason": "Different models and subsystems: Qwen3-vl-embedding video StopIteration vs MoE MPS histogram kernel issue." + "reason": "Different subsystems and errors: Qwen3-vl StopIteration on video vs MPS histogram not implemented for Int." }, { "left": "issue:43550", "right": "issue:44910", "accept": false, - "reason": "Different concrete bugs: torch.compile + SDPA on Bamba vs Flash Attention crashing on Qwen3.5 3D position_ids." + "reason": "Different code paths: Bamba torch.compile/SDPA failure vs Qwen3.5 Flash Attention crash from 3D position_ids." }, { "left": "issue:43645", "right": "issue:43646", "accept": true, - "reason": "Same underlying regression in custom model initialization under Transformers v5; one report is notebook-specific, the other is the general case." + "reason": "Same underlying Transformers v5 custom-model initialization regression; one is a notebook manifestation of the same breakage." }, { "left": "issue:42915", "right": "issue:43957", "accept": false, - "reason": "Different failures: Qwen3Moe FP8 config vs dequantized model loading with gpu+cpu offload." + "reason": "Qwen3-MoE FineGrainedFP8 failure is unrelated to meta-device loading regressions." }, { "left": "issue:43366", "right": "issue:44589", "accept": false, - "reason": "Unrelated: GGUF gpt-oss support vs Float8 storage-object loading error." + "reason": "GGUF gpt-oss support is unrelated to Float8 storage lookup failures." }, { - "left": "issue:42617", - "right": "issue:43366", + "left": "issue:44560", + "right": "issue:44918", "accept": false, - "reason": "No shared bug: 3d_parallel.py execution failure vs GGUF architecture support." + "reason": "Different Qwen3-vl problems: video StopIteration vs embedding unpacking with TRL SFT trainer." }, { - "left": "issue:44560", - "right": "issue:44918", + "left": "issue:42617", + "right": "issue:43366", "accept": false, - "reason": "Both Qwen3.5-related, but different failures: video StopIteration vs trl SFT embedding unpacking." + "reason": "3d_parallel.py execution failure is not the same bug as GGUF gpt-oss support." }, { "left": "issue:36010", "right": "issue:42915", "accept": false, - "reason": "Completely different: GenerationMixin import error vs Qwen3Moe FP8 failure." + "reason": "ImportError for GenerationMixin is unrelated to Qwen3Moe FineGrainedFP8 failure." }, { "left": "issue:44661", "right": "issue:45823", "accept": false, - "reason": "Different areas: tokenizer mapping-name restriction vs Gemma4 PLE device mismatch." + "reason": "TOKENIZER_MAPPING_NAMES add-new-model-like bug is unrelated to Gemma4 PLE device mismatch." }, { "left": "issue:44560", "right": "issue:44805", "accept": false, - "reason": "Different symptoms and code paths: video StopIteration vs mask/tensor shape mismatch." + "reason": "Different runtime errors in different Qwen3-vl contexts: video StopIteration vs mask/tensor shape mismatch." }, { "left": "issue:43632", "right": "issue:45834", "accept": false, - "reason": "Different regressions: _is_hf_initialized flag breakage vs Kosmos2.5 long-OCR index error." + "reason": "_is_hf_initialized flag regression is unrelated to Kosmos2.5 OCR index errors." }, { "left": "issue:38175", "right": "issue:43994", "accept": true, - "reason": "Same model family and same symptom class: SigLIP2 producing nonsensical/zero-probability outputs." + "reason": "Both report the same SigLIP2 bad-output regression on google/siglip2-base-patch16-224, just with different symptoms." }, { "left": "issue:43638", "right": "issue:44661", "accept": false, - "reason": "Different issues: ZeRO-3/BertModel index error vs add-new-model-like tokenizer-mapping failure." + "reason": "Zero-3 BertModel indexing error and add-new-model-like TOKENIZER_MAPPING_NAMES issue are distinct." }, { "left": "issue:43596", "right": "issue:44661", "accept": false, - "reason": "Different bugs: ZeRO-3 init index error vs tokenizer-mapping-name restriction." + "reason": "DeepSpeed ZeRO-3 BertModel index error is unrelated to TOKENIZER_MAPPING_NAMES handling." }, { "left": "issue:44661", "right": "issue:44805", "accept": false, - "reason": "Unrelated: tokenizer/model registration issue vs tensor mask shape mismatch." + "reason": "TOKENIZER_MAPPING_NAMES bug and mask/tensor shape mismatch are unrelated." }, { "left": "issue:44877", "right": "issue:45030", "accept": false, - "reason": "Both involve config validation, but the titles point to different model configs and likely different fixes." + "reason": "Strict config loading for granite_speech is not the same as glm4v config validation failure." }, { "left": "issue:43317", "right": "issue:44756", "accept": false, - "reason": "Different concerns: dequantized model offload failure vs mmap OOM on Strix Halo." + "reason": "device_map/offload loading failure is unrelated to the Strix Halo mmap OOM workaround request." }, { "left": "issue:43644", "right": "issue:44534", "accept": true, - "reason": "Near-identical duplicate reports of Transformers v5 filling non-persistent buffers with junk." + "reason": "Same regression: Transformers v5 fills non-persistent buffers with junk; 44534 is a later duplicate phrasing of the same bug." }, { "left": "issue:38175", "right": "issue:39692", "accept": false, - "reason": "Both mention SigLIP2, but one is bad model outputs while the other is a broken documentation example with additional issues." + "reason": "SigLIP2 wrong outputs are not the same as the documentation example's model/processor mismatch and quantization errors." }, { "left": "issue:42491", "right": "issue:43720", "accept": false, - "reason": "Different topics: Qwen3-MoE LoRA compatibility across versions vs BitNet packed-weight loading." + "reason": "Qwen3 MoE LoRA hf4.x/hf5.x incompatibility is unrelated to BitNet packed-weight loading in accelerate." }, { "left": "issue:41720", "right": "issue:42915", "accept": false, - "reason": "Different concrete failures: auto device mapping cudaErrorAssert vs FineGrainedFP8Config on Qwen3Moe." + "reason": "Qwen3 auto device-mapping cudaErrorAssert and Qwen3Moe FineGrainedFP8 are different failures." }, { "left": "issue:44960", "right": "issue:44977", "accept": false, - "reason": "Different models and bugs: GLM5 vs Qwen3.5 flash-attention generation failure." + "reason": "GLM5 and Qwen3.5 flash-attention generation failures are unrelated bugs." }, { "left": "issue:44451", "right": "issue:44488", - "accept": true, - "reason": "Same underlying model-loading regression affecting different BERT checkpoints; both report current/latest version cannot load the model." + "accept": false, + "reason": "Both are model-loading failures, but they concern different checkpoints and do not clearly share the same concrete bug." }, { "left": "issue:45313", "right": "issue:45357", "accept": false, - "reason": "Different Qwen3.5 bugs: ZeRO-3 weight loading vs incorrect visual encoder keys on save_pretrained." + "reason": "Qwen3.5 ZeRO-3 weight loading and save_pretrained visual-encoder key saving are different regressions." }, { - "left": "issue:39692", - "right": "issue:43994", + "left": "issue:42915", + "right": "issue:44756", + "accept": false, + "reason": "Qwen3Moe FP8 failure is unrelated to Strix Halo mmap OOM mitigation." + }, + { + "left": "issue:42491", + "right": "issue:45276", + "accept": false, + "reason": "Qwen3 MoE LoRA compatibility is unrelated to Gemma4 resize_token_embeddings not propagating." + }, + { + "left": "issue:43708", + "right": "issue:44075", "accept": false, - "reason": "Same model family but different problem type: docs example errors vs inference producing nonsensical outputs." + "reason": "Trainer resume/max_steps calculation and SGD args not used are different Trainer bugs." } ] }, "evaluator_result": { "accept": true, - "feedback": "The clustering is grounded in the packet and the accepted soft-edge pairs are conservative enough: 43644/44534, 43645/43646, 38175/43994, and 44451/44488 are all strongly title-consistent duplicates. The rejected pairs are generally cautious and do not overclaim beyond the available evidence. Summary is appropriately conservative." + "feedback": "The summary is grounded in the packet and stays conservative. The accepted soft pairs are plausible duplicates based on the issue titles, and the rejected pairs are framed narrowly without overclaiming. The representative issue choice for the non-persistent-buffer regression is also supported by the packet." }, "evaluator_used": true, "retried": false, @@ -1933,13 +1971,13 @@ "component_index": 1, "component_count": 17, "review_unit_index": 6, - "review_unit_count": 49, + "review_unit_count": 50, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-427", + "cluster_id": "cluster-10000-428", "nodes": [ "issue:36010", "issue:39404", @@ -1960,6 +1998,7 @@ "issue:43873", "issue:43931", "issue:43957", + "issue:43994", "issue:44155", "issue:44164", "issue:44291", @@ -1992,6 +2031,7 @@ "soft_pairs": [ "issue:45488|issue:45812", "issue:44488|issue:44991", + "issue:39692|issue:43994", "issue:43644|issue:43645", "issue:43257|issue:43854", "issue:43854|issue:43931", @@ -2003,16 +2043,16 @@ "issue:42491|issue:45292", "issue:44910|issue:45588", "issue:44164|issue:44291", - "issue:36010|issue:39404", "issue:45276|issue:45335", + "issue:36010|issue:39404", "issue:41720|issue:44155", "issue:43404|issue:43454", "issue:39692|issue:44960", "issue:43425|issue:43957", - "issue:42947|issue:45704", "issue:44402|issue:44488", - "issue:45663|issue:45759", + "issue:42947|issue:45704", "issue:43873|issue:44387", + "issue:45663|issue:45759", "issue:44291|issue:44756", "issue:45663|issue:45859", "issue:44155|issue:45823", @@ -2026,254 +2066,260 @@ "issue:43854|issue:45216", "issue:43381|issue:44928" ], - "prepared_review_unit_hash": "27e2ce7913dba9da643da04bd087aa1d397130f9a1b068b615384f8bd6f9ae4b", + "prepared_review_unit_hash": "120350c1ca5b6d606835b0f2a1df39ef2dbaa6f41d79eed764191b09103f9e63", "component_budget": { - "node_count": 427, - "item_count": 427, - "soft_pair_count": 1576, - "serialized_chars": 816182, - "estimated_input_tokens": 204046, - "estimated_eval_tokens": 408348 + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 }, "budget": { - "node_count": 47, - "item_count": 47, - "soft_pair_count": 35, - "serialized_chars": 37753, - "estimated_input_tokens": 9439, - "estimated_eval_tokens": 19134 + "node_count": 48, + "item_count": 48, + "soft_pair_count": 36, + "serialized_chars": 38639, + "estimated_input_tokens": 9660, + "estimated_eval_tokens": 19576 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This set is mostly a grab bag of unrelated Transformers issues. The only clear duplicate cluster is around `resize_token_embeddings` failing to propagate to downstream embeddings/output heads (45276/45292/45335).", - "confidence": 0.82, - "canonical_issue_reason": "Issue 45276 is the strongest representative of the only coherent duplicate cluster: it describes the core `resize_token_embeddings` propagation bug, and it has direct near-duplicates/narrower variants in 45292 and 45335.", + "summary": "This cluster is still mostly unrelated, with two plausible same-area issue families: `resize_token_embeddings` regressions around 45276/45292/45335 and `lm_head` weight-tying reports around 43404/43454. However, the packet only shows title-level similarity, so I would not treat those soft edges as confirmed duplicates. No PRs are present.", + "confidence": 0.61, + "canonical_issue_reason": "45276 is the best anchor for the resize-token-embedding subgroup because it is the broadest concrete report in that area, but the related items should be treated as plausible regressions rather than confirmed duplicates.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 45276 is the best overall issue to anchor triage because it is the clearest, broadest statement of the shared resize-token-embedding regression and matches the accepted duplicate neighborhood.", + "best_issue_reason": "45276 is the strongest single issue to represent the cluster\u2019s main coherent subgroup, though the evidence is still too weak to collapse the nearby variants into one confirmed duplicate set.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:45488", "right": "issue:45812", "accept": false, - "reason": "Both are tokenizer regressions, but they affect different tokenizer stacks/models and describe different failure mechanisms (`LlamaTokenizer` pre-tokenizer override vs wrong token IDs)." + "reason": "Both are tokenizer/model-output regressions, but they affect different model families and the packet does not show the same underlying bug." }, { "left": "issue:44488", "right": "issue:44991", "accept": false, - "reason": "Both are tokenizer loading failures, but the affected models and reported breakages are different; no sign of the same underlying bug." + "reason": "Both are loading failures, but for different models and different failure descriptions; title similarity is not enough to merge them." + }, + { + "left": "issue:39692", + "right": "issue:43994", + "accept": false, + "reason": "Same model family is mentioned, but one is a docs/example problem and the other is an inference quality issue." }, { "left": "issue:43644", "right": "issue:43645", "accept": false, - "reason": "Both are Transformers 5 regressions, but one is about non-persistent buffers being junk-filled and the other is about custom model initialization in Jupyter notebooks." + "reason": "Both are Transformers 5.0.0 regressions, but they describe different code paths and symptoms." }, { "left": "issue:43257", "right": "issue:43854", "accept": false, - "reason": "Different models and different failure modes; both involve loading, but not the same concrete code path." + "reason": "Different model-loading defects with different targets; no shared concrete bug is evident." }, { "left": "issue:43854", "right": "issue:43931", "accept": false, - "reason": "Unrelated load failures for different models; one is a unit-test loading issue, the other is a shape mismatch on a Qwen3-VL checkpoint." + "reason": "Both concern loading, but the model families and reported failures are different." }, { "left": "issue:43856", "right": "issue:43873", "accept": false, - "reason": "Both mention memory/performance, but one is Qwen3 MoE training memory usage and the other is quantization/offloading behavior; not the same bug." + "reason": "Both mention efficiency/memory, but one is training-memory behavior and the other is quantization/offloading behavior." }, { "left": "issue:42491", "right": "issue:43931", "accept": false, - "reason": "Different models and unrelated load-time failures; no shared underlying bug." + "reason": "Distinct issues: LoRA compatibility across branches versus a shape-mismatch load failure." }, { "left": "issue:45362", "right": "issue:45406", "accept": false, - "reason": "One is a Qwen3.5 chat crash, the other is a Gemma4 processor attribute error in serve; unrelated." + "reason": "Different components and failure modes; no clear common code-path bug from the packet." }, { "left": "issue:43856", "right": "issue:44387", "accept": false, - "reason": "Both are memory-related, but one is Qwen3 MoE training memory usage and the other is int4 quantization reserved-memory/OOM; too broad to be duplicates." + "reason": "Both are memory-related, but they point to different root causes and different workflows." }, { "left": "issue:44960", "right": "issue:46050", "accept": false, - "reason": "Different models and different save/load regressions; not the same issue." + "reason": "Unrelated model/version report versus a save/load corruption report." }, { "left": "issue:42491", "right": "issue:45292", "accept": false, - "reason": "Unrelated: LoRA/qwen3_moe compatibility versus a token-embedding resize propagation bug." + "reason": "Different model/version compatibility problem versus a resize-token-embedding regression." }, { "left": "issue:44910", "right": "issue:45588", "accept": false, - "reason": "Both touch attention code, but one is a Qwen3.5 3D position_ids flash-attention crash and the other is an `s_aux=None` AttributeError in sink-less models." + "reason": "Both involve attention code, but the failure modes are different and not mergeable as one fix." }, { "left": "issue:44164", "right": "issue:44291", "accept": false, - "reason": "Both involve model serialization/loading internals, but the concrete bugs differ (`extra_state` handling vs unexpected `_is_hf_initialized` during init_empty_weights)." + "reason": "Save/load extra_state handling is unrelated to init_empty_weights argument handling." }, { - "left": "issue:36010", - "right": "issue:39404", + "left": "issue:45276", + "right": "issue:45335", "accept": false, - "reason": "Import error for `GenerationMixin` versus Whisper pipeline `return_language` regression; unrelated." + "reason": "They look like the same family, but the packet only gives similar titles and not enough evidence to confirm the same underlying regression." }, { - "left": "issue:45276", - "right": "issue:45335", - "accept": true, - "reason": "Same underlying `resize_token_embeddings` propagation bug: both say resized embeddings are not reflected in output/decoder embeddings, just on different model families." + "left": "issue:36010", + "right": "issue:39404", + "accept": false, + "reason": "ImportError for GenerationMixin is unrelated to Whisper pipeline `return_language` behavior." }, { "left": "issue:41720", "right": "issue:44155", "accept": false, - "reason": "Different model/runtime problems: Qwen3 auto device-mapping crash versus AudioFlamingo3 batched inference token/embedding leakage." + "reason": "Different models and failure modes; they are not the same concrete bug." }, { "left": "issue:43404", "right": "issue:43454", "accept": false, - "reason": "Both are weight-tying complaints, but they target different model classes and there is no evidence they are the same code path bug." + "reason": "Both are lm_head weight-tying complaints, but the packet only shows title similarity across different model families." }, { "left": "issue:39692", "right": "issue:44960", "accept": false, - "reason": "SigLIP2 docs/example errors are unrelated to the GLM5 model-loading issue." + "reason": "Completely different topics: SigLIP2 example errors versus a GLM5 issue." }, { "left": "issue:43425", "right": "issue:43957", "accept": false, - "reason": "Both are compatibility/load issues, but one is torch version incompatibility and the other is meta-device loading regressions; not duplicates." - }, - { - "left": "issue:42947", - "right": "issue:45704", - "accept": false, - "reason": "Completely different topics: PEFT gradient checkpointing versus an apex RMSNorm memory leak." + "reason": "Unrelated compatibility/load problems with different failure surfaces." }, { "left": "issue:44402", "right": "issue:44488", "accept": false, - "reason": "Both are model/tokenizer loading issues, but they concern different models and likely different root causes." + "reason": "Both are tokenizer/vocab reports, but they concern different models and different loading symptoms." }, { - "left": "issue:45663", - "right": "issue:45759", + "left": "issue:42947", + "right": "issue:45704", "accept": false, - "reason": "Different model architectures and different bugs: Gemma-4 FSDP2 attention state versus Qwen multimodal config unwrapping." + "reason": "Gradient-checkpointing/PEFT behavior is unrelated to the RMSNorm memory leak report." }, { "left": "issue:43873", "right": "issue:44387", "accept": false, - "reason": "Both are memory-related, but one is quantization/offloading behavior and the other is a specific int4 reserved-memory regression." + "reason": "Quantization/offloading behavior and int4 OOM are different defects." + }, + { + "left": "issue:45663", + "right": "issue:45759", + "accept": false, + "reason": "Different composite-model issues affecting different code paths." }, { "left": "issue:44291", "right": "issue:44756", "accept": false, - "reason": "Meta initialization TypeError and mmap-on-Strix-Halo OOM are unrelated." + "reason": "Different loading/init bug versus a platform-specific mmap/OOM issue." }, { "left": "issue:45663", "right": "issue:45859", "accept": false, - "reason": "Different models and code paths: Gemma-4 FSDP2 attention state versus Qwen3_5 MoE rotary embedding CPU-offload incompatibility." + "reason": "FSDP2 attention-state bug is not the same as CPU-offload incompatibility." }, { "left": "issue:44155", "right": "issue:45823", "accept": false, - "reason": "Different subsystems and failures: batched inference leakage versus device-map mismatch in Gemma4 PLE." + "reason": "Different model families and unrelated forward/device mismatch behavior." }, { "left": "issue:44292", "right": "issue:44912", "accept": false, - "reason": "Both are quantization-related loading issues, but they involve different quantization formats and symptoms (NVFP4 load error vs MXFP4 fallback to bf16)." + "reason": "Both are quantization-related, but they affect different models and failure modes." }, { "left": "issue:44387", "right": "issue:44756", "accept": false, - "reason": "Different OOM causes; one is int4 quantization memory growth, the other is mmap on Strix Halo." + "reason": "Int4 quantization OOM is unrelated to disabling mmap on Strix Halo." }, { "left": "issue:43335", "right": "issue:45663", "accept": false, - "reason": "SwitchTransformers sparse-layer construction and Gemma-4 FSDP2 shared_kv_states are unrelated bugs." + "reason": "Sparse-layer config bug is unrelated to the Gemma-4 FSDP2 attention-state issue." }, { "left": "issue:43957", "right": "issue:44756", "accept": false, - "reason": "Meta-device loading regressions and Strix Halo mmap/OOM are different issues." + "reason": "Meta-device loading failures and Strix Halo mmap/OOM are not the same bug." }, { "left": "issue:42947", "right": "issue:43381", "accept": false, - "reason": "Both mention gradient checkpointing, but one is ineffective PEFT LoRA checkpointing and the other is checkpointing in eval mode; different bugs." + "reason": "Both mention gradient checkpointing, but they describe different constraints and code paths." }, { "left": "issue:45276", "right": "issue:45292", - "accept": true, - "reason": "Same resize-token-embedding regression: both report resized embeddings not propagating to the model outputs/embedding modules." + "accept": false, + "reason": "This is the same title family, but the packet does not provide enough detail to confirm the exact same underlying regression." }, { "left": "issue:44164", "right": "issue:44756", "accept": false, - "reason": "Serialization `extra_state` handling and mmap-related OOM are unrelated." + "reason": "Save/load extra_state handling is unrelated to mmap-induced OOM." }, { "left": "issue:43854", "right": "issue:45216", "accept": false, - "reason": "Different model-loading regressions for different models and different root causes." + "reason": "Different model-loading regressions with different symptoms and targets." }, { "left": "issue:43381", "right": "issue:44928", "accept": false, - "reason": "Related to gradient/checkpointing and attention behavior, but not the same concrete bug; one is eval-mode checkpointing and the other is Qwen3.5 3D position_ids causing NaNs via SDPA fallback." + "reason": "Eval-mode checkpointing restriction is unrelated to the RLHF gradient-explosion report." } ] }, "evaluator_result": { - "accept": true, - "feedback": "The triage summary is grounded in the packet, and the one accepted cluster is conservative and well-supported: 45276/45292/45335 all describe the same resize_token_embeddings propagation problem across related model variants. The other pairwise rejections are appropriately cautious given the differing titles and apparent failure modes." + "accept": false, + "feedback": "The summary is mostly grounded, but it overstates confidence on the duplicate subclusters. In particular, the analyst marks 45276\u219445335 and 43404\u219443454 (and 45276\u219445292) as true duplicates even though the packet only shows similar titles, not enough to justify \u2018same underlying bug\u2019 with certainty. Those are better treated as plausible same-area regressions rather than confirmed duplicates. The overall cluster being mostly unrelated is fine, and \u2018No PRs are present\u2019 is correct, but the soft-edge acceptances are not conservative enough." }, "evaluator_used": true, - "retried": false, - "accepted_nontrivial_soft_edge": true, + "retried": true, + "accepted_nontrivial_soft_edge": false, "error_kind": null, "error_message": null }, @@ -2282,13 +2328,13 @@ "component_index": 1, "component_count": 17, "review_unit_index": 7, - "review_unit_count": 49, + "review_unit_count": 50, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-427", + "cluster_id": "cluster-10000-428", "nodes": [ "issue:36032", "issue:39404", @@ -2369,24 +2415,24 @@ "issue:44164|issue:45759", "issue:36032|issue:43452", "issue:44671|issue:44991", - "issue:39404|issue:44671", "issue:43596|issue:44683", + "issue:39404|issue:44671", "issue:43856|issue:44756" ], - "prepared_review_unit_hash": "04b758e1d17dcbff8e304ed02d3a78c9a1a7206e502615b16a56928215f60eae", + "prepared_review_unit_hash": "cc646bd8f79efe8341865da36b0cefa834062491e38a8cd075722dace945986f", "component_budget": { - "node_count": 427, - "item_count": 427, - "soft_pair_count": 1576, - "serialized_chars": 816182, - "estimated_input_tokens": 204046, - "estimated_eval_tokens": 408348 + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 32, - "serialized_chars": 37358, + "serialized_chars": 37357, "estimated_input_tokens": 9340, "estimated_eval_tokens": 18936 }, @@ -2395,120 +2441,120 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly a grab-bag of unrelated Transformers v5 regressions. Only two soft edges look like true duplicates: custom model initialization (43645/43646) and resize_token_embeddings not propagating to tied output embeddings (45292/45335).", - "confidence": 0.87, - "canonical_issue_reason": "issue:43646 is the broadest representative of the custom-model initialization regression; issue:43645 is a narrower Jupyter-notebook repro of the same underlying bug.", + "summary": "The cluster is mostly a loose collection of distinct Transformers v5 regressions and model-specific loading/compatibility bugs. Only one soft pair looks like a true duplicate: the two resize_token_embeddings reports.", + "confidence": 0.84, + "canonical_issue_reason": "issue:43611 is the broadest representative of the generic Transformers 5 model-loading regressions in this set; the other issues are narrower model-, pipeline-, or docs-specific failures.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43646 best fits the main duplicate-worthy bug in the set and is phrased more generally than the notebook-specific repro.", + "best_issue_reason": "issue:43611 is the best single summary item for the cluster\u2019s main theme of v5 loading/init breakages, even though most other items are not duplicates of it.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:45684", "right": "issue:45698", "accept": false, - "reason": "Both involve save/from_pretrained custom-model flows, but one is permissions propagation and the other is loading the wrong module; not the same concrete bug." + "reason": "Both involve custom-model save/load flows, but one is about read-only source-file permissions and the other about loading the wrong module after save_pretrained; different bugs." }, { "left": "issue:44479", "right": "issue:44877", "accept": false, - "reason": "Different model families and different failures; same broad v5 regression theme is not enough." + "reason": "Different failure modes and model families: video-input regression vs strict-config loading for granite_speech." }, { "left": "issue:45588", "right": "issue:45663", "accept": false, - "reason": "Unrelated code paths: flash_attention sink handling vs Gemma4 FSDP2 shared_kv_states rebuilding." + "reason": "Sink-less flash_attention crash vs Gemma4 FSDP2 shared_kv_states rebuild issue; unrelated code paths." }, { "left": "issue:43646", "right": "issue:43708", "accept": false, - "reason": "Custom model initialization and Trainer resume/max_steps are unrelated bugs." + "reason": "Custom model initialization in notebooks vs Trainer resume_from_checkpoint max_steps calculation; not the same bug." }, { "left": "issue:43645", "right": "issue:44164", "accept": false, - "reason": "Notebook custom-model initialization is unrelated to save/from_pretrained extra_state handling." + "reason": "Notebook custom-model init regression vs save/from_pretrained extra_state handling; different underlying problems." }, { "left": "issue:45292", "right": "issue:45335", "accept": true, - "reason": "Same underlying bug: resize_token_embeddings does not update tied output embeddings; 45335 is a T5Gemma-specific instance." + "reason": "Same bug: resize_token_embeddings updates output embeddings incorrectly, including t5gemma decoder.embed_tokens." }, { "left": "issue:43828", "right": "issue:43957", "accept": false, - "reason": "Both mention dtype/loading issues, but the model paths and failure modes differ." + "reason": "Both are dtype/loading regressions, but one is autocast on Phi-tiny-MoE and the other is torch.device('meta') loading; distinct issues." }, { "left": "issue:44402", "right": "issue:44451", "accept": false, - "reason": "A tokenizer vocab-size mismatch and a generic model load failure are not enough to treat as the same bug." + "reason": "Both concern tokenizer/model loading, but the first is a vocab-size mismatch report and the second is a generic ScandiBERT load failure; not enough evidence of a single bug." }, { "left": "issue:43425", "right": "issue:43828", "accept": false, - "reason": "Torch version incompatibility is a different problem from an autocast dtype mismatch." + "reason": "Torch version incompatibility is too broad; the concrete symptoms and code paths differ." }, { "left": "issue:43257", "right": "issue:44863", "accept": false, - "reason": "Different models and different loading failures; no clear shared code-path bug." + "reason": "Different model-specific checkpoint loading problems: Qwen3 MoE conversion vs NemotronH checkpoint loading." }, { "left": "issue:39404", "right": "issue:45701", "accept": false, - "reason": "Whisper pipeline return_language and generic tokenization changes are different issues." + "reason": "Whisper pipeline return_language regression and version-dependent tokenization changes are unrelated." }, { "left": "issue:44589", "right": "issue:44756", "accept": false, - "reason": "Float8 storage lookup failure is unrelated to mmap OOM behavior." + "reason": "Float8 storage lookup error vs disabling mmap on Strix Halo to avoid OOM; unrelated symptoms and fixes." }, { "left": "issue:44683", "right": "issue:45823", "accept": false, - "reason": "Compiled flex_attention on newer torch and Gemma4 PLE device mismatch are unrelated." + "reason": "Compiled flex_attention on torch>=2.9 and Gemma4 PLE device mismatch are different failures." }, { "left": "issue:43425", "right": "issue:45823", "accept": false, - "reason": "Torch incompatibility and Gemma4 device mismatch are different bugs." + "reason": "Generic torch incompatibility does not match the specific Gemma4 PLE device-mapping bug." }, { "left": "issue:43611", "right": "issue:43645", "accept": false, - "reason": "base_model_prefix loading failures are not the same as the custom-model initialization regression." + "reason": "Both are Transformers 5 loading/init regressions, but one is base_model_prefix loading and the other is Jupyter custom-model initialization; separate bugs." }, { "left": "issue:43425", "right": "issue:43716", "accept": false, - "reason": "Torch 2.10 incompatibility and Mistral-3 dtype mismatch do not share the same concrete bug." + "reason": "Torch incompatibility vs Mistral-3 image preprocessor/model dtype mismatch; not the same issue." }, { "left": "issue:44530", "right": "issue:45084", "accept": false, - "reason": "PagedAttentionCache linear_attention handling is unrelated to template-node compilation errors." + "reason": "PagedAttentionCache linear_attention crash and non-template node compilation error are different failures." }, { "left": "issue:43906", "right": "issue:44530", "accept": false, - "reason": "A reproduction of another issue and a Qwen3.5 cache crash are different problems." + "reason": "An isolated reproduction of another issue vs a Qwen3.5 PagedAttentionCache crash; not duplicates." }, { "left": "issue:39404", @@ -2520,85 +2566,85 @@ "left": "issue:44265", "right": "issue:44898", "accept": false, - "reason": "torch.export/exportability and Perceiver resolution handling are different code paths." + "reason": "torch.export with torch_compilable_check and Perceiver image classification failure are different model/code-path issues." }, { "left": "issue:44479", "right": "issue:44805", "accept": false, - "reason": "Video-input regression and mask-shape IndexError are not the same bug." + "reason": "Video-input regression for Qwen VL models vs a generic mask/tensor shape mismatch; no clear duplicate evidence." }, { "left": "issue:43901", "right": "issue:44509", "accept": false, - "reason": "Both are docs issues, but they cover different removed/legacy pipeline mentions." + "reason": "Both are docs-related, but one is about return_all_scores docs drift and the other about removed pipeline tasks; separate documentation issues." }, { "left": "issue:43994", "right": "issue:44079", "accept": false, - "reason": "SigLIP2 nonsense outputs and ModelOutput key assignment are unrelated." + "reason": "Siglip2 nonsensical outputs and ModelOutput key assignment are different bugs." }, { "left": "issue:43425", "right": "issue:44918", "accept": false, - "reason": "Torch version incompatibility and Qwen3.5 embedding unpacking failure are different issues." + "reason": "Generic torch incompatibility vs Qwen3.5 embedding unpacking with TRL SFT trainer; unrelated." }, { "left": "issue:43299", "right": "issue:43931", "accept": false, - "reason": "Same model family, but the reported symptoms are different enough that they do not look like one concrete bug." + "reason": "Both are Qwen3-VL loading reports, but one is a MoE loading regression in transformers 5.0.0.dev0 and the other is a weight-shape mismatch for a specific checkpoint; not the same concrete bug." }, { "left": "issue:44589", "right": "issue:45084", "accept": false, - "reason": "Float8 storage lookup and non-template-node compilation failures are unrelated." + "reason": "Different compile/runtime failures with no shared code path." }, { "left": "issue:44164", "right": "issue:45759", "accept": false, - "reason": "extra_state save/load handling is unrelated to from_config text_config unwrapping." + "reason": "extra_state save/load handling is unrelated to AutoModelForCausalLM.from_config text_config unwrapping." }, { "left": "issue:36032", "right": "issue:43452", "accept": false, - "reason": "Tokenizer special-token method conflict is unrelated to gguf_file from_pretrained breakage." + "reason": "Tokenizer special-token attribute conflict vs gguf_file breaking AutoTokenizer/AutoModelForCausalLM.from_pretrained; distinct problems." }, { "left": "issue:44671", "right": "issue:44991", "accept": false, - "reason": "CamemBERT MLM predictions and loading a tokenizer for est-roberta are different bugs." + "reason": "CamemBERT masked-LM prediction regression and EMBEDDIA tokenizer loading failure are different bugs." }, { - "left": "issue:39404", - "right": "issue:44671", + "left": "issue:43596", + "right": "issue:44683", "accept": false, - "reason": "Whisper pipeline behavior and CamemBERT MLM outputs are unrelated." + "reason": "Deepspeed ZeRO3/BertModel index error and compiled flex_attention on torch>=2.9 are unrelated." }, { - "left": "issue:43596", - "right": "issue:44683", + "left": "issue:39404", + "right": "issue:44671", "accept": false, - "reason": "Deepspeed zero3/BertModel indexing failure is unrelated to flex_attention compile issues." + "reason": "Whisper pipeline regression and CamemBERT masked-LM regression affect different models and code paths." }, { "left": "issue:43856", "right": "issue:44756", "accept": false, - "reason": "Qwen3 MoE training memory usage and Strix Halo mmap OOM are different performance problems." + "reason": "Qwen3 MoE training memory usage and Strix Halo mmap OOM avoidance are different performance/memory issues, not duplicates." } ] }, "evaluator_result": { "accept": true, - "feedback": "The summary is grounded in the packet and stays conservative overall. The lone accepted soft edge, issue:45292 vs issue:45335, is a reasonable same-bug call based on the titles, and the other pairs are rejected with appropriately cautious reasoning. The note about custom model initialization is also supported by the two issue titles, even though that exact pair is not one of the listed soft-edge verdicts." + "feedback": "The summary is grounded in the packet and stays conservative: it characterizes the cluster as mostly distinct v5 loading/init regressions, and it only treats the resize_token_embeddings pair (issue:45292 / issue:45335) as a true duplicate, which is well supported by the titles and linkage. The other soft-edge denials are phrased cautiously and do not overstate duplicate evidence." }, "evaluator_used": true, "retried": false, @@ -2611,13 +2657,13 @@ "component_index": 1, "component_count": 17, "review_unit_index": 8, - "review_unit_count": 49, + "review_unit_count": 50, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-427", + "cluster_id": "cluster-10000-428", "nodes": [ "issue:32090", "issue:34567", @@ -2681,15 +2727,15 @@ "issue:40990|issue:43421", "issue:43716|issue:45084", "issue:44206|issue:44220", - "issue:36032|issue:42915", "issue:42175|issue:46178", - "issue:43643|issue:44530", + "issue:36032|issue:42915", "issue:43883|issue:44488", + "issue:43643|issue:44530", "issue:34567|issue:40444", "issue:44811|issue:45584", "issue:42915|issue:44530", - "issue:43541|issue:45084", "issue:44451|issue:44987", + "issue:43541|issue:45084", "issue:43644|issue:43646", "issue:44464|issue:45987", "issue:43452|issue:43482", @@ -2702,226 +2748,220 @@ "issue:43262|issue:43377", "issue:43089|issue:45987" ], - "prepared_review_unit_hash": "93ac9d0bacd2aeb50ec17a7163c990d4ae26c97b9d6cf48b0eb626411682178a", + "prepared_review_unit_hash": "130286c9eb703ed8cbcffb2f777d229f885238245502ce7aa3682eec97473f6a", "component_budget": { - "node_count": 427, - "item_count": 427, - "soft_pair_count": 1576, - "serialized_chars": 816182, - "estimated_input_tokens": 204046, - "estimated_eval_tokens": 408348 + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 32, - "serialized_chars": 37337, - "estimated_input_tokens": 9335, - "estimated_eval_tokens": 18926 + "serialized_chars": 37342, + "estimated_input_tokens": 9336, + "estimated_eval_tokens": 18928 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is a loose embedding-based grouping of mostly unrelated issues. They share broad Transformers ecosystem themes, but the soft pairs do not describe the same concrete bug or change, so none should be merged as duplicates.", - "confidence": 0.96, - "canonical_issue_reason": null, + "summary": "This cluster is mostly a loose bag of unrelated issue reports. A few titles share broad themes (Transformers v5 regressions, Whisper/audio, model loading), but none of the soft-edge pairs look like the same concrete bug/change.", + "confidence": 0.92, + "canonical_issue_reason": "Issue 43646 is the broadest Transformers-v5 regression report in the set and the best umbrella representative if one must be chosen, but the cluster does not contain a true duplicate set.", "canonical_pr_reason": null, - "best_issue_reason": null, + "best_issue_reason": "Issue 43646 is the best single issue to anchor the cluster because it is the most general and sits near several other v5 initialization/loading regressions; still, it should be treated as a loose thematic representative, not a duplicate canonical.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44991", "right": "issue:45081", "accept": false, - "reason": "Both involve tokenizer loading, but one is a generic tokenizer failure on a specific model and the other is a Mistral regex patch crash; different code paths and fixes." + "reason": "Both are loading failures, but one is a tokenizer load regression for EMBEDDIA/est-roberta and the other is a Mistral regex patch crash; different code paths and symptoms." }, { "left": "issue:32090", "right": "issue:39404", "accept": false, - "reason": "Trainer broadcast NoneType error and Whisper return_language pipeline regression are unrelated subsystems and failure modes." + "reason": "Trainer GPU broadcast NoneType error and Whisper return_language pipeline regression are unrelated subsystems and failure modes." }, { "left": "issue:44164", "right": "issue:44479", "accept": false, - "reason": "save/from_pretrained extra_state handling and Qwen video input regression are separate model lifecycle vs multimodal input bugs." + "reason": "save/from_pretrained extra_state handling and Qwen video-input regression are different bugs with no shared concrete code path." }, { "left": "issue:44737", "right": "issue:45859", "accept": false, - "reason": "Both mention device/offload behavior, but XLNet CPU arange work and Qwen rotary embedding CPU offload are distinct implementations." + "reason": "Both mention device/offload behavior, but XLNet CPU arange placement and Qwen3_5Moe CPU offload are distinct model-specific issues." }, { "left": "issue:43582", "right": "issue:43632", "accept": false, - "reason": "Apple Silicon allocator warmup TypeError is unrelated to the Transformers v5 _is_hf_initialized flag regression." + "reason": "Apple Silicon allocator warmup TypeError and _is_hf_initialized flag breakage are unrelated regression areas." }, { "left": "issue:44661", "right": "issue:44683", "accept": false, - "reason": "Model registration helper failure and compiled flex_attention on torch>=2.9 are unrelated code paths." + "reason": "Tokenizer mapping-name handling and compiled flex_attention on torch>=2.9 are separate features and failure paths." }, { "left": "issue:44488", "right": "issue:45081", "accept": false, - "reason": "Both are tokenizer loading complaints, but they target different models and different root causes." + "reason": "Both are tokenization/model-loading reports, but one is a model load failure for cjvt/sleng-bert and the other is a Mistral-specific regex patch crash." }, { "left": "issue:44488", "right": "issue:44987", "accept": false, - "reason": "Both concern loading models, but the affected models and failure mechanisms differ; not the same defect." + "reason": "Both fail on model loading, but they target different models and likely different underlying regressions." }, { "left": "issue:43611", "right": "issue:43644", "accept": false, - "reason": "Base model prefix loading breakage and non-persistent buffers being filled with junk are separate v5 regressions." + "reason": "Both are Transformers v5 regressions, but base_model_prefix loading and non-persistent buffer initialization are not the same bug." }, { "left": "issue:40990", "right": "issue:43421", "accept": false, - "reason": "High perplexity on one model/dataset and runtime special-token post-processor updates are unrelated issues." + "reason": "High perplexity on gpt-oss and TokenizersBackend post-processor updates are unrelated." }, { "left": "issue:43716", "right": "issue:45084", "accept": false, - "reason": "Image preprocessor dtype mismatch and template-node compilation error are different subsystems and bugs." + "reason": "Image preprocessor dtype mismatch and template-node compilation failure are unrelated code paths." }, { "left": "issue:44206", "right": "issue:44220", "accept": false, - "reason": "The first is a regression from an unsupported center arg; the second is a different fbank extraction issue with no clear same root cause." - }, - { - "left": "issue:36032", - "right": "issue:42915", - "accept": false, - "reason": "Tokenizer method-name conflict and Qwen3Moe FP8 config failure are unrelated." + "reason": "Both are audio feature-extraction related, but one is an unsupported center argument in LasrFeatureExtractor and the other is a different _torch_extract_fbank_features issue." }, { "left": "issue:42175", "right": "issue:46178", "accept": false, - "reason": "Missing tensorflow backend in a torch-only install and missing tensorflow_text in BACKENDS_MAPPING are related only at a very broad packaging level, not the same bug." + "reason": "Backend packaging omissions for tensorflow vs tensorflow_text mapping errors are different backend-definition problems." }, { - "left": "issue:43643", - "right": "issue:44530", + "left": "issue:36032", + "right": "issue:42915", "accept": false, - "reason": "Missing fields from AutoConfig with trust_remote_code and PagedAttentionCache invalid linear_attention group type are unrelated." + "reason": "T5 tokenizer add_special_tokens name conflict and Qwen3Moe FP8 config failure are unrelated." }, { "left": "issue:43883", "right": "issue:44488", "accept": false, - "reason": "Missing all_tied_weights_keys in Molmo and a tokenizer loading failure for cjvt/sleng-bert are different model-specific issues." + "reason": "Molmo tied-weights attribute error and tokenizer loading failures for sleng-bert are unrelated." + }, + { + "left": "issue:43643", + "right": "issue:44530", + "accept": false, + "reason": "trust_remote_code config field loss and PagedAttentionCache linear_attention crash are different bugs." }, { "left": "issue:34567", "right": "issue:40444", "accept": false, - "reason": "TrainerState token counter not updating and iterable dataset multi-image finetuning failure are not the same underlying problem." + "reason": "TrainerState token counting and multi-image IterableDataset finetuning failure are unrelated." }, { "left": "issue:44811", "right": "issue:45584", "accept": false, - "reason": "Whisper batch_decode skip_special_tokens handling and empty-transcription generation after align_special_tokens are distinct bugs." + "reason": "Both involve Whisper special-token handling, but batch_decode skip_special_tokens and empty-transcription generation after align_special_tokens are different concrete failures." }, { "left": "issue:42915", "right": "issue:44530", "accept": false, - "reason": "Qwen3Moe FP8 config failure and Qwen3.5 PagedAttentionCache invalid group type are different model/runtime failures." + "reason": "Qwen3Moe FineGrainedFP8Config failure and Qwen3.5 PagedAttentionCache linear_attention crash are distinct issues." }, { - "left": "issue:43541", - "right": "issue:45084", + "left": "issue:44451", + "right": "issue:44987", "accept": false, - "reason": "Grouped_mm tracing failure in Mixtral and template-node compilation failure are unrelated compile-time issues." + "reason": "Cannot load ScandiBERT and transformers>=5.1.0 fails loading physical-intelligence/fast are separate model-loading regressions." }, { - "left": "issue:44451", - "right": "issue:44987", + "left": "issue:43541", + "right": "issue:45084", "accept": false, - "reason": "Both are model-loading regressions, but they concern different models and likely different loader paths." + "reason": "Mixtral grouped_mm torchdynamo tracing failure and template-node compilation error are unrelated." }, { "left": "issue:43644", "right": "issue:43646", "accept": false, - "reason": "Non-persistent buffer initialization corruption and custom model initialization breakage are related only by the v5 release, not the same bug." + "reason": "Both are broad Transformers v5 regressions, but one is about non-persistent buffers and the other about custom model initialization; too different to merge." }, { "left": "issue:44464", "right": "issue:45987", "accept": false, - "reason": "Chunked generation inconsistency under compiled forward and StaticCache.get_seq_length returning a tensor are different generation/cache issues." + "reason": "Chunked generation inconsistency with compiled forward and StaticCache seq-length type violation are different generation/cache bugs." }, { "left": "issue:43452", "right": "issue:43482", "accept": false, - "reason": "gguf_file argument breakage in from_pretrained APIs and Qwen2.5-GGUF loading failure are separate GGUF-related regressions." + "reason": "Both touch GGUF loading, but one is a general gguf_file breakage for tokenizer/model loaders and the other is a Qwen2.5-GGUF-specific v5 failure." }, { "left": "issue:36032", "right": "issue:43645", "accept": false, - "reason": "Tokenizer special-token method conflict and Jupyter custom model initialization breakage are unrelated." + "reason": "Tokenizer special-token method conflict and Jupyter custom model initialization are unrelated." }, { "left": "issue:44849", "right": "issue:45325", "accept": false, - "reason": "Qwen3.5 hidden-state bug and Qwen2.5-VL rope index scaling bug are different model-specific failures." + "reason": "Qwen3.5 hidden-states bug and Qwen2.5-VL rope-index temporal scaling bug are different model-specific regressions." }, { "left": "issue:43317", "right": "issue:43482", "accept": false, - "reason": "Dequantized model loading with device_map auto/offload and Qwen2.5-GGUF loading failure are different loading scenarios." + "reason": "Dequantized model offload loading failure and Qwen2.5-GGUF loading failure are separate load paths and not the same bug." }, { "left": "issue:44987", "right": "issue:44991", "accept": false, - "reason": "Both are versioned loading failures, but they hit different models and do not indicate the same code-path problem." - }, - { - "left": "issue:43883", - "right": "issue:44991", - "accept": false, - "reason": "Molmo tied-weights attribute failure and tokenizer loading failure for a different model are unrelated." + "reason": "Both are version-gated loading failures, but they affect different models and no shared concrete defect is evident." }, { "left": "issue:45084", "right": "issue:45507", "accept": false, - "reason": "Template compilation TypeError and GraniteMoEHybrid invalid method call are separate issues." + "reason": "Template compilation error and GraniteMoEHybrid invalid method calls are unrelated." }, { "left": "issue:43262", "right": "issue:43377", "accept": false, - "reason": "Audio chat-template sampling-rate default and MIMI batched-vs-single padding-mask mismatch are different audio bugs." + "reason": "Audio chat-template sampling-rate default and MIMI batched-vs-single padding mask issue are different audio bugs." }, { "left": "issue:43089", "right": "issue:45987", "accept": false, - "reason": "Generation GPU-sync overhead and StaticCache seq-length return type are not the same defect." + "reason": "Generation performance overhead and StaticCache seq-length return type are unrelated." } ] }, @@ -2937,13 +2977,13 @@ "component_index": 1, "component_count": 17, "review_unit_index": 9, - "review_unit_count": 49, + "review_unit_count": 50, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-427", + "cluster_id": "cluster-10000-428", "nodes": [ "issue:33666", "issue:36032", @@ -2999,8 +3039,8 @@ "issue:43646|issue:44534", "issue:43322|issue:43540", "issue:43404|issue:44466", - "issue:43994|issue:44610", "issue:44403|issue:45081", + "issue:43994|issue:44610", "issue:42915|issue:45084", "issue:43377|issue:43819", "issue:44265|issue:44610", @@ -3015,253 +3055,256 @@ "issue:44479|issue:45325", "issue:42915|issue:44589", "issue:43404|issue:45127", - "issue:40444|issue:45325", "issue:43296|issue:43531", + "issue:40444|issue:45325", "issue:36032|issue:41720", "issue:43611|issue:44534", "issue:43531|issue:44877", "issue:43425|issue:44403", + "issue:41553|issue:43582", "issue:40990|issue:44811", "issue:43632|issue:44534", - "issue:41553|issue:43582", "issue:44079|issue:45663", - "issue:33666|issue:40444", "issue:43792|issue:44589", + "issue:33666|issue:40444", "issue:44683|issue:45020" ], - "prepared_review_unit_hash": "5cbb1dff378c8a9ed8f7a39c263b785c314bc387dab318e3fc4acc1045900ed3", + "prepared_review_unit_hash": "1fe41e250d77384c381f5151ded2d51d50763e5ba279e284180c8c212da783ce", "component_budget": { - "node_count": 427, - "item_count": 427, - "soft_pair_count": 1576, - "serialized_chars": 816182, - "estimated_input_tokens": 204046, - "estimated_eval_tokens": 408348 + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 33, - "serialized_chars": 37471, - "estimated_input_tokens": 9368, - "estimated_eval_tokens": 18992 + "serialized_chars": 37467, + "estimated_input_tokens": 9367, + "estimated_eval_tokens": 18990 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The cluster is mostly a loose collection of model-specific regressions and a few broad Transformers v5 compatibility reports. The soft edges are all superficial similarities rather than the same underlying bug, so none should be merged as duplicates. No pull requests are present.", - "confidence": 0.91, - "canonical_issue_reason": "Issue 45020 is the best umbrella candidate: it is the broadest report in the set and explicitly covers recent-version loading breakages involving remote_code, which makes it the closest canonical anchor among these heterogeneous issues.", + "summary": "This set is mostly heterogeneous regression reports across different models and subsystems. I only see one plausible duplicate pair around Qwen3.5 save_pretrained checkpoint corruption; the rest are too far apart to safely merge.", + "confidence": 0.72, + "canonical_issue_reason": "issue:45216 is the broader, earlier report about Qwen3.5 save_pretrained producing an incorrect checkpoint; issue:45357 reads like a narrower follow-up focused on the visual encoder keys, so 45216 is the better canonical issue for that pair.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 45020 is the most representative global issue because it is broad enough to subsume related v5/regression reports, even though many of the others are distinct model-specific failures.", + "best_issue_reason": "issue:45216 is the strongest single issue to anchor the only likely duplicate cluster because it states the regression at the broader save_pretrained level and has the clearest canonical scope.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43645", "right": "issue:44534", "accept": false, - "reason": "Different failure modes: notebook/custom-model initialization vs non-persistent buffer corruption." + "reason": "Both mention Transformers v5 regressions, but one is about custom model init in notebooks and the other about non-persistent buffers being serialized incorrectly; different concrete bugs." }, { "left": "issue:43646", "right": "issue:44534", "accept": false, - "reason": "Both are v5 regressions, but one is custom model init and the other is buffer state handling; not the same bug." + "reason": "Custom model initialization failure and non-persistent buffer junk are separate v5 issues with different failure modes and code paths." }, { "left": "issue:43322", "right": "issue:43540", "accept": false, - "reason": "Different model families and symptoms: Llava Next segfault vs Qwen3OmniMoe video-processing ValueError." + "reason": "Both are multimodal loading/processing failures, but Llava-Next segfault and Qwen3OmniMoe video ValueError are not the same underlying bug." }, { "left": "issue:43404", "right": "issue:44466", "accept": false, - "reason": "Different scope: Mistral3 tied lm_head weights vs generic v5 save_pretrained serialization inconsistency." + "reason": "Both involve tied lm_head weights, but one is a model-specific tie failure and the other is a broader save/serialization inconsistency; not the same concrete bug." }, { - "left": "issue:43994", - "right": "issue:44610", + "left": "issue:44403", + "right": "issue:45081", "accept": false, - "reason": "Unrelated models and APIs: siglip2 inference/pipeline output quality vs OmDet-Turbo processor image-size mismatch." + "reason": "Unnecessary logging noise and a Mistral tokenizer crash are unrelated." }, { - "left": "issue:44403", - "right": "issue:45081", + "left": "issue:43994", + "right": "issue:44610", "accept": false, - "reason": "Generic loading noise vs a specific Mistral tokenizer regex crash; not the same code path." + "reason": "Different models and different defects: bad SigLIP2 outputs versus an OmDet-Turbo processor size mismatch." }, { "left": "issue:42915", "right": "issue:45084", "accept": false, - "reason": "Qwen3MoE FP8 failure and non-template-node compile error are different subsystems." + "reason": "Qwen3Moe FP8 config failure and a template-node compilation TypeError are unrelated." }, { "left": "issue:43377", "right": "issue:43819", "accept": false, - "reason": "Both involve audio/audio-codec areas, but one is padding-mask batching and the other is DAC STE mismatch; distinct bugs." + "reason": "Both are audio-model correctness bugs, but MIMI padding-mask behavior and DAC missing-STE behavior are different code paths and fixes." }, { "left": "issue:44265", "right": "issue:44610", "accept": false, - "reason": "Different features: torch.export/torch_compilable_check vs image preprocessing size expectations." + "reason": "torch.export failures from torch_compilable_check are unrelated to the OmDet-Turbo image-size regression." }, { "left": "issue:44451", "right": "issue:45081", "accept": false, - "reason": "Different model and stack: ScandiBERT load failure vs Mistral tokenizer backend tokenization crash." + "reason": "Loading ScandiBERT and the Mistral regex patch crash are separate issues." }, { "left": "issue:44987", "right": "issue:45081", "accept": false, - "reason": "Different regression classes: model loading on transformers>=5.1 vs tokenizer regex patching for Mistral." + "reason": "A failing remote-code model load and a tokenizer regex patch crash are not the same bug." }, { "left": "issue:44220", "right": "issue:44683", "accept": false, - "reason": "Audio feature extraction bug is unrelated to compiled flex_attention on newer torch versions." + "reason": "Audio fbank extraction and compiled flex_attention on new torch versions are unrelated." }, { "left": "issue:42915", "right": "issue:44291", "accept": false, - "reason": "Different root causes: Qwen3MoE FP8 config failure vs init_empty_weights/_is_hf_initialized argument handling." + "reason": "Different failures: FP8 config support for Qwen3Moe versus init_empty_weights/_is_hf_initialized argument handling." }, { "left": "issue:45216", "right": "issue:45357", - "accept": false, - "reason": "Related model family, but different regression points: save_pretrained checkpoint correctness vs visual encoder key serialization." + "accept": true, + "reason": "Both describe the same Qwen3.5 save_pretrained regression; 45357 looks like a narrower symptom of incorrect visual-encoder key serialization within the broader checkpoint corruption reported in 45216." }, { "left": "issue:43632", "right": "issue:44756", "accept": false, - "reason": "_is_hf_initialized breakage is unrelated to mmap/Strix Halo OOM behavior." + "reason": "_is_hf_initialized breakage and Strix Halo mmap OOM are unrelated." }, { "left": "issue:44062", "right": "issue:44987", "accept": false, - "reason": "Tokenizer AddedToken special-argument conflict is a different bug from loading physical-intelligence/fast." + "reason": "AddedToken special-argument duplication and a model load failure for physical-intelligence/fast are different tokenizer/model issues." }, { "left": "issue:36032", "right": "issue:43742", "accept": false, - "reason": "T5 tokenizer constructor conflict and MobileLLM key error are unrelated loading issues." + "reason": "T5 tokenizer special-token conflict and MobileLLM key error are unrelated loading failures." }, { "left": "issue:44479", "right": "issue:45325", "accept": false, - "reason": "Both are Qwen VL regressions, but one is general video-input breakage and the other is rope-index scaling for still images." + "reason": "Both are Qwen VL-family regressions, but one is a broad video-input regression and the other is a specific still-image rope-index scaling bug; not enough to treat as one issue." }, { "left": "issue:42915", "right": "issue:44589", "accept": false, - "reason": "Different failure domains: Qwen3MoE FP8 config vs missing Float8 storage type." + "reason": "FP8-config failure for Qwen3Moe and missing Float8 storage support are different failures." }, { "left": "issue:43404", "right": "issue:45127", "accept": false, - "reason": "Tied lm_head weights in Mistral3 is not the same as LoRA merge collapse with extended vocabulary." + "reason": "Tie-weight serialization in Mistral3 and LoRA-merge collapse with extended vocab are separate tied-weights problems with different triggers." }, { - "left": "issue:40444", - "right": "issue:45325", + "left": "issue:43296", + "right": "issue:43531", "accept": false, - "reason": "IterableDataset multi-image finetuning failure is separate from Qwen2.5-VL still-image temporal position-id scaling." + "reason": "PaddleOCR-VL load failure and Qwen3-MoE sliding_window behavior are unrelated." }, { - "left": "issue:43296", - "right": "issue:43531", + "left": "issue:40444", + "right": "issue:45325", "accept": false, - "reason": "PaddleOCR-VL load failure in vLLM is unrelated to Qwen3-MoE sliding_window behavior." + "reason": "Qwen2.5-VL finetuning with multiple images per prompt and still-image temporal scaling are not the same bug." }, { "left": "issue:36032", "right": "issue:41720", "accept": false, - "reason": "Tokenizer load conflict and Qwen3 auto device mapping cudaErrorAssert are different problems." + "reason": "Tokenizer load conflict and Qwen3 auto device-mapping cudaErrorAssert are unrelated." }, { "left": "issue:43611", "right": "issue:44534", "accept": false, - "reason": "base_model_prefix loading breakage and non-persistent buffer junk are different v5 regressions." + "reason": "base_model_prefix loading breakage and non-persistent buffer junk are distinct v5 regressions." }, { "left": "issue:43531", "right": "issue:44877", "accept": false, - "reason": "Sliding_window handling in Qwen3-MoE is unrelated to strict config blocking granite_speech loading." + "reason": "Qwen3-MoE sliding_window handling and granite_speech config strictness are unrelated." }, { "left": "issue:43425", "right": "issue:44403", "accept": false, - "reason": "Torch version compatibility issue is not the same as generic transformer loading noise." + "reason": "Torch 2.10 incompatibility and generic transformer-loading noise are unrelated." + }, + { + "left": "issue:41553", + "right": "issue:43582", + "accept": false, + "reason": "Voxtral tokenizer error messaging and AppleSilicon warmup TypeError are unrelated." }, { "left": "issue:40990", "right": "issue:44811", "accept": false, - "reason": "High perplexity on GPT-OSS is unrelated to Whisper batch_decode skip_special_tokens handling." + "reason": "High perplexity on gpt-oss-20b and Whisper batch_decode skip_special_tokens handling are unrelated." }, { "left": "issue:43632", "right": "issue:44534", "accept": false, - "reason": "They are both v5-related, but one is _is_hf_initialized and the other is buffer initialization; no shared code path." - }, - { - "left": "issue:41553", - "right": "issue:43582", - "accept": false, - "reason": "Bad AutoTokenizer error for Voxtral and AppleSilicon TypeError in caching_allocator_warmup are unrelated." + "reason": "_is_hf_initialized regression and junk-filled non-persistent buffers are different v5 issues." }, { "left": "issue:44079", "right": "issue:45663", "accept": false, - "reason": "ModelOutput key assignment and Gemma4 FSDP2 shared_kv_states KeyError are different mechanisms." + "reason": "ModelOutput key assignment and Gemma-4 FSDP2 shared_kv_states behavior are unrelated." }, { - "left": "issue:33666", - "right": "issue:40444", + "left": "issue:43792", + "right": "issue:44589", "accept": false, - "reason": "Qwen2-VL multi-GPU training and Qwen2.5-VL iterable multi-image finetuning are different training bugs." + "reason": "Whisper load failure and missing Float8 storage object are unrelated." }, { - "left": "issue:43792", - "right": "issue:44589", + "left": "issue:33666", + "right": "issue:40444", "accept": false, - "reason": "Whisper runtime failure and missing Float8 storage object are unrelated." + "reason": "Both are Qwen VL training reports, but multi-GPU training and iterable-dataset multi-image finetuning are different problems." }, { "left": "issue:44683", "right": "issue:45020", "accept": false, - "reason": "Compiled flex_attention on torch>=2.9 is a torch backend compatibility issue, not the remote_code loading regressions tracked by 45020." + "reason": "Compiled flex_attention on newer torch and remote_code breakage are unrelated." } ] }, - "evaluator_result": null, - "evaluator_used": false, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet and stays conservative. The single accepted duplicate pair (issue:45216 / issue:45357) is well-supported by both titles, and the rest of the soft-edge rejections avoid overclaiming sameness across clearly different bugs and subsystems." + }, + "evaluator_used": true, "retried": false, - "accepted_nontrivial_soft_edge": false, + "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null }, @@ -3270,13 +3313,13 @@ "component_index": 1, "component_count": 17, "review_unit_index": 10, - "review_unit_count": 49, + "review_unit_count": 50, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-427", + "cluster_id": "cluster-10000-428", "nodes": [ "issue:35707", "issue:40444", @@ -3338,238 +3381,238 @@ "issue:44829|issue:45479", "issue:43482|issue:45538", "issue:44291|issue:44589", - "issue:44079|issue:44534", "issue:43482|issue:43525", + "issue:44079|issue:44534", "issue:43818|issue:45357", "issue:43377|issue:45491", - "issue:40444|issue:45381", "issue:41720|issue:45081", "issue:43531|issue:44291", - "issue:43749|issue:44079", + "issue:40444|issue:45381", "issue:44265|issue:45072", "issue:43644|issue:44079", + "issue:43749|issue:44079", "issue:45325|issue:45381", - "issue:43632|issue:45072", "issue:42175|issue:43296", + "issue:43632|issue:45072", "issue:43611|issue:43646", + "issue:43792|issue:43957", "issue:43994|issue:44898", "issue:45325|issue:45834", - "issue:43792|issue:43957", "issue:43454|issue:44466", "issue:42915|issue:43482" ], - "prepared_review_unit_hash": "f7a95ba0ac21acc6ff840648984d710722ceeed994c82360eefae0251c23d0b7", + "prepared_review_unit_hash": "712e408dc75365ed2629d8556d8ff1e2010840ab33c68351df634787951ae548", "component_budget": { - "node_count": 427, - "item_count": 427, - "soft_pair_count": 1576, - "serialized_chars": 816182, - "estimated_input_tokens": 204046, - "estimated_eval_tokens": 408348 + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 29, - "serialized_chars": 36689, - "estimated_input_tokens": 9173, - "estimated_eval_tokens": 18602 + "serialized_chars": 36687, + "estimated_input_tokens": 9172, + "estimated_eval_tokens": 18600 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The cluster still centers on Transformers v5 loading/init regressions, with issue 43611 the best anchor, but the soft-edge evidence is too thin to merge any of the candidate pairs as duplicates.", - "confidence": 0.69, - "canonical_issue_reason": "Issue 43611 is the clearest representative of the cluster: a concrete Transformers v5 model-loading/init regression with broad impact across related failures.", + "summary": "Mostly a grab bag of unrelated Transformers issues. The only plausible duplicate pair is the v5 custom model initialization/loading regression around `base_model_prefix`; the rest differ by model, subsystem, or failure mode. No PRs are present.", + "confidence": 0.43, + "canonical_issue_reason": "issue:43646 is the best canonical issue: it is the broadest and most central v5 model-initialization regression, and it can naturally subsume the more specific `base_model_prefix` loading failure in issue:43611.", "canonical_pr_reason": null, - "best_issue_reason": "43611 is the best single issue anchor because it is broad, concrete, and captures the main v5 loading/init regression theme better than the narrower model-specific reports.", + "best_issue_reason": "issue:43646 has the widest scope among the candidates and best represents the underlying v5 custom-model init/load breakage.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43421", "right": "issue:44568", "accept": false, - "reason": "Both are tokenizer-related, but one is runtime special-token/post-processor updates and the other is a specific tokenizer add_special_tokens regression; not the same concrete bug." + "reason": "Both are tokenizer-related, but one is about runtime post-processor updates and the other about `add_special_tokens`/BOS-EOS behavior for a specific tokenizer; not the same concrete bug." }, { "left": "issue:44403", "right": "issue:44991", "accept": false, - "reason": "Loading-noise/reporting behavior vs a tokenizer-loading failure for a specific model repo; unrelated code paths." + "reason": "One is generic loading noise, the other is a tokenizer load failure for a specific model; different symptoms and code paths." }, { "left": "issue:35707", "right": "issue:40444", "accept": false, - "reason": "Progressive generation with inputs_embeds/past_key_values is different from VL fine-tuning with multiple images per prompt." + "reason": "Different tasks entirely: generation with `inputs_embeds/past_key_values` versus Qwen2.5-VL finetuning with multiple images." }, { "left": "issue:43381", "right": "issue:45923", "accept": false, - "reason": "Eval-mode gradient checkpointing behavior is separate from a missing supports_gradient_checkpointing flag on a trust_remote_code model." + "reason": "Gradient-checkpointing-in-eval and missing `supports_gradient_checkpointing` on a trust-remote-code model are related only broadly, not the same bug." }, { "left": "issue:44462", "right": "issue:45812", "accept": false, - "reason": "Both mention tokenizer handling, but one is ignoring tokenizer.json and the other is a Granite token-ID regression; too broad to call identical." + "reason": "Both involve AutoTokenizer, but one ignores `tokenizer.json` while the other returns wrong token IDs for Granite models; different failures." }, { "left": "issue:42175", "right": "issue:45542", "accept": false, - "reason": "Packaging/backend installation issues are different from a runtime undefined tf backend error." + "reason": "Both mention TensorFlow backend packaging, but one is an install-extra/dependency issue and the other is a runtime undefined-backend error with tensorboard; not clearly the same fix." }, { "left": "issue:43317", "right": "issue:43856", "accept": false, - "reason": "Device-map/offload loading failure and Qwen3 MoE training memory usage are different problems." + "reason": "One is device-map/offload loading failure; the other is MoE training memory inefficiency. Different problem class." }, { "left": "issue:44829", "right": "issue:45479", "accept": false, - "reason": "Flash-attention training degeneration and num_labels=1 zero-loss classification are not the same bug." + "reason": "Different sequence-classification bugs: flash-attention-3 training degeneration versus incorrect loss for `num_labels=1`." }, { "left": "issue:43482", "right": "issue:45538", "accept": false, - "reason": "GGUF loading failure and CLIPTokenizer model_max_length behavior are unrelated." + "reason": "Qwen2.5-GGUF loading on v5 versus CLIPTokenizer `model_max_length` constant; unrelated." }, { "left": "issue:44291", "right": "issue:44589", "accept": false, - "reason": "Unexpected _is_hf_initialized argument handling and missing Float8 storage lookup are different failure modes." + "reason": "Both are v5 initialization-related, but one is an unexpected `_is_hf_initialized` argument and the other is a missing Float8 storage type; different failures." }, { - "left": "issue:44079", - "right": "issue:44534", + "left": "issue:43482", + "right": "issue:43525", "accept": false, - "reason": "ModelOutput key assignment and non-persistent buffer initialization are different code paths." + "reason": "Both are model-loading issues, but model-specific config/tokenization problems with different root causes." }, { - "left": "issue:43482", - "right": "issue:43525", + "left": "issue:44079", + "right": "issue:44534", "accept": false, - "reason": "GGUF loading and missing pad_token_id on Llama4Config are unrelated." + "reason": "ModelOutput key assignment and non-persistent buffer junk are unrelated runtime/state bugs." }, { "left": "issue:43818", "right": "issue:45357", "accept": false, - "reason": "Video-LLaVA tower weight/temporal-attention issues are unrelated to Qwen3.5 visual encoder key serialization." + "reason": "Video-LLaVA tower/weight-sharing bug versus Qwen3.5 visual encoder save-pretrained regression; different models and code paths." }, { "left": "issue:43377", "right": "issue:45491", "accept": false, - "reason": "Batch-vs-single padding-mask mismatch is different from Gemma3 NaNs caused by all-padding sliding windows." - }, - { - "left": "issue:40444", - "right": "issue:45381", - "accept": false, - "reason": "Multi-image prompt finetuning failure is unrelated to Qwen2.5-VL video vision_position_ids." + "reason": "Both are batching/padding issues, but they affect different models and different mechanisms." }, { "left": "issue:41720", "right": "issue:45081", "accept": false, - "reason": "Qwen3 auto device mapping assert and Mistral tokenizer regex crash are different subsystems." + "reason": "CUDA assert during Qwen3 auto device mapping is unrelated to the Mistral tokenizer regex patch crash." }, { "left": "issue:43531", "right": "issue:44291", "accept": false, - "reason": "Qwen3-MoE sliding_window behavior is not the same as the meta-init _is_hf_initialized argument issue." + "reason": "Sliding-window handling in Qwen3-MoE is a different bug from `_is_hf_initialized`/`init_empty_weights` initialization issues." }, { - "left": "issue:43749", - "right": "issue:44079", + "left": "issue:40444", + "right": "issue:45381", "accept": false, - "reason": "FSDP RAM-efficient loading and ModelOutput key bookkeeping are unrelated." + "reason": "Multimodal finetuning with multiple images is unrelated to Qwen2.5-VL vision position IDs for video input." }, { "left": "issue:44265", "right": "issue:45072", "accept": false, - "reason": "torch.export/check compatibility and bfloat16 dtype mismatches are different problems." + "reason": "`torch.export`/`torch_compilable_check` failure and dtype mismatch in inference are unrelated." }, { "left": "issue:43644", "right": "issue:44079", "accept": false, - "reason": "Non-persistent buffer junk fill is separate from ModelOutput key assignment." + "reason": "Non-persistent buffer initialization bug versus ModelOutput key bookkeeping bug; different components." }, { - "left": "issue:45325", - "right": "issue:45381", + "left": "issue:43749", + "right": "issue:44079", "accept": false, - "reason": "Both are Qwen2.5-VL position-id bugs, but one is still-image temporal scaling and the other is video input position IDs; the packet doesn't show a single concrete shared defect." + "reason": "FSDP CPU RAM efficient loading failure is unrelated to ModelOutput key handling." }, { - "left": "issue:43632", - "right": "issue:45072", + "left": "issue:45325", + "right": "issue:45381", "accept": false, - "reason": "_is_hf_initialized breakage and dtype mismatch inference are different v5 regressions." + "reason": "Both are Qwen2.5-VL positional-id bugs, but one is still-image temporal scaling and the other is video-input vision positions; not enough to treat as the same concrete bug." }, { "left": "issue:42175", "right": "issue:43296", "accept": false, - "reason": "Missing TensorFlow backend after pip install and PaddleOCR-VL load failure are distinct issues." + "reason": "TensorFlow backend packaging versus loading PaddleOCR-VL in vLLM 0.11.0; unrelated." + }, + { + "left": "issue:43632", + "right": "issue:45072", + "accept": false, + "reason": "`_is_hf_initialized` flag breakage is unrelated to SwitchTransformers/TimmWrapperModel bfloat16 dtype mismatches." }, { "left": "issue:43611", "right": "issue:43646", + "accept": true, + "reason": "Both point to the same Transformers v5 custom-model initialization/loading regression; `base_model_prefix` loading is a specific manifestation of the broader custom init breakage." + }, + { + "left": "issue:43792", + "right": "issue:43957", "accept": false, - "reason": "Both are broad v5 loading/init regressions, but the titles point to different failure surfaces rather than one merged bug." + "reason": "Whisper run failure and meta-device loading failure are different model-loading problems." }, { "left": "issue:43994", "right": "issue:44898", "accept": false, - "reason": "Different vision-model failures with no clear shared code path." + "reason": "Different model families and different failure modes: SigLIP nonsensical outputs versus Perceiver resolution/interpolation failure." }, { "left": "issue:45325", "right": "issue:45834", "accept": false, - "reason": "Same model family, but position-id scaling and long OCR index error are different bugs." - }, - { - "left": "issue:43792", - "right": "issue:43957", - "accept": false, - "reason": "Meta-device loading regressions are related thematically, but init_empty_weights/_is_hf_initialized and torch.device(\"meta\") breakage are not shown to be the same concrete bug." + "reason": "Both are vision-language issues, but one is a positional-index regression and the other is long OCR input index errors; not the same bug." }, { "left": "issue:43454", "right": "issue:44466", "accept": false, - "reason": "Missing lm_head tying and device-dependent serialization of tied weights are different tied-weight failures." + "reason": "AyaVision weight tying bug versus inconsistent lm_head serialization; related only by weights, not the same issue." }, { "left": "issue:42915", "right": "issue:43482", "accept": false, - "reason": "Qwen3Moe FP8 config failure and Qwen2.5-GGUF loading failure are unrelated." + "reason": "Qwen3MoE FP8 configuration failure and Qwen2.5-GGUF loading failure are unrelated." } ] }, "evaluator_result": { - "accept": false, - "feedback": "The overall cluster summary is grounded, but the soft-edge judgments are not conservative enough. In particular, the accepted pairs (43421\u219444568, 44462\u219445812, 43792\u219443957) are supported only at a broad thematic level from the titles; the packet does not justify treating them as the same concrete bug. The other verdicts are mostly careful, but because a few accept=true calls overstate evidence, I\u2019d reject this review." + "accept": true, + "feedback": "The summary is grounded in the packet and stays conservative overall. The only accepted soft pair, issue:43611 \u2194 issue:43646, is a plausible same v5 init/load regression by title, and the rest are reasonably rejected as different models, subsystems, or failure modes. No PRs are present." }, "evaluator_used": true, - "retried": true, - "accepted_nontrivial_soft_edge": false, + "retried": false, + "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null }, @@ -3578,13 +3621,13 @@ "component_index": 1, "component_count": 17, "review_unit_index": 11, - "review_unit_count": 49, + "review_unit_count": 50, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-427", + "cluster_id": "cluster-10000-428", "nodes": [ "issue:34634", "issue:35707", @@ -3635,13 +3678,13 @@ "issue:46050" ], "soft_pairs": [ - "issue:44387|issue:44617", "issue:43296|issue:45800", + "issue:44387|issue:44617", "issue:43701|issue:45325", "issue:44479|issue:45823", "issue:43881|issue:44488", - "issue:43606|issue:44265", "issue:44291|issue:44877", + "issue:43606|issue:44265", "issue:43262|issue:43746", "issue:43873|issue:45701", "issue:34634|issue:35707", @@ -3654,32 +3697,32 @@ "issue:43454|issue:43550", "issue:39692|issue:40444", "issue:43377|issue:44610", - "issue:43454|issue:45072", "issue:43317|issue:43873", + "issue:43454|issue:45072", "issue:43572|issue:45759", "issue:43257|issue:45923", "issue:43377|issue:44898", "issue:40990|issue:42947", "issue:43322|issue:46050", "issue:43526|issue:43883", - "issue:43454|issue:43720", "issue:41720|issue:43322", - "issue:43881|issue:44991" + "issue:43881|issue:44991", + "issue:43454|issue:43720" ], - "prepared_review_unit_hash": "8fe5a8e942d7ba43a26f53d8e32550bc9245a3ae19285c413bc4605e9285b4b7", + "prepared_review_unit_hash": "b52f508eec2c595d1794fd24b7abd2360fee7b5303182f014f9f8fbd0fedf037", "component_budget": { - "node_count": 427, - "item_count": 427, - "soft_pair_count": 1576, - "serialized_chars": 816182, - "estimated_input_tokens": 204046, - "estimated_eval_tokens": 408348 + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 }, "budget": { "node_count": 47, "item_count": 47, "soft_pair_count": 30, - "serialized_chars": 36267, + "serialized_chars": 36268, "estimated_input_tokens": 9067, "estimated_eval_tokens": 18390 }, @@ -3688,192 +3731,192 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This set is mostly a heterogeneous grab bag of unrelated bugs across loading, quantization, tokenization, multimodal processors, and device placement. I don\u2019t see any pair that clearly describes the same underlying fix, so none of the soft edges should be merged.", - "confidence": 0.96, - "canonical_issue_reason": "Issue 43873 is the broadest and most referenced report, so it is the closest thing to a cluster hub; however, the overall set is too diverse to treat it as a true canonical duplicate anchor.", + "summary": "The cluster is mostly a set of unrelated model-loading, tokenizer, quantization, and device-map bugs. None of the soft pairs look like true duplicates under a strict triage standard; the closest themes are offloading/quantization and tokenizer regressions, but they still appear to be distinct concrete failures. I\u2019d use the broad offloading/quantization report as the cluster representative.", + "confidence": 0.38, + "canonical_issue_reason": "issue:43873 is the broadest and most central issue in this set: it directly covers offloading behavior with quantization, has the most discussion/inbound references, and best represents the nearby offload/device-map/quantization failures without being overly model-specific.", "canonical_pr_reason": null, - "best_issue_reason": "43873 is the best representative by scope and inbound references, but it should be viewed as a broad related report rather than a real duplicate target.", + "best_issue_reason": "issue:43873 is the best single issue to anchor the cluster because it is open, high-signal, and spans the main recurring theme across the nearby reports: quantized/offloaded model loading and device placement.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:44387", - "right": "issue:44617", + "left": "issue:43296", + "right": "issue:45800", "accept": false, - "reason": "Both mention OOM, but one is int4 quantization memory growth and the other is a specific video-model CUDA OOM; different failure modes." + "reason": "Different failures: a specific PaddleOCR-VL load issue vs a broad torch/transformers version incompatibility." }, { - "left": "issue:43296", - "right": "issue:45800", + "left": "issue:44387", + "right": "issue:44617", "accept": false, - "reason": "PaddleOCR-VL loading in vLLM and a torch/transformers compatibility issue are different problems with different roots." + "reason": "Both mention OOM, but the models and root causes are unrelated; one is int4 quantization memory growth, the other is Sam3Video CUDA OOM." }, { "left": "issue:43701", "right": "issue:45325", "accept": false, - "reason": "Checkpoint key mismatch is unrelated to the Qwen2.5-VL rope index temporal-position bug." + "reason": "Checkpoint key mismatch is unrelated to the Qwen2.5-VL rope-index scaling bug." }, { "left": "issue:44479", "right": "issue:45823", "accept": false, - "reason": "A video-input regression in Qwen-family models is not the same as a Gemma4 PLE device-mismatch forward pass issue." + "reason": "Distinct model families and symptoms: a video-input regression in Qwen variants vs a Gemma4 device-mismatch forward bug." }, { "left": "issue:43881", "right": "issue:44488", "accept": false, - "reason": "Both are loading failures, but they involve different models and no shared concrete code path is evident." + "reason": "Both are load failures, but they affect different models and likely different code paths." }, { - "left": "issue:43606", - "right": "issue:44265", + "left": "issue:44291", + "right": "issue:44877", "accept": false, - "reason": "CPU offload device mismatch for Bark-small is unrelated to torch.export failures involving torch_compilable_check." + "reason": "One is an init-empty-weights argument regression; the other is a strict config loading issue. Same area, different bug." }, { - "left": "issue:44291", - "right": "issue:44877", + "left": "issue:43606", + "right": "issue:44265", "accept": false, - "reason": "The init_empty_weights argument error and the strict-config granite_speech loading problem are distinct loader/config issues." + "reason": "Bark CPU offload device mismatch and torch.export torch_compilable_check failure are unrelated." }, { "left": "issue:43262", "right": "issue:43746", "accept": false, - "reason": "Audio chat-template sample-rate handling is unrelated to PEFT adapter checkpoint loading." + "reason": "Audio chat-template sampling-rate handling is unrelated to PEFT adapter checkpoint loading." }, { "left": "issue:43873", "right": "issue:45701", "accept": false, - "reason": "Quantization/offloading behavior and tokenizer changes are different subsystems and different bugs." + "reason": "Quantization/offloading bug vs tokenizer behavior change; not the same concrete issue." }, { "left": "issue:34634", "right": "issue:35707", "accept": false, - "reason": "BarkProcessor voice_preset and progressive generation with inputs_embeds/past_key_values are unrelated." + "reason": "BarkProcessor voice_preset and progressive generation with past_key_values are different bugs." }, { "left": "issue:40990", "right": "issue:45584", "accept": false, - "reason": "High perplexity on GPT-OSS and Whisper empty-transcription generation failure are not the same bug." + "reason": "High perplexity on one model is unrelated to Whisper empty-transcription generation failure." }, { "left": "issue:44568", "right": "issue:45920", "accept": false, - "reason": "Both are tokenizer regressions, but one is a specific special-token omission and the other is a broader wrong-token-IDs report across many models." + "reason": "Both are tokenizer regressions, but they describe different failure modes and model sets; too broad to treat as one bug." }, { "left": "issue:35707", "right": "issue:41720", "accept": false, - "reason": "Progressive generation with past_key_values is unrelated to a Qwen3 auto device-map CUDA assert." + "reason": "Progressive generation with inputs_embeds/past_key_values is unrelated to Qwen3 auto device-map CUDA asserts." }, { "left": "issue:44186", "right": "issue:44265", "accept": false, - "reason": "Tokenizer crashes on NER/padding are unrelated to export failures with torch_compilable_check." + "reason": "Tokenizer crashes on NER/batched padding are unrelated to torch.export failures." }, { "left": "issue:43606", "right": "issue:43819", "accept": false, - "reason": "A device mismatch under CPU offload is not the same as DAC.from_latents diverging from forward due to missing STE." + "reason": "Both involve audio/codecs, but CPU offload device mismatch and DAC STE mismatch are different problems." }, { "left": "issue:41720", "right": "issue:45823", "accept": false, - "reason": "Both involve device_map=auto, but the affected models and concrete failures differ enough that they are not the same bug." + "reason": "Both are device-map related, but they fail in different models and likely different code paths; not a duplicate." }, { "left": "issue:43454", "right": "issue:43550", "accept": false, - "reason": "Weight-tying for AyaVision and torch.compile SDPA issues in Bamba are unrelated." + "reason": "Weight-tying in AyaVision is unrelated to Bamba SDPA torch.compile failure." }, { "left": "issue:39692", "right": "issue:40444", "accept": false, - "reason": "A SigLIP2 doc example mismatch/quantization issue is unrelated to Qwen2.5-VL multi-image dataset finetuning failures." + "reason": "Doc-example/model-processor mismatch in SigLIP2 is unrelated to Qwen2.5-VL multi-image IterableDataset finetuning failure." }, { "left": "issue:43377", "right": "issue:44610", "accept": false, - "reason": "MIMI encoder padding-mask behavior and OmDet-Turbo processor size mismatch are different code paths." + "reason": "Missing padding-mask support in MIMI encoder is unrelated to OmDet-Turbo image-size mismatch." }, { - "left": "issue:43454", - "right": "issue:45072", + "left": "issue:43317", + "right": "issue:43873", "accept": false, - "reason": "AyaVision lm_head tying and bfloat16 dtype mismatches in SwitchTransformers/TimmWrapperModel are not the same underlying issue." + "reason": "These are the closest pair, but one is a dequantized-model load failure on gpu+cpu offload while the other is a broader quantization offloading complaint; not enough evidence they are the same concrete bug." }, { - "left": "issue:43317", - "right": "issue:43873", + "left": "issue:43454", + "right": "issue:45072", "accept": false, - "reason": "They are both offload/quantization-adjacent, but one is dequantized model loading with device_map=auto and the other is a broader offloading regression report." + "reason": "Weight tying and bfloat16 dtype mismatch are different issues." }, { "left": "issue:43572", "right": "issue:45759", "accept": false, - "reason": "Missing pad_token_idx in StableLmConfig is unrelated to unwrapping text_config in composite Qwen multimodal configs." + "reason": "StableLm config pad-token handling and composite Qwen config unwrapping are different config-path bugs." }, { "left": "issue:43257", "right": "issue:45923", "accept": false, - "reason": "Qwen3 MoE weight conversion under accelerate+deepspeed and a missing gradient-checkpointing flag are different issues." + "reason": "MoE weight conversion under accelerate+deepspeed is unrelated to missing supports_gradient_checkpointing on a trust_remote_code variant." }, { "left": "issue:43377", "right": "issue:44898", "accept": false, - "reason": "A batched-vs-single padding-mask bug in MIMI has nothing to do with Perceiver non-default-resolution image classification." + "reason": "Padding-mask correctness in MIMI encoder is unrelated to Perceiver non-default-resolution failure." }, { "left": "issue:40990", "right": "issue:42947", "accept": false, - "reason": "Perplexity on WikiText-2 and ineffective gradient checkpointing with LoRA are unrelated." + "reason": "Perplexity anomaly and gradient-checkpointing ineffectiveness are unrelated." }, { "left": "issue:43322", "right": "issue:46050", "accept": false, - "reason": "Llava Next loading segfaults and sinq quantized save/load breakage are different failure classes." + "reason": "Llava Next load crash and sinq quantized save/load breakage are different bugs." }, { "left": "issue:43526", "right": "issue:43883", "accept": false, - "reason": "BeitImageProcessorFast label reduction and Molmo tied-weights attribute errors are unrelated." - }, - { - "left": "issue:43454", - "right": "issue:43720", - "accept": false, - "reason": "AyaVision weight tying and BitNet packed-weight unpacking during accelerate loading are different bugs." + "reason": "BeitImageProcessorFast label reduction and Molmo tied-weights attribute error are unrelated." }, { "left": "issue:41720", "right": "issue:43322", "accept": false, - "reason": "Both are multimodal/model-loading failures, but they affect different models and symptoms; too broad to merge." + "reason": "Both are model-loading/runtime crashes, but the affected models and failure signatures are different." }, { "left": "issue:43881", "right": "issue:44991", "accept": false, - "reason": "A glm-4v-9b load failure is unrelated to tokenizer loading for EMBEDDIA/est-roberta." + "reason": "Model loading failure for glm-4v-9b vs tokenizer loading failure for EMBEDDIA/est-roberta are unrelated." + }, + { + "left": "issue:43454", + "right": "issue:43720", + "accept": false, + "reason": "AyaVision weight tying and BitNet packed-weight unpacking are distinct loading/model-structure bugs." } ] }, @@ -3889,13 +3932,13 @@ "component_index": 1, "component_count": 17, "review_unit_index": 12, - "review_unit_count": 49, + "review_unit_count": 50, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-427", + "cluster_id": "cluster-10000-428", "nodes": [ "issue:36010", "issue:36331", @@ -3947,7 +3990,7 @@ ], "soft_pairs": [ "issue:43749|issue:44534", - "issue:43296|issue:44991", + "issue:43296|issue:44877", "issue:44466|issue:45357", "issue:41720|issue:44977", "issue:42617|issue:42915", @@ -3955,9 +3998,9 @@ "issue:44008|issue:45823", "issue:44387|issue:45704", "issue:44186|issue:45072", - "issue:43404|issue:45357", "issue:36331|issue:42617", "issue:43577|issue:43708", + "issue:43404|issue:45357", "issue:43873|issue:44756", "issue:36331|issue:42915", "issue:44877|issue:44991", @@ -3966,8 +4009,8 @@ "issue:44423|issue:45406", "issue:45632|issue:45684", "issue:41720|issue:42617", - "issue:44610|issue:44898", "issue:44479|issue:44918", + "issue:44610|issue:44898", "issue:43550|issue:43606", "issue:44186|issue:44610", "issue:42915|issue:43317", @@ -3975,223 +4018,217 @@ "issue:43526|issue:43881", "issue:41720|issue:45161", "issue:44671|issue:44743", - "issue:36010|issue:42617", - "issue:45081|issue:45812" + "issue:45081|issue:45812", + "issue:36010|issue:42617" ], - "prepared_review_unit_hash": "e8c260a028022adacac71ca8ebb87c49035e0a23492c19eaaf9b25bd22410f6f", + "prepared_review_unit_hash": "f10fb66cea39a10eb12a9e6be0a5bba2f9576ad9de5d3ad2f286b39fd81a89e8", "component_budget": { - "node_count": 427, - "item_count": 427, - "soft_pair_count": 1576, - "serialized_chars": 816182, - "estimated_input_tokens": 204046, - "estimated_eval_tokens": 408348 + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 }, "budget": { "node_count": 47, "item_count": 47, "soft_pair_count": 31, - "serialized_chars": 36659, - "estimated_input_tokens": 9165, - "estimated_eval_tokens": 18586 + "serialized_chars": 36671, + "estimated_input_tokens": 9168, + "estimated_eval_tokens": 18592 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "These items are mostly distinct Transformers bug reports that only loosely resemble each other by subsystem or symptom. I do not see a true duplicate cluster; the closest umbrella is the open offloading/quantization report, but the soft pairs are not mergeable as duplicates.", - "confidence": 0.87, - "canonical_issue_reason": "No exact duplicate center emerges; issue 43873 is the closest umbrella because it is open, active, and covers offloading/quantization behavior that overlaps the most with the nearby loading/offload reports.", - "canonical_pr_reason": null, - "best_issue_reason": "Issue 43873 is the best cluster anchor by scope and activity, but it is still not a duplicate of the others; the rest are mostly model-, tokenizer-, or API-specific regressions.", - "best_pr_reason": null, + "summary": "Mostly a loose theme cluster around Transformers loading/device/offload/tokenization regressions, but the soft pairs describe different concrete bugs or models rather than true duplicates. I would not merge any of the candidate pairs.", + "confidence": 0.82, + "canonical_issue_reason": "Issue #43873 is the broadest representative of the cluster\u2019s shared offloading/loading pain points, but it is still a distinct bug from the others.", + "canonical_pr_reason": "No pull requests are present in this cluster.", + "best_issue_reason": "#43873 is the best single issue anchor because it is the most general and central of the loading/offload-related reports here.", + "best_pr_reason": "No pull requests are present in this cluster.", "soft_edge_verdicts": [ { "left": "issue:43749", "right": "issue:44534", "accept": false, - "reason": "Both are runtime/loading regressions, but one is CPU-efficient loading broken and the other is non-persistent buffer initialization junk; different code paths and root causes." + "reason": "Both are memory/loading regressions, but one is FSDP CPU RAM efficient loading and the other is non-persistent buffer initialization; different failure modes and code paths." }, { "left": "issue:43296", - "right": "issue:44991", + "right": "issue:44877", "accept": false, - "reason": "Both involve loading failures, but one is a vLLM/PaddleOCR-VL model load issue and the other is a tokenizer-loading regression for est-roberta; not the same bug." + "reason": "One is a model-load failure in vLLM, the other is strict config blocking granite_speech; unrelated root causes." }, { "left": "issue:44466", "right": "issue:45357", "accept": false, - "reason": "Both touch save_pretrained serialization, but one is tied lm_head weight serialization and the other is incorrect visual-encoder keys; too different to merge." + "reason": "Both involve saving weights, but one is tied lm_head serialization by device and the other is incorrect visual encoder keys for Qwen3.5." }, { "left": "issue:41720", "right": "issue:44977", "accept": false, - "reason": "Different failures: auto device mapping on Qwen3 versus flash-attention generation on Qwen3.5; not the same concrete code path." + "reason": "Different Qwen3 family problems: CUDA assert under auto device mapping versus flash-attention generation failure." }, { "left": "issue:42617", "right": "issue:42915", "accept": false, - "reason": "A 3d_parallel script failure and a Qwen3Moe FineGrainedFP8Config failure are unrelated beyond both being training/runtime problems." + "reason": "Completely different contexts: 3d_parallel.py execution failure versus FineGrainedFP8Config failure on Qwen3Moe." }, { "left": "issue:40990", "right": "issue:45701", "accept": false, - "reason": "Perplexity on WikiText-2 and a tokenization-version change are related only broadly; the reports do not describe the same underlying defect." + "reason": "Perplexity on a specific model/dataset is not the same bug as tokenizer-version-dependent tokenization changes." }, { "left": "issue:44008", "right": "issue:45823", "accept": false, - "reason": "Different models and different bugs: Gemma3n variable-name collision versus Gemma4 PLE device mismatch." + "reason": "Gemma3n variable-name collision in forward is unrelated to Gemma4 PLE device mismatch." }, { "left": "issue:44387", "right": "issue:45704", "accept": false, - "reason": "Both are memory-related, but one is int4 quantization reserved-memory growth and the other is an apex RMSNorm leak; distinct causes." + "reason": "Both mention memory, but one is int4 quantization OOM in Transformers 5.x and the other is a T5/apex RMSNorm leak." }, { "left": "issue:44186", "right": "issue:45072", "accept": false, - "reason": "Tokenizer batching/truncation crashes and bfloat16 dtype mismatches are unrelated failures in different parts of the stack." - }, - { - "left": "issue:43404", - "right": "issue:45357", - "accept": false, - "reason": "One is an lm_head tying bug in Mistral3ForConditionalGeneration, the other is save_pretrained writing wrong visual-encoder keys; not the same issue." + "reason": "Tokenizer crashes on batching/padding are unrelated to dtype mismatches in SwitchTransformers/TimmWrapperModel inference." }, { "left": "issue:36331", "right": "issue:42617", "accept": false, - "reason": "CustomTrainer.compute_loss signature breakage and a 3d_parallel script failure are unrelated." + "reason": "Trainer compute_loss API break and 3d_parallel.py failure are different bugs with no shared code path." }, { "left": "issue:43577", "right": "issue:43708", "accept": false, - "reason": "Blip2 dtype loading mismatch and Trainer max_steps miscalculation on resume are separate bugs with no shared code path." + "reason": "Blip2 dtype loading is separate from Trainer resume_from_checkpoint step calculation." + }, + { + "left": "issue:43404", + "right": "issue:45357", + "accept": false, + "reason": "Both touch weights/serialization, but one is missing lm_head tying in Mistral3 and the other is incorrect visual encoder keys in Qwen3.5." }, { "left": "issue:43873", "right": "issue:44756", "accept": false, - "reason": "Both mention offloading/memory, but 43873 is quantization/offload behavior while 44756 is mmap-related OOM on Strix Halo; not the same defect." + "reason": "Both are offloading-related, but one is quantization/offload behavior and the other is mmap OOM on Strix Halo; not the same bug." }, { "left": "issue:36331", "right": "issue:42915", "accept": false, - "reason": "A trainer API regression and a Qwen3Moe FP8 loading/config bug are unrelated." + "reason": "A custom trainer signature regression is unrelated to a Qwen3Moe FP8 load failure." }, { "left": "issue:44877", "right": "issue:44991", "accept": false, - "reason": "Strict config loading for granite_speech and tokenizer loading for est-roberta are different failure modes." + "reason": "Strict config loading for granite_speech is unrelated to tokenizer loading failure for EMBEDDIA/est-roberta." }, { "left": "issue:43550", "right": "issue:43720", "accept": false, - "reason": "torch.compile with SDPA on Bamba is a different code path from packed-weight unpacking during accelerate loading for BitNet." + "reason": "Both are model-loading/runtime issues, but one is torch.compile SDPA on Bamba and the other is accelerate loading of BitNet packed weights." }, { "left": "issue:43450", "right": "issue:45701", "accept": false, - "reason": "Video batch-shape bugs and tokenization-version drift are unrelated." + "reason": "Incorrect video batch shapes are unrelated to tokenization changing across versions." }, { "left": "issue:44423", "right": "issue:45406", "accept": false, - "reason": "Both affect serve, but one is a continuous-batching multimodal crash from passing a string to .to(), and the other is Gemma4Processor missing _tokenizer; different bugs." + "reason": "Both affect serve paths, but one crashes on multimodal continuous batching and the other lacks Gemma4Processor._tokenizer." }, { "left": "issue:45632", "right": "issue:45684", "accept": false, - "reason": "A cache-path collision and permission propagation from save_pretrained are separate filesystem behaviors." + "reason": "One is a cache-path collision for local models, the other is permission propagation in save_pretrained; distinct filesystem bugs." }, { "left": "issue:41720", - "right": "issue:42617", + "right": "issue:45161", "accept": false, - "reason": "Qwen3 auto device mapping asserts and 3d_parallel execution failure are not the same underlying issue." + "reason": "Both concern Qwen/GPT-OSS parallelism, but one is an auto device-map CUDA assert and the other is TP not working for GPT-OSS MoE." }, { - "left": "issue:44610", - "right": "issue:44898", + "left": "issue:44479", + "right": "issue:44918", "accept": false, - "reason": "Both are processor/model input-size mismatches, but for different models and different preprocessing expectations; not mergeable as one bug." + "reason": "Qwen video-input regression and TRL SFT embedding unpacking failure are different code paths and symptoms." }, { - "left": "issue:44479", - "right": "issue:44918", + "left": "issue:44610", + "right": "issue:44898", "accept": false, - "reason": "A v5 video-input regression for Qwen VL models and Qwen3.5 embedding unpacking with TRL SFT trainer are distinct multimodal/training problems." + "reason": "Different vision preprocessing bugs: OmDet-Turbo produces the wrong resolution, while Perceiver fails on non-default resolution despite interpolate_pos_encoding." }, { "left": "issue:43550", "right": "issue:43606", "accept": false, - "reason": "Both involve model execution under special loading/compile settings, but SDPA+torch.compile and CPU-offload device mismatch are different code paths." + "reason": "Both involve device/compile behavior, but Bamba SDPA compile failure is not the same as suno/bark-small CPU-offload device mismatch." }, { "left": "issue:44186", "right": "issue:44610", "accept": false, - "reason": "A tokenizer crash on NER/batched padding and an image processor size mismatch are unrelated." + "reason": "Tokenizer batching/truncation crash and image processor resolution mismatch are unrelated." }, { "left": "issue:42915", "right": "issue:43317", "accept": false, - "reason": "Qwen3Moe FineGrainedFP8Config failure and dequantized model loading with gpu+cpu offload are different quantization/offload bugs." + "reason": "Qwen3Moe FineGrainedFP8 loading and dequantized-model CPU/GPU offload failure are distinct bugs." }, { "left": "issue:43295", "right": "issue:45701", "accept": false, - "reason": "A regression around processor.tokenizer access and a general tokenization-version change are not the same code-path bug." + "reason": "Processor.tokenizer regression is not the same as broader tokenization changes across versions." }, { "left": "issue:43526", "right": "issue:43881", "accept": false, - "reason": "BeitImageProcessorFast label reduction and glm-4v-9b loading failure are unrelated." - }, - { - "left": "issue:41720", - "right": "issue:45161", - "accept": false, - "reason": "Qwen3 auto device mapping on A800 and GPT-OSS MoE tensor-parallel issues are separate parallelization failures." + "reason": "A BeitImageProcessorFast label reduction bug is unrelated to glm-4v-9b loading failure." }, { "left": "issue:44671", "right": "issue:44743", "accept": false, - "reason": "CamemBERT masked-LM prediction regression and Qwen3.5 recurrent-state reset are unrelated model-specific bugs." + "reason": "CamemBERT masked LM prediction errors and Qwen3.5 recurrent-state reset are different model-specific issues." }, { - "left": "issue:36010", - "right": "issue:42617", + "left": "issue:45081", + "right": "issue:45812", "accept": false, - "reason": "An ImportError for GenerationMixin and a 3d_parallel execution failure have no substantive overlap." + "reason": "Both are tokenizer-related, but one is a Mistral regex patch crash and the other is wrong token IDs for Granite models; different root causes." }, { - "left": "issue:45081", - "right": "issue:45812", + "left": "issue:36010", + "right": "issue:42617", "accept": false, - "reason": "Both are tokenizer-related, but one is a Mistral regex patch crash and the other is wrong token IDs for Granite models; different failures and fixes." + "reason": "GenerationMixin import error is unrelated to running 3d_parallel.py." } ] }, @@ -4207,13 +4244,13 @@ "component_index": 1, "component_count": 17, "review_unit_index": 13, - "review_unit_count": 49, + "review_unit_count": 50, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-427", + "cluster_id": "cluster-10000-428", "nodes": [ "issue:29159", "issue:30064", @@ -4222,7 +4259,6 @@ "issue:34634", "issue:35707", "issue:36032", - "issue:36331", "issue:41720", "issue:42490", "issue:42915", @@ -4245,6 +4281,7 @@ "issue:44077", "issue:44186", "issue:44361", + "issue:44451", "issue:44492", "issue:44509", "issue:44514", @@ -4252,6 +4289,7 @@ "issue:44625", "issue:44898", "issue:44913", + "issue:44991", "issue:45059", "issue:45072", "issue:45381", @@ -4261,17 +4299,15 @@ "issue:45701", "issue:45704", "issue:45759", - "issue:45812", - "issue:45859" + "issue:45812" ], "soft_pairs": [ - "issue:43316|issue:44913", "issue:43531|issue:43572", + "issue:43316|issue:44913", "issue:43540|issue:43782", "issue:42915|issue:43322", "issue:43421|issue:45812", "issue:44186|issue:44361", - "issue:36331|issue:41720", "issue:43388|issue:45059", "issue:29159|issue:36032", "issue:45406|issue:45464", @@ -4285,177 +4321,171 @@ "issue:44492|issue:44509", "issue:43606|issue:43653", "issue:34634|issue:41720", - "issue:42490|issue:43329", "issue:43531|issue:45759", + "issue:42490|issue:43329", "issue:34567|issue:35707", "issue:44514|issue:44545", "issue:44625|issue:45479", "issue:43421|issue:43927", "issue:44898|issue:45072", "issue:43782|issue:45381", - "issue:44077|issue:44625", - "issue:45759|issue:45859" + "issue:44451|issue:44991", + "issue:44077|issue:44625" ], - "prepared_review_unit_hash": "fa7595952a31bd266194a79e4240a8e7812b3e457c7dc21899291b6ca804acd5", + "prepared_review_unit_hash": "83016af477d902585c2d2c6c832137cd9acbae67bda6a240d62bb8ca0653b457", "component_budget": { - "node_count": 427, - "item_count": 427, - "soft_pair_count": 1576, - "serialized_chars": 816182, - "estimated_input_tokens": 204046, - "estimated_eval_tokens": 408348 + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 30, - "serialized_chars": 36848, - "estimated_input_tokens": 9212, - "estimated_eval_tokens": 18680 + "soft_pair_count": 29, + "serialized_chars": 36509, + "estimated_input_tokens": 9128, + "estimated_eval_tokens": 18512 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly heterogeneous: many issues share broad themes like tokenization, model loading, or multimodal processors, but they are not the same underlying bug. The only clear duplicate-like pair is the Qwen2_5_VLProcessor batched padding=False crash.", - "confidence": 0.84, - "canonical_issue_reason": "Issue 44545 is the best canonical issue because it matches 44514 exactly on the same Qwen2_5_VLProcessor batched-input padding=False crash and has slightly stronger reference signal.", + "summary": "This cluster is mostly heterogeneous issue reports, not a single duplicate set. The only clear duplicate pair is the Qwen2_5_VLProcessor.apply_chat_template padding=False crash; the rest only share broad subsystem/model-family similarity.", + "confidence": 0.97, + "canonical_issue_reason": "No single issue cleanly represents the whole cluster; the only true duplicate subcluster is issue:44514/issue:44545, and issue:44545 is the best anchor because it has the stronger discussion/inbound reference signal for the same exact bug.", "canonical_pr_reason": null, - "best_issue_reason": "44545 is the strongest representative in the set: it is the only exact duplicate pair and appears to be the more canonical report of that bug.", + "best_issue_reason": "issue:44545 is the strongest issue candidate because it exactly matches issue:44514's bug report and has slightly more triage signal, making it the best dedupe target for that subcluster.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:43316", - "right": "issue:44913", + "left": "issue:43531", + "right": "issue:43572", "accept": false, - "reason": "Both are config-related, but one is Gemma3TextConfig API mismatch and the other is GPTNeoXConfig rotary_pct reload behavior; different bugs." + "reason": "Both are model/config generation bugs, but they affect different models and different parameters; not the same underlying defect." }, { - "left": "issue:43531", - "right": "issue:43572", + "left": "issue:43316", + "right": "issue:44913", "accept": false, - "reason": "Different models and fields: Qwen3-MoE sliding_window vs StableLmConfig pad_token_idx after v5 update." + "reason": "Both concern config behavior, but one is a Gemma3TextConfig API mismatch and the other is rotary_pct persistence on reload; different bug paths." }, { "left": "issue:43540", "right": "issue:43782", "accept": false, - "reason": "Both are Qwen multimodal failures, but one is video-input processing and the other is from_pretrained weight_only loading; not the same code path." + "reason": "Different multimodal Qwen bugs: one is video-input processing, the other is from_pretrained weight_only loading." }, { "left": "issue:42915", "right": "issue:43322", "accept": false, - "reason": "Qwen3Moe FP8 config failure and Llava Next segmentation fault are unrelated loading/runtime problems." + "reason": "Unrelated failures in different models and code paths; one is FP8 config loading, the other a Llava Next segfault." }, { "left": "issue:43421", "right": "issue:45812", "accept": false, - "reason": "Both involve tokenizers, but runtime post-processor updates and wrong Granite token IDs are distinct bugs." + "reason": "Both are tokenizer-related, but one is runtime post-processor refresh and the other is a silent token-id regression for Granite models." }, { "left": "issue:44186", "right": "issue:44361", "accept": false, - "reason": "Different tokenizer classes and failure modes: LayoutLMv2Tokenizer padding/truncation crash vs MLukeTokenizer AttributeError." - }, - { - "left": "issue:36331", - "right": "issue:41720", - "accept": false, - "reason": "Trainer compute_loss signature mismatch is unrelated to Qwen3 auto device mapping cudaErrorAssert." + "reason": "Both are tokenizer task crashes, but on different tokenizer implementations and different failure modes." }, { "left": "issue:43388", "right": "issue:45059", "accept": false, - "reason": "Metric gathering label truncation and SAM3 text/box behavior are unrelated." + "reason": "These are unrelated: metric gathering label truncation versus SAM3 text/bounding-box behavior." }, { "left": "issue:29159", "right": "issue:36032", "accept": false, - "reason": "General slow/fast tokenizer inconsistency is not the same bug as T5 add_special_tokens name conflict." + "reason": "Both mention tokenizers, but one is slow/fast inconsistency and the other is a T5 loading AttributeError; different bugs." }, { "left": "issue:45406", "right": "issue:45464", "accept": false, - "reason": "Gemma4Processor missing _tokenizer in serve is unrelated to Qwen3.5 streaming inference failures." + "reason": "Serving crashes in different places: Gemma4Processor missing _tokenizer versus Qwen3.5 streaming inference failure." }, { "left": "issue:33357", "right": "issue:35707", "accept": false, - "reason": "MacOS CLIP bus error and progressive generation with inputs_embeds/past_key_values are different issues." + "reason": "Different failures entirely: MacOS CLIP bus error versus generation with inputs_embeds/past_key_values." }, { "left": "issue:30064", "right": "issue:34634", "accept": false, - "reason": "Image processor void segmentation maps and BarkProcessor voice_preset are unrelated." + "reason": "Different processors and symptoms; image segmentation-map handling is unrelated to Bark voice_preset handling." }, { "left": "issue:41720", "right": "issue:43746", "accept": false, - "reason": "Qwen3 auto device mapping cudaErrorAssert and GraniteSpeech PEFT local checkpoint loading are different bugs." + "reason": "Different model/loading problems with no shared concrete code path." }, { "left": "issue:43606", "right": "issue:43720", "accept": false, - "reason": "CPU offload device mismatch for Bark-small is unrelated to BitNet packed-weight unpacking during accelerate loading." + "reason": "Both involve accelerate/CI loading, but one is CPU offload device mismatch and the other is packed-weight unpacking." }, { "left": "issue:43550", "right": "issue:43653", "accept": false, - "reason": "torch.compile/SDPA failure and tokenizer special-token registration are unrelated." + "reason": "Different subsystems: torch.compile/SDPA versus tokenizer special-token registration." }, { "left": "issue:43421", "right": "issue:45701", "accept": false, - "reason": "A runtime post-processor update feature gap is not the same as a broad tokenization regression across versions." + "reason": "Similar broad tokenizer theme, but 45701 is a generic version regression while 43421 is a runtime post-processor update bug." }, { "left": "issue:43873", "right": "issue:45704", "accept": false, - "reason": "Quantization/offloading behavior and T5 RMSNorm memory leak are unrelated." + "reason": "Unrelated issues: quantized offloading versus a T5 RMSNorm memory leak." }, { "left": "issue:44492", "right": "issue:44509", "accept": false, - "reason": "A cache-strategy typo and outdated docs about removed pipeline tasks are not duplicates." + "reason": "Both are documentation-ish, but they concern different content and different fixes." }, { "left": "issue:43606", "right": "issue:43653", "accept": false, - "reason": "Bark CPU offload mismatch and BigBirdTokenizer mask token registration are different bugs." + "reason": "Distinct bugs in different models/components; no shared underlying defect." }, { "left": "issue:34634", "right": "issue:41720", "accept": false, - "reason": "BarkProcessor voice_preset and Qwen3 auto device mapping failures do not share the same underlying cause." + "reason": "Different model/runtime failures with no common code path." }, { - "left": "issue:42490", - "right": "issue:43329", + "left": "issue:43531", + "right": "issue:45759", "accept": false, - "reason": "Local path trailing-slash import warning and multimodal token counting video-branch bug are unrelated." + "reason": "Both touch Qwen3 model config behavior, but sliding_window handling and text_config unwrapping are separate bugs." }, { - "left": "issue:43531", - "right": "issue:45759", + "left": "issue:42490", + "right": "issue:43329", "accept": false, - "reason": "Both touch Qwen3 configs, but sliding_window behavior and text_config unwrapping are different code paths." + "reason": "Completely different issues: trailing-slash local model import warning versus multimodal token counting in video branch." }, { "left": "issue:34567", @@ -4467,49 +4497,49 @@ "left": "issue:44514", "right": "issue:44545", "accept": true, - "reason": "Exact same Qwen2_5_VLProcessor batched-input padding=False crash; this is a clear duplicate pair." + "reason": "Exact same crash description and same code path in Qwen2_5_VLProcessor.apply_chat_template with batched input and padding=False." }, { "left": "issue:44625", "right": "issue:45479", "accept": false, - "reason": "Qwen3.5 num_labels propagation and zero-loss single-label classification are distinct sequence-classification issues." + "reason": "Config propagation bug versus a cross-model loss-degeneration problem; not the same bug." }, { "left": "issue:43421", "right": "issue:43927", "accept": false, - "reason": "Runtime special-token post-processor syncing and DiaConfig token-ID loss after save/load are not the same bug." + "reason": "Both involve special-token/config persistence, but one is runtime post-processor syncing and the other is saved custom token IDs being lost." }, { "left": "issue:44898", "right": "issue:45072", "accept": false, - "reason": "Perceiver interpolation failure and dtype mismatch in SwitchTransformers/TimmWrapperModel are unrelated." + "reason": "Different model failures: interpolation on Perceiver image classification versus bfloat16 dtype mismatches." }, { "left": "issue:43782", "right": "issue:45381", "accept": false, - "reason": "Qwen3VL weight_only loading and qwen2.5-vl video vision_position_ids are different multimodal bugs." + "reason": "Both are Qwen VL-related, but one is a load-time weight_only error and the other is wrong video vision_position_ids." }, { - "left": "issue:44077", - "right": "issue:44625", + "left": "issue:44451", + "right": "issue:44991", "accept": false, - "reason": "PatchTSMixer post_init policy and Qwen3.5 num_labels propagation are unrelated config issues." + "reason": "Different tokenizer-loading regressions affecting different models; not the same underlying issue." }, { - "left": "issue:45759", - "right": "issue:45859", + "left": "issue:44077", + "right": "issue:44625", "accept": false, - "reason": "Composite config unwrapping and CPU offload incompatibility in Qwen3_5MoeTextRotaryEmbedding are different problems." + "reason": "PatchTSMixer post_init validation and Qwen3.5 num_labels propagation are unrelated." } ] }, "evaluator_result": { "accept": true, - "feedback": "Grounded overall. The heterogeneous-cluster summary matches the packet, and the 44514/44545 pair is the only clearly duplicate-looking match by title/body. The other soft-edge rejections are conservative and consistent with the issue titles." + "feedback": "The summary is grounded in the packet and stays conservative. The only clear duplicate pair in the evidence is issue:44514/issue:44545, and the choice of issue:44545 as the anchor is supported by its slightly higher inbound/discussion signal. The soft-edge rejections are also cautious and generally consistent with the titles and descriptions provided." }, "evaluator_used": true, "retried": false, @@ -4522,13 +4552,13 @@ "component_index": 1, "component_count": 17, "review_unit_index": 14, - "review_unit_count": 49, + "review_unit_count": 50, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-427", + "cluster_id": "cluster-10000-428", "nodes": [ "issue:30064", "issue:36010", @@ -4565,7 +4595,6 @@ "issue:44918", "issue:44936", "issue:44987", - "issue:44991", "issue:45042", "issue:45081", "issue:45356", @@ -4580,10 +4609,10 @@ "issue:46050" ], "soft_pairs": [ - "issue:44451|issue:44991", + "issue:45759|issue:45859", "issue:43531|issue:44589", - "issue:44987|issue:45042", "issue:44368|issue:45357", + "issue:44987|issue:45042", "issue:44062|issue:44451", "issue:45588|issue:45759", "issue:39692|issue:44610", @@ -4595,8 +4624,8 @@ "issue:42886|issue:44843", "issue:43381|issue:43844", "issue:43421|issue:44186", - "issue:43994|issue:44671", "issue:45704|issue:45859", + "issue:43994|issue:44671", "issue:41720|issue:45704", "issue:43531|issue:43645", "issue:43746|issue:43873", @@ -4605,227 +4634,227 @@ "issue:40444|issue:43746", "issue:43596|issue:44220", "issue:45356|issue:45907", - "issue:43720|issue:43819", "issue:30064|issue:36010", + "issue:43720|issue:43819", "issue:43653|issue:43720", - "issue:43540|issue:44008", "issue:44220|issue:44661", + "issue:43540|issue:44008", "issue:43618|issue:43697" ], - "prepared_review_unit_hash": "6f6178b840e0f9961edfcaca7258a5296644b72a6b79d49589b5bca32d0cd319", + "prepared_review_unit_hash": "9ff48fa0394695b65daf12c29ed7a5c0554e5f279194e3a83bbb3aa3f2c47593", "component_budget": { - "node_count": 427, - "item_count": 427, - "soft_pair_count": 1576, - "serialized_chars": 816182, - "estimated_input_tokens": 204046, - "estimated_eval_tokens": 408348 + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 }, "budget": { - "node_count": 48, - "item_count": 48, + "node_count": 47, + "item_count": 47, "soft_pair_count": 31, - "serialized_chars": 37379, - "estimated_input_tokens": 9345, - "estimated_eval_tokens": 18946 + "serialized_chars": 36808, + "estimated_input_tokens": 9202, + "estimated_eval_tokens": 18660 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is a grab bag of mostly unrelated Transformers issues across tokenizers, multimodal models, quantization, training, and offline-mode regressions. I don\u2019t see any pair that clearly represents the same underlying bug/change, so no soft merges are justified.", - "confidence": 0.93, - "canonical_issue_reason": "issue:42886 is a clean, self-contained regression report with a specific version and reproducible offline-cache symptom, making it a reasonable representative issue for the cluster.", + "summary": "This cluster is mostly heterogeneous: the soft pairs share superficial model/tokenizer/offload wording, but they describe different bugs and code paths, so none should be deduplicated as the same underlying change or regression.", + "confidence": 0.94, + "canonical_issue_reason": "Issue 43873 is the closest representative of the cluster because it is the most discussed/high-signal item and sits in the broad quantization/offloading theme that recurs in a few nearby issues.", "canonical_pr_reason": null, - "best_issue_reason": "issue:42886 is the strongest standalone issue to keep as the cluster representative because it has a precise failure mode and clear reproduction context.", + "best_issue_reason": "Issue 43873 is the best single issue to anchor the cluster: it has the strongest engagement and is the most broadly framed, actionable bug among otherwise unrelated reports.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:44451", - "right": "issue:44991", + "left": "issue:45759", + "right": "issue:45859", "accept": false, - "reason": "Both are model/tokenizer loading failures, but they concern different models and likely different root causes." + "reason": "Same Qwen family, but different failures: composite config unwrapping vs CPU-offload rotary embedding behavior." }, { "left": "issue:43531", "right": "issue:44589", "accept": false, - "reason": "Completely different subsystems: Qwen3-MoE sliding window behavior vs a float8 storage import/type error." + "reason": "Qwen3-MoE sliding-window handling and missing Float8 storage are unrelated bugs." }, { - "left": "issue:44987", - "right": "issue:45042", + "left": "issue:44368", + "right": "issue:45357", "accept": false, - "reason": "Both are regressions, but one is a model-loading failure and the other is an image-processor torchvision dependency issue." + "reason": "One is a tie_word_embeddings warning; the other is a save_pretrained key regression for visual encoder weights." }, { - "left": "issue:44368", - "right": "issue:45357", + "left": "issue:44987", + "right": "issue:45042", "accept": false, - "reason": "Unrelated bugs: a warning about tie_word_embeddings vs incorrect visual encoder keys on save_pretrained." + "reason": "Different subsystems: loading a model repo versus PIL image processors incorrectly requiring torchvision." }, { "left": "issue:44062", "right": "issue:44451", "accept": false, - "reason": "Both involve tokenizers, but one is an AddedToken argument bug and the other is a model load failure for ScandiBERT." + "reason": "AddedToken constructor duplication and inability to load ScandiBERT are not the same issue." }, { "left": "issue:45588", "right": "issue:45759", "accept": false, - "reason": "Different code paths: flash_attention sink handling vs AutoModelForCausalLM/config unwrapping for multimodal Qwen configs." + "reason": "flash_attention s_aux handling and Qwen multimodal text_config unwrapping are separate code paths." }, { "left": "issue:39692", "right": "issue:44610", "accept": false, - "reason": "Both mention processor/model mismatches, but they affect different models and different concrete failures." + "reason": "Both are processor/model issues, but one is a SigLIP2 docs/model mismatch and the other is an OmDet-Turbo image-size mismatch." }, { "left": "issue:43746", "right": "issue:46050", "accept": false, - "reason": "Both are save/load related, but PEFT local checkpoint loading and sinq quantized round-tripping are distinct problems." + "reason": "Local checkpoint loading with PEFT adapters and quantized sinq save/load corruption are different persistence bugs." }, { "left": "issue:45704", "right": "issue:45923", "accept": false, - "reason": "Unrelated: apex FusedRMSNorm memory leak in T5 vs a missing gradient_checkpointing flag on a trust_remote_code model." + "reason": "Apex RMSNorm memory leak and a missing supports_gradient_checkpointing flag are unrelated." }, { "left": "issue:44918", "right": "issue:44936", "accept": false, - "reason": "Both mention trainer workflows, but one is embedding unpacking in SFT and the other is evaluate() failing after train()." + "reason": "Trainer embedding unpacking and evaluate-after-train failure do not share the same root cause." }, { "left": "issue:43421", "right": "issue:43653", "accept": false, - "reason": "Both concern tokenizer special tokens, but one is a feature gap in TokenizersBackend and the other is a BigBirdTokenizer CI/decoding bug." + "reason": "Both touch special tokens, but runtime post-processor refresh and BigBird mask-token registration are distinct problems." }, { "left": "issue:45663", "right": "issue:45923", "accept": false, - "reason": "Different failures: Gemma-4 FSDP2 shared_kv_states logic vs a missing support flag on a different model variant." + "reason": "Gemma4 FSDP2 shared_kv_states KeyError is unrelated to a missing gradient-checkpointing capability flag." }, { "left": "issue:42886", "right": "issue:44843", "accept": false, - "reason": "Same offline-mode theme, but one is generic cache loading and the other is a specific _patch_mistral_regex model_info call; not clearly the same bug." + "reason": "Both involve offline/tokenizer behavior, but one is generic cache loading and the other is a specific _patch_mistral_regex model_info call." }, { "left": "issue:43381", "right": "issue:43844", "accept": false, - "reason": "Different training/runtime behaviors: eval-mode gradient checkpointing vs abnormal gradient growth with ZeRO-3." + "reason": "Eval-mode gradient checkpointing and ZeRO-3 gradient growth are different training failures." }, { "left": "issue:43421", "right": "issue:44186", "accept": false, - "reason": "Both are tokenizer issues, but runtime special-token updates and LayoutLMv2 NER padding/truncation failures are distinct." + "reason": "Special-token post-processor updates and LayoutLMv2 NER/padding crashes are separate tokenizer bugs." }, { - "left": "issue:43994", - "right": "issue:44671", + "left": "issue:45704", + "right": "issue:45859", "accept": false, - "reason": "Different model-specific regressions: SigLIP2 incorrect outputs vs CamemBERT masked-LM prediction changes." + "reason": "T5 RMSNorm memory leak and Qwen offload incompatibility are unrelated." }, { - "left": "issue:45704", - "right": "issue:45859", + "left": "issue:43994", + "right": "issue:44671", "accept": false, - "reason": "Different bugs in different models: T5/apex RMSNorm leak vs Qwen3_5MoeTextRotaryEmbedding CPU offload incompatibility." + "reason": "Different models and symptoms: SigLIP2 nonsensical outputs versus CamemBERT masked-LM regressions." }, { "left": "issue:41720", "right": "issue:45704", "accept": false, - "reason": "No common code path: Qwen3 auto device mapping CUDA assert vs T5 RMSNorm memory leak." + "reason": "Qwen3 device-mapping CUDA asserts and a T5 RMSNorm memory leak are not the same bug." }, { "left": "issue:43531", "right": "issue:43645", "accept": false, - "reason": "Sliding-window behavior in Qwen3-MoE and notebook custom-model initialization in Transformers 5.0 are unrelated." + "reason": "Qwen3-MoE sliding-window behavior and Jupyter custom-model initialization are unrelated." }, { "left": "issue:43746", "right": "issue:43873", "accept": false, - "reason": "Both touch loading/offloading themes, but one is PEFT checkpoint loading and the other is quantization offloading behavior." + "reason": "Checkpoint loading with PEFT adapters and quantization/offloading behavior are different code paths." }, { "left": "issue:45081", "right": "issue:45356", "accept": false, - "reason": "Both involve tokenizer regex handling, but one is a crash in patch_mistral_regex and the other is a codec/ warning regression for Kimi-K2.5." + "reason": "Both are tokenizer regressions, but one is _patch_mistral_regex crash behavior and the other is Kimi-K2.5 codec handling/warning issues." }, { "left": "issue:39692", "right": "issue:43873", "accept": false, - "reason": "Unrelated: SigLIP2 docs/model mismatch vs quantization offloading not working as expected." + "reason": "SigLIP2 example errors and quantization offloading problems are unrelated." }, { "left": "issue:40444", "right": "issue:43746", "accept": false, - "reason": "Different model families and failures: Qwen2.5-VL multi-image dataset handling vs GraniteSpeech local checkpoint loading." + "reason": "Qwen2.5-VL multi-image iterable-dataset finetuning and local checkpoint loading with PEFT are different bugs." }, { "left": "issue:43596", "right": "issue:44220", "accept": false, - "reason": "No overlap beyond generic training/audio data processing; deepspeed zero3 BertModel index error and fbank feature extraction are separate bugs." + "reason": "ZeRO-3/BertModel init failure and _torch_extract_fbank_features are unrelated." }, { "left": "issue:45356", "right": "issue:45907", "accept": false, - "reason": "Tokenizer codec/regex regression vs list_repo_files missing revision forwarding are unrelated." + "reason": "Kimi tokenizer regression and list_repo_files revision forwarding are not the same issue." }, { - "left": "issue:43720", - "right": "issue:43819", + "left": "issue:30064", + "right": "issue:36010", "accept": false, - "reason": "Both are model/quantization related, but packed-weight loading and DAC STE mismatch are different concrete issues." + "reason": "Image processor void-segmentation-map failure and GenerationMixin import error are unrelated." }, { - "left": "issue:30064", - "right": "issue:36010", + "left": "issue:43720", + "right": "issue:43819", "accept": false, - "reason": "Completely unrelated: image processor void segmentation maps vs a GenerationMixin import error." + "reason": "BitNet packed-weight loading and DAC forward/from_latents mismatch are different model bugs." }, { "left": "issue:43653", "right": "issue:43720", "accept": false, - "reason": "Tokenizer special-token registration and BitNet packed-weight loading are unrelated." + "reason": "BigBird special-token registration and BitNet packed-weight loading are unrelated." }, { - "left": "issue:43540", - "right": "issue:44008", + "left": "issue:44220", + "right": "issue:44661", "accept": false, - "reason": "Different multimodal failures: Qwen3OmniMoe video processing vs Gemma 3n audio_mel_mask variable collision." + "reason": "Audio fbank extraction and add-new-model-like/tokenizer-mapping behavior are different problems." }, { - "left": "issue:44220", - "right": "issue:44661", + "left": "issue:43540", + "right": "issue:44008", "accept": false, - "reason": "Audio feature extraction and TOKENIZER_MAPPING_NAMES add-new-model-like behavior are unrelated." + "reason": "Video-input processing in Qwen3OmniMoe and Gemma3n audio_mel_mask variable collision are separate multimodal bugs." }, { "left": "issue:43618", "right": "issue:43697", "accept": false, - "reason": "Both are v5 regressions, but the affected models and failure modes are different, so they are not the same bug." + "reason": "Missing CLIPOutput attentions and RTDetrV2 output drift in v5 are distinct regressions." } ] }, @@ -4841,13 +4870,13 @@ "component_index": 1, "component_count": 17, "review_unit_index": 15, - "review_unit_count": 49, + "review_unit_count": 50, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-427", + "cluster_id": "cluster-10000-428", "nodes": [ "issue:30064", "issue:36331", @@ -4901,9 +4930,9 @@ "soft_pairs": [ "issue:45183|issue:45479", "issue:43550|issue:45072", - "issue:43550|issue:43819", - "issue:43295|issue:45834", "issue:43388|issue:46032", + "issue:43295|issue:45834", + "issue:43550|issue:43819", "issue:43716|issue:44008", "issue:44062|issue:44488", "issue:43322|issue:45685", @@ -4927,196 +4956,196 @@ "issue:44186|issue:44442", "issue:43644|issue:43749" ], - "prepared_review_unit_hash": "da1e89a6fdd673c4e30a7b6d3a587e30a5dfb21cfe6ea23ccafff86004bad8e0", + "prepared_review_unit_hash": "00fc4dfb87215b40e11935b4ad638dadfed10602ef525b7e71134ad4a409b294", "component_budget": { - "node_count": 427, - "item_count": 427, - "soft_pair_count": 1576, - "serialized_chars": 816182, - "estimated_input_tokens": 204046, - "estimated_eval_tokens": 408348 + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 27, - "serialized_chars": 36181, - "estimated_input_tokens": 9046, - "estimated_eval_tokens": 18348 + "serialized_chars": 36180, + "estimated_input_tokens": 9045, + "estimated_eval_tokens": 18346 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is a broad false positive: the items span unrelated tokenizer, multimodal, caching, loading, and model-specific regressions. None of the soft edges look like the same underlying bug/change.", - "confidence": 0.97, - "canonical_issue_reason": null, + "summary": "All proposed soft edges are false positives: they span different models, subsystems, and failure modes rather than the same underlying bug or change. There is no clear canonical duplicate target in this set.", + "confidence": 0.92, + "canonical_issue_reason": "No single issue is a credible canonical representative; the items are a heterogeneous mix of unrelated tokenizer, multimodal, cache, loading, and dtype bugs.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 45081 is the cleanest single, narrowly scoped regression in the set, but the cluster is too heterogeneous for a true canonical duplicate target.", + "best_issue_reason": "No issue stands out as a global duplicate target for the cluster, because the cluster does not collapse to one concrete defect.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:45183", "right": "issue:45479", "accept": false, - "reason": "Both are model validation/loss bugs, but one is an input XOR check and the other is a zero-loss classification bug; different failure modes and code paths." + "reason": "Both involve validation/classification behavior, but one is input-argument XOR checking across models and the other is a zero-loss bug for num_labels=1; different defects." }, { "left": "issue:43550", "right": "issue:45072", "accept": false, - "reason": "Different models and different bugs: SDPA/torch.compile failure in Bamba vs dtype mismatch in SwitchTransformers/TimmWrapperModel." + "reason": "Same general area of inference/runtime errors, but Bamba SDPA + torch.compile and dtype mismatches in SwitchTransformers/TimmWrapperModel are different code paths and failures." }, { - "left": "issue:43550", - "right": "issue:43819", + "left": "issue:43388", + "right": "issue:46032", "accept": false, - "reason": "Bamba compile/SDPA failure is unrelated to DAC.from_latents vs forward-pass mismatch with missing STE." + "reason": "gather_for_metrics label truncation and Mamba2Mixer cache corruption are unrelated subsystems with different symptoms." }, { "left": "issue:43295", "right": "issue:45834", "accept": false, - "reason": "One is a processor/tokenizer regression, the other is a long-OCR index error in Kosmos2.5; no shared root cause." + "reason": "Processor.tokenizer regression and Kosmos2.5 long-OCR index error are different multimodal loading/processing bugs." }, { - "left": "issue:43388", - "right": "issue:46032", + "left": "issue:43550", + "right": "issue:43819", "accept": false, - "reason": "gather_for_metrics last-batch label trimming is unrelated to Mamba2Mixer use_cache producing wrong results." + "reason": "Bamba torch.compile/SDPA failure and DAC.from_latents mismatch are unrelated model-specific bugs." }, { "left": "issue:43716", "right": "issue:44008", "accept": false, - "reason": "Different components and symptoms: image preprocessor/model dtype mismatch vs a Gemma3n forward variable-name collision causing AttributeError." + "reason": "Mistral-3 image preprocessor dtype mismatch is not the same as Gemma3n's variable-name collision in forward()." }, { "left": "issue:44062", "right": "issue:44488", "accept": false, - "reason": "AddedToken special-kwarg duplication is a tokenizer-construction bug, while sleng-bert failing to load is a separate model/tokenizer loading issue." + "reason": "AddedToken special-argument duplication and a tokenizer load failure for cjvt/sleng-bert are distinct tokenizer issues." }, { "left": "issue:43322", "right": "issue:45685", "accept": false, - "reason": "Llava Next loading segfault and MPS histogram on Int are unrelated runtime failures." + "reason": "Llava Next loading segfault and MPS histogram on Int in moe are different failure modes in different paths." }, { "left": "issue:43335", "right": "issue:43756", "accept": false, - "reason": "Both are configuration bugs, but one creates a sparse layer at zero sparse layers and the other drops RoPE layers; different models and logic." + "reason": "SwitchTransformers sparse-layer construction and Smollm3 RoPE-layer dropping are different model/config bugs." }, { "left": "issue:44186", "right": "issue:44898", "accept": false, - "reason": "LayoutLMv2 tokenizer crashes on NER/padding vs Perceiver image classification failing with interpolate_pos_encoding; unrelated subsystems." + "reason": "LayoutLMv2Tokenizer NER/padding crash and Perceiver non-default-resolution failure are unrelated." }, { "left": "issue:43653", "right": "issue:43819", "accept": false, - "reason": "BigBirdTokenizer special-token registration bug is not the same as DAC.from_latents diverging from forward due to missing STE." + "reason": "BigBirdTokenizer special-token registration and DAC.from_latents/forward mismatch do not share a concrete code-path bug." }, { "left": "issue:43950", "right": "issue:45987", "accept": false, - "reason": "Non-persistent buffer corruption on load is a different regression from StaticCache.get_seq_length returning the wrong type." + "reason": "Non-persistent buffer corruption on from_pretrained and StaticCache.get_seq_length returning a tensor are separate regression classes." }, { "left": "issue:43531", "right": "issue:43643", "accept": false, - "reason": "Qwen3-MoE sliding_window behavior and missing fields from trust_remote_code AutoConfig are unrelated bugs." + "reason": "Qwen3-MoE sliding_window behavior and trust_remote_code AutoConfig field loss are unrelated issues." }, { "left": "issue:44843", "right": "issue:45081", "accept": false, - "reason": "Both involve _patch_mistral_regex, but one fails because it calls model_info() offline and the other because it assumes backend_tokenizer exists; distinct root causes." + "reason": "Both mention Mistral regex patching, but one breaks offline mode via model_info() and the other crashes on a backend_tokenizer attribute assumption; not the same bug." }, { "left": "issue:43643", "right": "issue:44877", "accept": false, - "reason": "Missing fields with trust_remote_code is not the same as strict config blocking granite_speech loading." + "reason": "Missing config fields with trust_remote_code and strict config rejection for granite_speech are different config-loading problems." }, { "left": "issue:43540", "right": "issue:45381", "accept": false, - "reason": "Both are Qwen video issues, but they affect different model families and different code paths (video input parsing vs vision_position_ids)." + "reason": "Qwen3OmniMoe video processing ValueError and Qwen2.5-VL video vision_position_ids mismatch are different model-specific video bugs." }, { "left": "issue:44448", "right": "issue:44671", "accept": false, - "reason": "Both are v5 output regressions, but they concern different models and different generation/masked-LM paths." + "reason": "Pegasus v4/v5 output differences and CamemBERT MLM prediction errors are unrelated model behavior regressions." }, { "left": "issue:30064", "right": "issue:36331", "accept": false, - "reason": "ImageProcessor segmentation-map failure and CustomTrainer.compute_loss keyword regression are unrelated." + "reason": "Image processor void-segmentation handling and CustomTrainer.compute_loss keyword-arg failure are unrelated." }, { "left": "issue:43421", "right": "issue:44811", "accept": false, - "reason": "Runtime special-token/post-processor updates and Whisper batch_decode skip_special_tokens handling are different layers of the stack." + "reason": "Runtime special-token post-processor updates and Whisper batch_decode skip_special_tokens handling are different layers and symptoms." }, { "left": "issue:43577", "right": "issue:43818", "accept": false, - "reason": "BLIP2 dtype-loading bug and Video-LLaVA missing temporal attention/shared weights are unrelated model-specific issues." + "reason": "Blip2 dtype persistence and Video-LLaVA missing temporal attention/shared weights are unrelated defects." }, { "left": "issue:43782", "right": "issue:44977", "accept": false, - "reason": "from_pretrained weight_only=True load error is a different failure mode from flash-attention generation problems." + "reason": "Qwen3VL weight_only loading failure and Qwen3.5 flash-attention generation bug are distinct model/runtime issues." }, { "left": "issue:41720", "right": "issue:44991", "accept": false, - "reason": "Qwen3 auto-device-map cuda assert and EMBEDDIA tokenizer loading failure are unrelated." + "reason": "A800 cudaErrorAssert under auto device mapping and tokenizer load failure for est-roberta have no shared underlying code path." }, { "left": "issue:39692", "right": "issue:45834", "accept": false, - "reason": "SigLIP2 docs/model mismatch and Kosmos2.5 long OCR index error do not share a concrete bug." + "reason": "SigLIP2 docs example errors and Kosmos2.5 long OCR index error are unrelated." }, { "left": "issue:45584", "right": "issue:45834", "accept": false, - "reason": "Whisper empty-transcription generation after align_special_tokens is unrelated to Kosmos2.5 OCR index errors." + "reason": "Whisper empty-transcription generation failure and Kosmos2.5 OCR indexing are different modalities and bugs." }, { "left": "issue:44821", "right": "issue:45084", "accept": false, - "reason": "AutoImageProcessor URL loading and template-node compilation errors are entirely different code paths." + "reason": "AutoImageProcessor-from-URL loading and non-template-node compilation errors are unrelated." }, { "left": "issue:44186", "right": "issue:44442", "accept": false, - "reason": "LayoutLMv2 tokenizer crash and AutoTokenizer failing to load FastSpeech2ConformerTokenizer are different tokenizer issues." + "reason": "LayoutLMv2Tokenizer NER crash and AutoTokenizer loading FastSpeech2ConformerTokenizer are different tokenizer failures." }, { "left": "issue:43644", "right": "issue:43749", "accept": false, - "reason": "Non-persistent buffer corruption on from_pretrained and broken FSDP_CPU_RAM_EFFICIENT_LOADING are separate loading regressions." + "reason": "Non-persistent buffer corruption and broken FSDP CPU RAM efficient loading are separate regressions." } ] }, @@ -5132,13 +5161,13 @@ "component_index": 1, "component_count": 17, "review_unit_index": 16, - "review_unit_count": 49, + "review_unit_count": 50, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-427", + "cluster_id": "cluster-10000-428", "nodes": [ "issue:28218", "issue:29159", @@ -5147,7 +5176,6 @@ "issue:36331", "issue:42175", "issue:43262", - "issue:43329", "issue:43526", "issue:43577", "issue:43582", @@ -5185,246 +5213,244 @@ "issue:45325", "issue:45406", "issue:45538", - "issue:45600", "issue:45701", - "issue:45834" + "issue:45834", + "issue:45987", + "issue:46192" ], "soft_pairs": [ - "issue:43606|issue:45072", "issue:44561|issue:45020", + "issue:43606|issue:45072", "issue:44479|issue:45834", "issue:43825|issue:43994", "issue:34634|issue:36010", "issue:43577|issue:43582", + "issue:45987|issue:46192", "issue:44805|issue:44936", "issue:44485|issue:44509", "issue:43825|issue:43827", "issue:44671|issue:45701", "issue:28218|issue:29159", "issue:44610|issue:45072", - "issue:43720|issue:45072", "issue:43782|issue:45325", - "issue:43749|issue:44291", + "issue:43720|issue:45072", "issue:42175|issue:45081", + "issue:43749|issue:44291", "issue:45081|issue:45406", "issue:43582|issue:44857", "issue:43650|issue:43906", "issue:43262|issue:43819", "issue:43526|issue:43697", - "issue:34634|issue:36331", "issue:44060|issue:44368", + "issue:34634|issue:36331", "issue:43653|issue:44912", "issue:43526|issue:44554", "issue:43742|issue:45020", - "issue:43329|issue:45600", "issue:45084|issue:45538" ], - "prepared_review_unit_hash": "112c6a9b5f4b086de33666976d96923f42ef68caeb052f1f5e6d25c17d69dc23", + "prepared_review_unit_hash": "7b9d8ac8bfa88298c138081ecd3e815c3c6475ef469708c09451cb763070a4c7", "component_budget": { - "node_count": 427, - "item_count": 427, - "soft_pair_count": 1576, - "serialized_chars": 816182, - "estimated_input_tokens": 204046, - "estimated_eval_tokens": 408348 + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 28, - "serialized_chars": 36384, - "estimated_input_tokens": 9096, - "estimated_eval_tokens": 18448 + "serialized_chars": 36305, + "estimated_input_tokens": 9077, + "estimated_eval_tokens": 18410 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "Mostly a heterogeneous cluster of tokenizer, model-loading, and multimodal regression reports. Only a couple of pairs look like true duplicates; most share a subsystem or symptom but not the same underlying bug.", - "confidence": 0.79, - "canonical_issue_reason": "issue:45020 is the broadest umbrella report in the set and best fits the small remote_code/v5 regression subgroup, though the overall cluster is too mixed to have a single true canonical issue.", + "summary": "This is a noisy grab-bag of unrelated Transformers issues across tokenization, processors, model loading, training, and docs. I don\u2019t see a single underlying bug/change that unifies the cluster, so there\u2019s no strong canonical item.", + "confidence": 0.96, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "issue:45020 is the best representative issue because it captures the widest v5/remote_code breakage theme and has the broadest scope among the near-duplicates.", + "best_issue_reason": null, "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:43606", - "right": "issue:45072", + "left": "issue:44561", + "right": "issue:45020", "accept": false, - "reason": "Same broad inference symptom class, but different models and failure mechanisms; not the same underlying bug." + "reason": "Both mention remote_code breakage, but one is a specific v5 removal of is_torch_fx_available and the other is a broad report about recent versions; not clearly the same concrete bug." }, { - "left": "issue:44561", - "right": "issue:45020", - "accept": true, - "reason": "Both describe v5 breaking trust_remote_code/remote_code models via internal API removals; 44561 is a concrete instance of the broader 45020 regression." + "left": "issue:43606", + "right": "issue:45072", + "accept": false, + "reason": "Both involve dtype/device mismatches, but they affect different models and code paths (CPU offload vs bfloat16 inference)." }, { "left": "issue:44479", "right": "issue:45834", "accept": false, - "reason": "Both are multimodal input regressions, but one is Qwen-VL video indexing and the other is Kosmos2.5 OCR long-input indexing." + "reason": "Different model families and failures: a video-input regression in Qwen VL models versus a long-OCR index error in Kosmos2.5." }, { "left": "issue:43825", "right": "issue:43994", "accept": false, - "reason": "One is a pipeline error-message regression, the other is a model output correctness issue via AutoModel/pipeline; unrelated bugs." + "reason": "One is an error-message regression around removed translation tasks; the other is a bad AutoModel/pipeline result for SigLIP2. Different bugs." }, { "left": "issue:34634", "right": "issue:36010", "accept": false, - "reason": "BarkProcessor voice_preset and a GenerationMixin import error are unrelated." + "reason": "Completely different failures: BarkProcessor voice_preset handling versus GenerationMixin import errors." }, { "left": "issue:43577", "right": "issue:43582", "accept": false, - "reason": "Different model-loading/runtime failures on different platforms; no shared code-path bug is evident." + "reason": "Unrelated issues: BLIP2 dtype propagation versus an Apple Silicon allocator warmup TypeError." + }, + { + "left": "issue:45987", + "right": "issue:46192", + "accept": false, + "reason": "Different API surfaces and symptoms; one is StaticCache seq-length return type, the other is text_kwargs/return_tensors handling." }, { "left": "issue:44805", "right": "issue:44936", "accept": false, - "reason": "Mask/index shape mismatch and trainer.evaluate failing after train are distinct failures." + "reason": "A tensor mask shape mismatch is not the same bug as trainer.evaluate() failing after train()." }, { "left": "issue:44485", "right": "issue:44509", "accept": false, - "reason": "RoPE implementation discussion vs docs references to removed pipeline tasks are not the same issue." + "reason": "RoPE implementation discussion is unrelated to docs still referencing removed pipeline tasks." }, { "left": "issue:43825", "right": "issue:43827", "accept": false, - "reason": "Both concern v5 pipeline removals, but one is runtime messaging and the other is stale documentation." + "reason": "Both are about v5 pipeline removals, but one concerns runtime messaging and the other documentation; not the same concrete issue." }, { "left": "issue:44671", "right": "issue:45701", "accept": false, - "reason": "Both mention tokenization/version effects, but one is CamemBERT masked-LM correctness and the other is a broad tokenization change complaint." + "reason": "CamemBERT masked-LM regressions and version-dependent tokenization changes are too broad and not clearly the same bug." }, { "left": "issue:28218", "right": "issue:29159", - "accept": true, - "reason": "Both point to tokenizer inconsistency around added-token handling; 28218 is a concrete symptom of the broader slow/fast tokenizer mismatch in 29159." - }, - { - "left": "issue:44610", - "right": "issue:45072", "accept": false, - "reason": "Different model families and different regression types; only a generic inference mismatch theme overlaps." + "reason": "Both are tokenizer complaints, but the titles don\u2019t establish the same root cause; one is added-token spacing, the other slow/fast inconsistency." }, { - "left": "issue:43720", + "left": "issue:44610", "right": "issue:45072", "accept": false, - "reason": "Packed-weight loading failure and dtype mismatch are unrelated." + "reason": "Different models and failure modes: OmDet-Turbo processor size mismatch versus dtype mismatches in SwitchTransformers/TimmWrapperModel." }, { "left": "issue:43782", "right": "issue:45325", "accept": false, - "reason": "Both are Qwen VL issues, but one is a weight_only load error and the other is a RoPE/position-id scaling bug." + "reason": "One is a Qwen3VL weight_only load error, the other a Qwen2.5-VL rope-index scaling bug; different code paths and symptoms." }, { - "left": "issue:43749", - "right": "issue:44291", + "left": "issue:43720", + "right": "issue:45072", "accept": false, - "reason": "Both are loading-related regressions, but one is FSDP CPU RAM efficient loading and the other is init_empty_weights argument handling." + "reason": "BitNet packed-weight loading is unrelated to bfloat16 dtype mismatch issues in other models." }, { "left": "issue:42175", "right": "issue:45081", "accept": false, - "reason": "Backend packaging and Mistral regex patch loading are unrelated." + "reason": "Backend/package install configuration is unrelated to the Mistral tokenizer regex patch crash." + }, + { + "left": "issue:43749", + "right": "issue:44291", + "accept": false, + "reason": "Different loading mechanisms and failures: FSDP CPU RAM efficient loading versus init_empty_weights receiving an unexpected argument." }, { "left": "issue:45081", "right": "issue:45406", "accept": false, - "reason": "Tokenizer regex patch crash and Gemma4Processor missing _tokenizer are different code paths." + "reason": "Tokenizer regex patching and Gemma4Processor missing _tokenizer are different loading/processor bugs." }, { "left": "issue:43582", "right": "issue:44857", "accept": false, - "reason": "Apple Silicon allocator warmup TypeError vs CUDA AMP loss crash are unrelated." + "reason": "AppleSilicon warmup TypeError is unrelated to a CUDA/AMP crash in LwDetrImageLoss." }, { "left": "issue:43650", "right": "issue:43906", "accept": false, - "reason": "Generic placeholder issue and isolated reproduction report do not match on the underlying bug." + "reason": "These titles are not about the same bug; one is nonspecific 'ADD THE DATA', the other is an isolated reproduction of another issue." }, { "left": "issue:43262", "right": "issue:43819", "accept": false, - "reason": "Audio sampling-rate default and DAC forward/from_latents mismatch are unrelated." + "reason": "Audio chat-template sampling-rate defaults and DAC from_latents/STE mismatch are unrelated." }, { "left": "issue:43526", "right": "issue:43697", "accept": false, - "reason": "Label reduction bug in BeitImageProcessorFast vs RTDetrV2 output regression are different defects." + "reason": "BeitImageProcessorFast reduce_labels behavior and RTDetrV2 output differences are different model-specific bugs." }, { - "left": "issue:34634", - "right": "issue:36331", + "left": "issue:44060", + "right": "issue:44368", "accept": false, - "reason": "BarkProcessor voice_preset and CustomTrainer.compute_loss signature mismatch are unrelated." + "reason": "Both mention Qwen3/Qwen3.5 warnings, but one is a tied-weights warning bug and the other is a separate lora fine-tuning warning; not the same issue." }, { - "left": "issue:44060", - "right": "issue:44368", + "left": "issue:34634", + "right": "issue:36331", "accept": false, - "reason": "Both involve tied-weights warnings, but 44060 is an incorrect internal tie mapping while 44368 is a configuration/warning issue in fine-tuning; not the same bug." + "reason": "BarkProcessor voice_preset and CustomTrainer.compute_loss signature mismatch are unrelated." }, { "left": "issue:43653", "right": "issue:44912", "accept": false, - "reason": "BigBirdTokenizer special-token registration and MXFP4 quantization loading are unrelated." + "reason": "BigBirdTokenizer special-token decode failure is unrelated to MXFP4 quantization fallback." }, { "left": "issue:43526", "right": "issue:44554", "accept": false, - "reason": "Vision label processing and an MPS attention correctness issue are unrelated." + "reason": "A Beit image-processor label reduction bug is unrelated to an MPS attention correctness issue." }, { "left": "issue:43742", "right": "issue:45020", "accept": false, - "reason": "Both are model-loading regressions, but this is a model-specific key error while 45020 is a broad remote_code regression umbrella." - }, - { - "left": "issue:43329", - "right": "issue:45600", - "accept": false, - "reason": "Undefined multimodal-token helper variables and removed auto-mapping configs are unrelated CI/code-path problems." + "reason": "Loading facebook/MobileLLM-125M key errors and remote_code breakages are different failure classes." }, { "left": "issue:45084", "right": "issue:45538", "accept": false, - "reason": "Template compilation error and CLIPTokenizer model_max_length behavior are unrelated." + "reason": "Template compilation errors are unrelated to CLIPTokenizer model_max_length handling." } ] }, - "evaluator_result": { - "accept": true, - "feedback": "The summary is grounded and appropriately cautious. The cluster is described as mostly heterogeneous, and the rejected soft edges are conservative. The two accepted pairs are defensible from the packet: 44561/45020 is a clear remote_code/v5 regression linkage, and 28218/29159 is a plausible tokenizer added-token inconsistency relation. No mergeability issues apply here since these are issue pairings, not PR pairs." - }, - "evaluator_used": true, + "evaluator_result": null, + "evaluator_used": false, "retried": false, - "accepted_nontrivial_soft_edge": true, + "accepted_nontrivial_soft_edge": false, "error_kind": null, "error_message": null }, @@ -5433,13 +5459,13 @@ "component_index": 1, "component_count": 17, "review_unit_index": 17, - "review_unit_count": 49, + "review_unit_count": 50, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-427", + "cluster_id": "cluster-10000-428", "nodes": [ "issue:29159", "issue:34567", @@ -5484,20 +5510,21 @@ "issue:45042", "issue:45356", "issue:45459", - "issue:45488", + "issue:45600", "issue:45701", "issue:45759", "issue:45803", "issue:45859" ], "soft_pairs": [ + "issue:43329|issue:45600", "issue:43746|issue:44811", - "issue:44589|issue:45042", "issue:43452|issue:44991", + "issue:44589|issue:45042", "issue:43638|issue:44936", - "issue:43819|issue:43927", - "issue:45459|issue:45803", "issue:43550|issue:44912", + "issue:45459|issue:45803", + "issue:43819|issue:43927", "issue:36010|issue:43645", "issue:43257|issue:43329", "issue:43854|issue:44977", @@ -5505,9 +5532,9 @@ "issue:38175|issue:44610", "issue:43720|issue:44610", "issue:36010|issue:36331", - "issue:43262|issue:44811", "issue:43257|issue:43696", "issue:43454|issue:43577", + "issue:43262|issue:44811", "issue:43531|issue:45042", "issue:42490|issue:43257", "issue:43329|issue:44479", @@ -5519,217 +5546,216 @@ "issue:44568|issue:45356", "issue:43257|issue:43931", "issue:29159|issue:34567", - "issue:44849|issue:45701", - "issue:44568|issue:45488" + "issue:44849|issue:45701" ], - "prepared_review_unit_hash": "f15955feb45ee50da13545426c1b6063a628399c2e887c40c3340dad9e6e53f5", + "prepared_review_unit_hash": "cbe35caf3d26713b03d5195fcaf6eebe671272bba13e85f8bce8540dfde6dee1", "component_budget": { - "node_count": 427, - "item_count": 427, - "soft_pair_count": 1576, - "serialized_chars": 816182, - "estimated_input_tokens": 204046, - "estimated_eval_tokens": 408348 + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 30, - "serialized_chars": 36995, - "estimated_input_tokens": 9249, - "estimated_eval_tokens": 18754 + "serialized_chars": 36943, + "estimated_input_tokens": 9236, + "estimated_eval_tokens": 18728 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is not a true duplicate set: it mixes unrelated tokenizer, trainer, model-loading, multimodal/video, CPU-offload, and quantization bugs. Most soft edges are superficial topic matches only and should stay separate.", - "confidence": 0.9, - "canonical_issue_reason": "No single canonical issue fits the cluster well; the items span many unrelated code paths and failure modes, so there is no clear umbrella bug to normalize around.", + "summary": "The cluster is highly heterogeneous. The soft pairs mostly connect superficially similar Transformers issues across unrelated areas (tokenizers, loading, training, multimodal/video, offload, quantization), so none look like true duplicates.", + "confidence": 0.89, + "canonical_issue_reason": "issue:43329 is the most concrete, code-path-specific report in the densest soft-similarity pocket (multimodal/video token counting), so it is the best representative if a canonical must be chosen.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45488 is the best representative only because it is an active, broadly framed tokenizer regression with inbound references, but it still does not subsume the rest of the heterogeneous set.", + "best_issue_reason": "issue:43329 is the clearest representative of the only semi-coherent subtheme, but the cluster overall is not a true duplicate set.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:43746", - "right": "issue:44811", + "left": "issue:43329", + "right": "issue:45600", "accept": false, - "reason": "Different subsystems and symptoms: PEFT/local checkpoint loading vs Whisper batch_decode skip_special_tokens." + "reason": "Video token-counting undefined-variable bug vs removed config reference breaking CI; different subsystems and failure modes." }, { - "left": "issue:44589", - "right": "issue:45042", + "left": "issue:43746", + "right": "issue:44811", "accept": false, - "reason": "Unrelated bugs: float8 storage lookup vs PIL backend requiring torchvision." + "reason": "PEFT adapter local-checkpoint loading vs Whisper batch_decode skip_special_tokens behavior; unrelated code paths." }, { "left": "issue:43452", "right": "issue:44991", "accept": false, - "reason": "Both are tokenizer/model-load failures, but the concrete causes differ: gguf_file handling vs a specific tokenizer regression for est-roberta." + "reason": "gguf_file loader breakage vs tokenizer codec handling regression for a specific model; not the same bug." + }, + { + "left": "issue:44589", + "right": "issue:45042", + "accept": false, + "reason": "Float8 storage lookup error vs PIL backend unexpectedly requiring torchvision; distinct import/runtime issues." }, { "left": "issue:43638", "right": "issue:44936", "accept": false, - "reason": "Different training failures: ZeRO-3 with non-pretrained Bert IndexError vs evaluate() failing after train()." + "reason": "Deepspeed ZeRO-3 init index error vs trainer.evaluate failing after train; different trainer lifecycle bugs." }, { - "left": "issue:43819", - "right": "issue:43927", + "left": "issue:43550", + "right": "issue:44912", "accept": false, - "reason": "Different model bugs: DAC latent reconstruction mismatch vs Dia config losing custom token IDs." + "reason": "torch.compile + SDPA failure on Bamba vs MXFP4 quantization fallback on gpt-oss; not the same concrete path." }, { "left": "issue:45459", "right": "issue:45803", "accept": false, - "reason": "Same broad pattern of swallowed errors, but different code paths and root causes; not the same underlying bug." + "reason": "Tokenizer protobuf import-error masking vs bare except in FuyuBatchFeature conversion; both are error-handling issues, but not the same defect." }, { - "left": "issue:43550", - "right": "issue:44912", + "left": "issue:43819", + "right": "issue:43927", "accept": false, - "reason": "Different features: torch.compile/SDPA runtime issue vs MXFP4 quantization fallback." + "reason": "DAC latent reconstruction mismatch vs DiaConfig custom token IDs lost on save/load; different models and bug classes." }, { "left": "issue:36010", "right": "issue:43645", "accept": false, - "reason": "Importing GenerationMixin is unrelated to notebook custom-model initialization behavior." + "reason": "GenerationMixin import regression vs custom-model initialization in Jupyter; unrelated initialization/import behavior." }, { "left": "issue:43257", "right": "issue:43329", "accept": false, - "reason": "Different code paths: Qwen3 MoE weight conversion/loading vs multimodal token counting video-branch bug." + "reason": "Qwen3-MoE loading/conversion with accelerate+deepspeed vs multimodal video token counting undefined variables; different failure points." }, { "left": "issue:43854", "right": "issue:44977", "accept": false, - "reason": "Different models and failures: GLM-4.7-Flash loading tests vs Qwen3.5 flash-attention generation." + "reason": "Unit-test model loading for GLM-4.7 vs Qwen3.5 flash-attention generation bug; not the same underlying issue." }, { "left": "issue:39692", "right": "issue:44912", "accept": false, - "reason": "Both mention quantization, but one is a SigLIP2 docs/example problem and the other is a GPT-oss MXFP4 loading issue." + "reason": "SigLIP2 doc example/model-processor mismatch vs MXFP4 quantization fallback; one is documentation/examples, the other model loading/quantization." }, { "left": "issue:38175", "right": "issue:44610", "accept": false, - "reason": "Different symptom areas: SigLIP2 zero probabilities vs OmDet-Turbo processor size mismatch." + "reason": "Unexpected zero probabilities in SigLIP2 vs OmDet-Turbo processor image-size mismatch; different models and symptoms." }, { "left": "issue:43720", "right": "issue:44610", "accept": false, - "reason": "Packed-weight loading bug is unrelated to OmDet-Turbo image sizing." + "reason": "BitNet packed-weight unpacking during accelerate load vs processor output-size mismatch; not the same code path." }, { "left": "issue:36010", "right": "issue:36331", "accept": false, - "reason": "ImportError for GenerationMixin and CustomTrainer.compute_loss signature mismatch are unrelated." - }, - { - "left": "issue:43262", - "right": "issue:44811", - "accept": false, - "reason": "Both touch audio processors, but one is chat-template sample rate defaulting and the other is batch_decode skip_special_tokens handling." + "reason": "GenerationMixin import failure vs CustomTrainer.compute_loss signature mismatch; unrelated trainer/import problems." }, { "left": "issue:43257", "right": "issue:43696", "accept": false, - "reason": "Qwen3 MoE conversion/loading is unrelated to GPT-oss out-of-memory." + "reason": "Model weight conversion/load bug vs CUDA OOM on GPT-oss-20b; different root causes." }, { "left": "issue:43454", "right": "issue:43577", "accept": false, - "reason": "AyaVision lm_head tying and BLIP2 dtype propagation are separate model bugs." + "reason": "lm_head weight tying bug vs dtype propagation issue in BLIP2; both model-loading bugs, but not the same defect." + }, + { + "left": "issue:43262", + "right": "issue:44811", + "accept": false, + "reason": "Audio chat template sample-rate default bug vs Whisper batch_decode skip_special_tokens handling; separate processor APIs." }, { "left": "issue:43531", "right": "issue:45042", "accept": false, - "reason": "Sliding-window handling for Qwen3-MoE is unrelated to PIL/torchvision image processor requirements." + "reason": "Qwen3-MoE sliding_window behavior vs PIL backend torchvision dependency regression; unrelated areas." }, { "left": "issue:42490", "right": "issue:43257", "accept": false, - "reason": "Trailing-slash path import warning is not the same as Qwen3 MoE weight conversion." + "reason": "Trailing-slash local-path import warning vs Qwen3-MoE accelerate/deepspeed load failure; not the same issue." }, { "left": "issue:43329", "right": "issue:44479", "accept": false, - "reason": "Both are about video multimodal behavior, but one is an undefined-helper bug in token counting and the other is a broader regression report across models." + "reason": "Undefined multimodal video helper variables vs broad v5.3.0 video-input regression; overlapping topic, but not the same concrete bug." }, { "left": "issue:43643", "right": "issue:45759", "accept": false, - "reason": "Different config-loading issues: missing fields from trust_remote_code vs unwrapping text_config for composite multimodal configs." + "reason": "AutoConfig trust_remote_code missing fields vs AutoModelForCausalLM.from_config not unwrapping text_config; related config handling, but distinct." }, { "left": "issue:43749", "right": "issue:44756", "accept": false, - "reason": "Both involve memory/loading, but FSDP CPU-efficient loading and mmap on Strix Halo are not the same bug." + "reason": "FSDP CPU RAM efficient loading regression vs Strix Halo mmap OOM mitigation; different memory-loading issues." }, { "left": "issue:43329", "right": "issue:43450", "accept": false, - "reason": "Video token counting helper failure and batched video processor shape mismatch are distinct issues." + "reason": "Undefined video-token helper names vs batched video processor shape error; both multimodal/video, but separate bugs." }, { "left": "issue:43606", "right": "issue:45859", "accept": false, - "reason": "Both mention CPU offload, but they affect different models and code paths." + "reason": "Bark CPU-offload device mismatch vs Qwen3-MoE rotary embedding CPU-offload incompatibility; same general theme, different models/code paths." }, { "left": "issue:34567", "right": "issue:40990", "accept": false, - "reason": "TrainerState token counting and GPT-oss perplexity are unrelated." + "reason": "TrainerState token counter not updating vs high perplexity on GPT-oss-20b; no concrete overlap." }, { "left": "issue:44568", "right": "issue:45356", "accept": false, - "reason": "Separate tokenizer regressions affecting different models and mechanisms." + "reason": "Tokenizer add_special_tokens regression vs Kimi-K2.5 codec handling regression; both tokenizer issues, but different mechanics." }, { "left": "issue:43257", "right": "issue:43931", "accept": false, - "reason": "Different Qwen3 loading failures: MoE conversion vs VL weight-shape mismatch." + "reason": "Qwen3-MoE load/conversion issue vs Qwen3-VL-30B weight-shape mismatch; both model-loading problems, but not the same bug." }, { "left": "issue:29159", "right": "issue:34567", "accept": false, - "reason": "Tokenizer inconsistency and TrainerState token tracking are unrelated." + "reason": "Slow/fast tokenizer inconsistency vs TrainerState num_input_tokens_seen not updating; unrelated subsystems." }, { "left": "issue:44849", "right": "issue:45701", "accept": false, - "reason": "Qwen3.5 hidden-states bug is unrelated to a general tokenization-version change." - }, - { - "left": "issue:44568", - "right": "issue:45488", - "accept": false, - "reason": "Both are tokenizer issues, but they are different regressions with different models and failure modes." + "reason": "Qwen3.5 hidden-states bug vs broad tokenization changes across versions; too generic and not the same defect." } ] }, @@ -5745,13 +5771,13 @@ "component_index": 1, "component_count": 17, "review_unit_index": 18, - "review_unit_count": 49, + "review_unit_count": 50, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-427", + "cluster_id": "cluster-10000-428", "nodes": [ "issue:24643", "issue:28282", @@ -5784,7 +5810,7 @@ "issue:44479", "issue:44488", "issue:44560", - "issue:44737", + "issue:44568", "issue:44821", "issue:44898", "issue:44912", @@ -5796,21 +5822,22 @@ "issue:45357", "issue:45381", "issue:45405", + "issue:45488", "issue:45538", - "issue:45588", "issue:45823", "issue:45907", "issue:45987" ], "soft_pairs": [ + "issue:44568|issue:45488", "issue:42915|issue:43742", "issue:44291|issue:45042", + "issue:43697|issue:45357", "issue:42915|issue:44821", "issue:41720|issue:43540", - "issue:43697|issue:45357", + "issue:44987|issue:45538", "issue:44479|issue:45381", "issue:43452|issue:43582", - "issue:44987|issue:45538", "issue:43927|issue:44912", "issue:42915|issue:45538", "issue:33357|issue:34567", @@ -5828,240 +5855,239 @@ "issue:42915|issue:43749", "issue:43742|issue:44291", "issue:43906|issue:45381", - "issue:28282|issue:29127", "issue:42175|issue:44488", + "issue:28282|issue:29127", "issue:38175|issue:44898", "issue:43582|issue:43782", "issue:43404|issue:45987", - "issue:44737|issue:45588", "issue:43540|issue:45823" ], - "prepared_review_unit_hash": "304583ffdbe1bc6f50eed93593b0e441a157d11450558a90e610f4d89689af26", + "prepared_review_unit_hash": "97a25fd0508cfd78976e09eccf492e73a38df24fdf91ed86433874c03a9df201", "component_budget": { - "node_count": 427, - "item_count": 427, - "soft_pair_count": 1576, - "serialized_chars": 816182, - "estimated_input_tokens": 204046, - "estimated_eval_tokens": 408348 + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 32, - "serialized_chars": 37339, - "estimated_input_tokens": 9335, - "estimated_eval_tokens": 18926 + "serialized_chars": 37373, + "estimated_input_tokens": 9344, + "estimated_eval_tokens": 18944 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "Mostly unrelated issue reports; only one soft duplicate pair looks like the same underlying bug: a Qwen video-input regression involving wrong vision position ids. No PRs are present.", - "confidence": 0.79, - "canonical_issue_reason": "Issue 44479 is the broader regression report spanning multiple Qwen VL variants; 45381 appears to be a narrower instance of the same video position-id problem.", + "summary": "This cluster is mostly heterogeneous issue reports spanning unrelated models and subsystems. I found one plausible duplicate pair around Qwen video-input regression; the rest look like separate bugs and should not be merged.", + "confidence": 0.78, + "canonical_issue_reason": "issue:44479 is the best canonical issue because it describes the broader video-input regression across multiple Qwen model variants, while the paired report is a narrower symptom/reproduction.", "canonical_pr_reason": null, - "best_issue_reason": "44479 is the best cluster anchor because it is the most general and user-facing report of the shared Qwen video regression.", + "best_issue_reason": "issue:44479 is the strongest representative of the only likely duplicate thread in the cluster and has the broadest, most actionable bug description.", "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:44568", + "right": "issue:45488", + "accept": false, + "reason": "Both are tokenizer regressions, but they affect different tokenizer code paths and different models; not the same underlying bug." + }, { "left": "issue:42915", "right": "issue:43742", "accept": false, - "reason": "Different models and failures: Qwen3Moe FP8 conversion vs MobileLLM key loading." + "reason": "Different failure modes and model-loading paths: FineGrainedFP8Config vs a key error for MobileLLM." }, { "left": "issue:44291", "right": "issue:45042", "accept": false, - "reason": "Unrelated loader/init_empty_weights error versus PIL backend torchvision dependency issue." + "reason": "Unrelated subsystems: init_empty_weights loading bug vs PIL image processor torchvision dependency." + }, + { + "left": "issue:43697", + "right": "issue:45357", + "accept": false, + "reason": "Different bugs in different areas: RTDetrV2 output drift vs incorrect visual encoder keys on save_pretrained." }, { "left": "issue:42915", "right": "issue:44821", "accept": false, - "reason": "No shared code path; one is Qwen3Moe FP8 loading, the other AutoImageProcessor-from-URL loading." + "reason": "These are unrelated model/config issues; no evidence they share the same code-path or root cause." }, { "left": "issue:41720", "right": "issue:43540", "accept": false, - "reason": "Both are runtime failures, but the affected models and likely code paths differ; not enough evidence of the same bug." + "reason": "Both mention Qwen and runtime failures, but one is auto device mapping CUDA assert and the other is video-input handling in Qwen3OmniMoe." }, { - "left": "issue:43697", - "right": "issue:45357", + "left": "issue:44987", + "right": "issue:45538", "accept": false, - "reason": "Different problems: RTDetrV2 output changes vs save_pretrained writing incorrect visual encoder keys." + "reason": "Different problems: one is a model load failure, the other is tokenizer max_length behavior." }, { "left": "issue:44479", "right": "issue:45381", "accept": true, - "reason": "Both describe Qwen VL video-input regressions, and 45381\u2019s wrong vision_position_ids matches the broader video regression in 44479." + "reason": "Same underlying Qwen video-input regression; 45381 appears to be a narrower reproduction of the broader v5.3.0 video-position bug described in 44479." }, { "left": "issue:43452", "right": "issue:43582", "accept": false, - "reason": "Different subsystems entirely: gguf tokenizer/model loading versus Apple Silicon allocator warmup TypeError." - }, - { - "left": "issue:44987", - "right": "issue:45538", - "accept": false, - "reason": "Unrelated failures: model loading for a specific repo versus tokenizer max-length behavior." + "reason": "Unrelated: gguf_file tokenizer/model loading breakage vs an Apple Silicon allocator warmup TypeError." }, { "left": "issue:43927", "right": "issue:44912", "accept": false, - "reason": "Different bugs: custom token IDs lost on save/load versus MXFP4 quantization fallback." + "reason": "Different issues: custom token IDs lost on save/load vs MXFP4 quantization fallback during model loading." }, { "left": "issue:42915", "right": "issue:45538", "accept": false, - "reason": "Qwen3Moe FP8 loading and CLIPTokenizer max_length are unrelated." + "reason": "Completely different topics: FP8 config loading for Qwen3Moe vs CLIPTokenizer max_length handling." }, { "left": "issue:33357", "right": "issue:34567", "accept": false, - "reason": "Mac bus error on CLIP model loading is unrelated to TrainerState token-count tracking." + "reason": "One is a MacOS bus error on CLIP model loading; the other is TrainerState token counting not updating." }, { "left": "issue:43540", "right": "issue:44479", "accept": false, - "reason": "Both involve video inputs, but the model families and reported failures differ; not clearly the same underlying bug." + "reason": "Both are video-related, but they involve different Qwen model families and different failure modes; not the same bug." }, { "left": "issue:43540", "right": "issue:44560", "accept": false, - "reason": "Different Qwen video-input errors with different symptoms and model variants." + "reason": "Same broad model family and video inputs, but the reported errors and likely code paths differ." }, { "left": "issue:42175", "right": "issue:43824", "accept": false, - "reason": "Package/backend installation issue is unrelated to missing Qwen2_5_VL symbol imports." + "reason": "Both are import/backend availability problems, but one is TensorFlow packaging and the other is a missing Qwen class export." }, { "left": "issue:43317", "right": "issue:45538", "accept": false, - "reason": "Dequantized model offload failure is unrelated to CLIPTokenizer model_max_length behavior." + "reason": "Different areas: device-map/offload dequantization failure vs tokenizer max_length behavior." }, { "left": "issue:44987", "right": "issue:45907", "accept": false, - "reason": "Repo loading regression versus list_repo_files revision forwarding are unrelated." + "reason": "Unrelated: model loading failure vs a helper function not forwarding the revision argument." }, { "left": "issue:44062", "right": "issue:44991", "accept": false, - "reason": "Tokenizer AddedToken argument collision is not the same as est-roberta tokenizer loading failure." + "reason": "Both concern tokenizers, but one is an AddedToken argument collision and the other is a specific model tokenizer load failure." }, { "left": "issue:43257", "right": "issue:43756", "accept": false, - "reason": "Qwen3 MOE weight conversion and Smollm3 RoPE-layer mismatch are different model-specific issues." + "reason": "Different Qwen3Moe loading/conversion issue vs a RoPE-layer configuration bug in Smollm3." }, { "left": "issue:24643", "right": "issue:29159", "accept": false, - "reason": "DeepSpeed weight shape error and tokenizer slow/fast inconsistency are unrelated." + "reason": "No shared root cause: DeepSpeed training weight-shape error vs slow/fast tokenizer inconsistency." }, { "left": "issue:45137", "right": "issue:45405", "accept": false, - "reason": "DeepSpeed ZeRO3 deque error is unrelated to PEFT minimum-version packaging." + "reason": "Totally unrelated: ZeRO3 deque underflow vs an unreleased PEFT version bump." }, { "left": "issue:28282", "right": "issue:34689", "accept": false, - "reason": "Missing PyTorch import error is unrelated to Llama 3.2 Vision model loading regression." + "reason": "Different import/load errors for different models and environment states; not the same issue." }, { "left": "issue:43577", "right": "issue:45072", "accept": false, - "reason": "BLIP2 dtype retention and SwitchTransformers/TimmWrapperModel bfloat16 mismatches are different bugs." + "reason": "Different dtype-related bugs affecting different model classes and code paths." }, { "left": "issue:42915", "right": "issue:43749", "accept": false, - "reason": "Qwen3Moe FP8 loading and FSDP CPU RAM efficient loading are unrelated." + "reason": "Different topics: Qwen3Moe FP8 config failure vs FSDP CPU RAM efficient loading regression." }, { "left": "issue:43742", "right": "issue:44291", "accept": false, - "reason": "Different loading failures with different root causes and models." + "reason": "Different mechanisms and subsystems; no evidence of a common underlying bug." }, { "left": "issue:43906", "right": "issue:45381", "accept": false, - "reason": "A reproduction of another issue is not the same as the Qwen video-position-id regression." + "reason": "The second is a Qwen video-position bug; the first is only an isolated reproduction reference and does not establish the same root cause." }, { - "left": "issue:28282", - "right": "issue:29127", + "left": "issue:42175", + "right": "issue:44488", "accept": false, - "reason": "PyTorch import failure and LayoutLMv3 box-clarity error are unrelated." + "reason": "Packaging/backend installation issue vs a specific model-loading regression; not duplicates." }, { - "left": "issue:42175", - "right": "issue:44488", + "left": "issue:28282", + "right": "issue:29127", "accept": false, - "reason": "Backend installation omission and sleng-bert loading failure are not the same bug." + "reason": "Different bugs: missing PyTorch import vs a LayoutLMv3 box-content error message clarity issue." }, { "left": "issue:38175", "right": "issue:44898", "accept": false, - "reason": "Unexpected zero probabilities in SigLIP2 and Perceiver non-default-resolution failure are different model-specific issues." + "reason": "Different models and behaviors: unexpected zero probabilities vs Perceiver non-default resolution failure." }, { "left": "issue:43582", "right": "issue:43782", "accept": false, - "reason": "Apple Silicon allocator TypeError and Qwen3VL weight_only error are unrelated." + "reason": "Apple Silicon allocator warmup TypeError is unrelated to Qwen3VL weight_only loading failure." }, { "left": "issue:43404", "right": "issue:45987", "accept": false, - "reason": "lm_head weight tying in Mistral3 is unrelated to StaticCache.get_seq_length returning a tensor." - }, - { - "left": "issue:44737", - "right": "issue:45588", - "accept": false, - "reason": "XLNet CPU arange issue and flash-attention s_aux=None crash are unrelated." + "reason": "Different model families and different bugs: lm_head weight tying vs StaticCache sequence length return type." }, { "left": "issue:43540", "right": "issue:45823", "accept": false, - "reason": "Different device/input failures in different Qwen/Gemma models; not the same code-path bug." + "reason": "Different device/runtime bugs in different models; not the same code-path problem." } ] }, "evaluator_result": { "accept": true, - "feedback": "Grounded overall. The only accepted soft edge (44479/45381) is phrased conservatively enough given the titles, and the false rejections are reasonable. Minor caveat: \"user-facing\" is not directly evidenced in the packet, so that wording is a bit stronger than the available data supports." + "feedback": "The summary is grounded in the packet and stays conservative. The only plausible duplicate pair supported by the titles is 44479/45381, and the rest of the soft-edge rejections are appropriately cautious given the available evidence. No overstatement stands out." }, "evaluator_used": true, "retried": false, @@ -6074,13 +6100,13 @@ "component_index": 1, "component_count": 17, "review_unit_index": 19, - "review_unit_count": 49, + "review_unit_count": 50, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-427", + "cluster_id": "cluster-10000-428", "nodes": [ "issue:24643", "issue:29127", @@ -6098,7 +6124,6 @@ "issue:43742", "issue:43749", "issue:43756", - "issue:43819", "issue:43994", "issue:44062", "issue:44291", @@ -6106,8 +6131,8 @@ "issue:44514", "issue:44589", "issue:44625", + "issue:44737", "issue:44792", - "issue:44811", "issue:44821", "issue:44843", "issue:44877", @@ -6126,15 +6151,17 @@ "issue:45381", "issue:45405", "issue:45479", + "issue:45588", "issue:45600", "issue:45920", "issue:46121", "issue:46170" ], "soft_pairs": [ + "issue:44737|issue:45588", "issue:43526|issue:45479", - "issue:43493|issue:43994", "issue:43756|issue:45325", + "issue:43493|issue:43994", "issue:34567|issue:36331", "issue:44913|issue:46121", "issue:44291|issue:44821", @@ -6148,201 +6175,200 @@ "issue:45405|issue:45600", "issue:43531|issue:44821", "issue:44625|issue:44877", + "issue:43697|issue:45216", + "issue:43720|issue:44898", "issue:24643|issue:29127", "issue:44514|issue:45381", - "issue:43720|issue:44898", - "issue:43697|issue:45216", "issue:45372|issue:46170", - "issue:43688|issue:43749", "issue:41762|issue:44936", - "issue:44514|issue:45290", + "issue:43688|issue:43749", "issue:43452|issue:44843", - "issue:43819|issue:44811" + "issue:44514|issue:45290" ], - "prepared_review_unit_hash": "2c7e740c06f2cca79080ec9c50234c846021d88572d5dfb89923d768e4a748dc", + "prepared_review_unit_hash": "ecd87d52674b92ce5117d09baf5021e8c7696e2b05a3d3a1c511fdb900195d75", "component_budget": { - "node_count": 427, - "item_count": 427, - "soft_pair_count": 1576, - "serialized_chars": 816182, - "estimated_input_tokens": 204046, - "estimated_eval_tokens": 408348 + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 26, - "serialized_chars": 35986, - "estimated_input_tokens": 8997, - "estimated_eval_tokens": 18250 + "serialized_chars": 36040, + "estimated_input_tokens": 9010, + "estimated_eval_tokens": 18276 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "These issues are mostly unrelated regression reports across different models/components; none of the soft pairs look like true duplicates of the same underlying bug or change.", - "confidence": 0.91, - "canonical_issue_reason": "issue:45920 is the strongest representative overall: it documents a broad, concrete regression affecting multiple model families, with clear repros and active discussion.", + "summary": "This cluster is highly heterogeneous: the soft pairs mostly share only broad themes like loading, configs, or trainer behavior, but they point to different models/code paths and should not be merged as duplicates.", + "confidence": 0.94, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "issue:45920 has the clearest symptoms and widest impact, so it is the best standalone issue in this set.", + "best_issue_reason": null, "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:44737", + "right": "issue:45588", + "accept": false, + "reason": "XLNet CPU placement bug vs. flash-attention sinkless-model AttributeError; unrelated subsystems and failure modes." + }, { "left": "issue:43526", "right": "issue:45479", "accept": false, - "reason": "Both involve labels/classification, but one is a Beit image processor bug and the other is a sequence-classification loss issue; different code paths and symptoms." + "reason": "BEiT fast processor label reduction bug vs. sequence-classification loss degeneracy; different APIs and code paths." }, { - "left": "issue:43493", - "right": "issue:43994", + "left": "issue:43756", + "right": "issue:45325", "accept": false, - "reason": "Both mention SigLIP2, but one is a HF-vs-JAX implementation discrepancy and the other is an AutoModel/pipeline output problem; not the same defect." + "reason": "Smollm3 RoPE-layer truncation vs. Qwen2.5-VL position-id scaling; both are RoPE-related but for different models and bugs." }, { - "left": "issue:43756", - "right": "issue:45325", + "left": "issue:43493", + "right": "issue:43994", "accept": false, - "reason": "Both touch RoPE/positioning, but they concern different models and different bugs: Smollm3 layer dropping vs Qwen2.5-VL temporal position scaling." + "reason": "Both are SigLIP2-related, but one is an HF-vs-JAX discrepancy and the other is a checkpoint/pipeline output issue; not clearly the same concrete bug." }, { "left": "issue:34567", "right": "issue:36331", "accept": false, - "reason": "Trainer state token counting and custom compute_loss signature are separate trainer API issues with different failure modes." + "reason": "TrainerState token counting update bug vs. custom compute_loss signature break; different trainer internals." }, { "left": "issue:44913", "right": "issue:46121", "accept": false, - "reason": "One is a config value not persisting on reload; the other is a type error in rope validation for JSON-loaded lists. Related theme, different bug." + "reason": "Config reload default reset vs. TypeError converting JSON-loaded rope ignore keys; distinct config handling problems." }, { "left": "issue:44291", "right": "issue:44821", "accept": false, - "reason": "An init_empty_weights argument error and loading an image processor from URL are unrelated loading paths." + "reason": "init_empty_weights/_is_hf_initialized loading bug vs. loading AutoImageProcessor from URL; unrelated failures." }, { "left": "issue:44062", "right": "issue:45081", "accept": false, - "reason": "Both are tokenizer-related crashes, but one is AddedToken construction and the other is a Mistral regex patch failure; not the same root cause." + "reason": "AddedToken duplicate-special kwarg error vs. mistral regex/backend_tokenizer crash; both tokenizer-related but different code paths." }, { "left": "issue:44792", "right": "issue:44964", "accept": false, - "reason": "Both involve multimodal models, but one is a janus test failure and the other is a Phi-4 model loading regression." + "reason": "Janus image-generation test failure vs. Phi-4 multimodal load failure; different models and symptoms." }, { "left": "issue:45357", "right": "issue:45920", "accept": false, - "reason": "Qwen3.5 save_pretrained key mapping and AutoTokenizer token-ID corruption are different regression classes." + "reason": "Qwen3.5 save_pretrained key mismatch vs. AutoTokenizer wrong token IDs across many models; different bugs and components." }, { "left": "issue:45042", "right": "issue:45081", "accept": false, - "reason": "Image processor torchvision dependency and Mistral regex/tokenizer crash are unrelated." + "reason": "PIL backend torchvision dependency regression vs. mistral regex crash; no shared underlying fix." }, { "left": "issue:43452", "right": "issue:44488", "accept": false, - "reason": "GGUF/AutoTokenizer-AutoModel loading failure and sleng-bert model loading failure do not indicate the same underlying bug." + "reason": "gguf_file loader regression vs. specific sleng-bert loading failure; too broad and not same bug." }, { "left": "issue:43742", "right": "issue:44589", "accept": false, - "reason": "A model key error and a missing Float8 storage type error are separate loading/runtime problems." + "reason": "MobileLLM key lookup failure vs. Float8 storage lookup error; both are load-time issues but unrelated." }, { "left": "issue:43643", "right": "issue:45375", "accept": false, - "reason": "Both are config-field issues, but one is generic trust_remote_code behavior and the other is a specific Qwen3_5MoeVisionConfig field dropped by strict parsing." + "reason": "AutoConfig trust_remote_code field loss vs. Qwen3_5MoeVisionConfig strict-field drop; different mechanisms and models." }, { "left": "issue:45405", "right": "issue:45600", "accept": false, - "reason": "A version pin bump and removed config references in auto_mappings are different CI/breakage causes." + "reason": "Version pin mismatch vs. removed config references in auto_mappings; unrelated maintenance issues." }, { "left": "issue:43531", "right": "issue:44821", "accept": false, - "reason": "Sliding-window behavior in Qwen3-MoE and AutoImageProcessor URL loading are unrelated." + "reason": "Qwen3-MoE sliding_window behavior vs. AutoImageProcessor URL loading; no shared code path." }, { "left": "issue:44625", "right": "issue:44877", "accept": false, - "reason": "Num_labels propagation and strict config rejection are different config-handling regressions." - }, - { - "left": "issue:24643", - "right": "issue:29127", - "accept": false, - "reason": "DeepSpeed training weight-shape runtime error and a LayoutLMv3 error-message clarity issue are not the same bug." + "reason": "num_labels propagation bug in Qwen3.5 vs. strict config blocking granite_speech loading; different config surfaces." }, { - "left": "issue:44514", - "right": "issue:45381", + "left": "issue:43697", + "right": "issue:45216", "accept": false, - "reason": "Both are Qwen2.5-VL issues, but one is chat-template batching and the other is video position IDs; different code paths." + "reason": "RTDetrV2 output regression vs. Qwen3.5 save_pretrained regression; different model families and behavior." }, { "left": "issue:43720", "right": "issue:44898", "accept": false, - "reason": "BitNet packed-weight loading and Perceiver interpolation failures are unrelated model-specific issues." + "reason": "BitNet packed-weight loading bug vs. Perceiver non-default-res interpolation failure; unrelated model-loading problems." }, { - "left": "issue:43697", - "right": "issue:45216", + "left": "issue:24643", + "right": "issue:29127", "accept": false, - "reason": "RTDetrV2 output drift and Qwen3.5 save_pretrained corruption are distinct regressions." + "reason": "DeepSpeed training weight-shape runtime error vs. LayoutLMv3 error-message clarity issue; one is a bug, the other a UX/reporting issue." }, { - "left": "issue:45372", - "right": "issue:46170", + "left": "issue:44514", + "right": "issue:45381", "accept": false, - "reason": "Gemma 4 processor loading fails due to mistral_common import changes, while the Windows ROCm issue is about torch.distributed dependency handling." + "reason": "Qwen2.5-VL chat-template batch crash vs. Qwen2.5-VL video position-id bug; same family but different inputs and code paths." }, { - "left": "issue:43688", - "right": "issue:43749", + "left": "issue:45372", + "right": "issue:46170", "accept": false, - "reason": "Auxiliary loss normalization and FSDP CPU RAM-efficient loading are different training/runtime concerns." + "reason": "Gemma 4 processor import failure vs. Windows/ROCm torch.distributed dependency issue; unrelated platform/library problems." }, { "left": "issue:41762", "right": "issue:44936", "accept": false, - "reason": "Gemma3 ZeRO-3 loading failure and trainer.evaluate after train are unrelated trainer/model-loading bugs." + "reason": "Gemma3 ZeRO-3 loading IndexError vs. trainer.evaluate after train; different stack traces and lifecycle stages." }, { - "left": "issue:44514", - "right": "issue:45290", + "left": "issue:43688", + "right": "issue:43749", "accept": false, - "reason": "Both involve apply_chat_template, but they fail on different inputs and in different branches of the templating logic." + "reason": "Aux-loss normalization bug vs. FSDP CPU RAM-efficient loading regression; different model/training subsystems." }, { "left": "issue:43452", "right": "issue:44843", "accept": false, - "reason": "GGUF loading breakage and unconditional model_info() offline failure are separate tokenizer-loading regressions." + "reason": "General gguf loading breakage vs. mistral_regex offline-model_info call; both loading-related but not the same issue." }, { - "left": "issue:43819", - "right": "issue:44811", + "left": "issue:44514", + "right": "issue:45290", "accept": false, - "reason": "DAC latent decoding mismatch and Whisper batch_decode skip_special_tokens behavior are unrelated." + "reason": "apply_chat_template batched-padding crash vs. tool-call/no-content assistant crash; both chat-template issues but distinct triggers and fixes." } ] }, @@ -6358,13 +6384,13 @@ "component_index": 1, "component_count": 17, "review_unit_index": 20, - "review_unit_count": 49, + "review_unit_count": 50, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-427", + "cluster_id": "cluster-10000-428", "nodes": [ "issue:28282", "issue:29159", @@ -6380,13 +6406,13 @@ "issue:43531", "issue:43550", "issue:43577", - "issue:43643", "issue:43646", "issue:43653", "issue:43688", "issue:43697", "issue:43742", "issue:43749", + "issue:43819", "issue:43906", "issue:43927", "issue:44361", @@ -6395,10 +6421,10 @@ "issue:44466", "issue:44484", "issue:44493", - "issue:44561", "issue:44568", "issue:44704", "issue:44743", + "issue:44811", "issue:44843", "issue:44936", "issue:44977", @@ -6417,244 +6443,244 @@ ], "soft_pairs": [ "issue:44466|issue:45005", - "issue:44493|issue:46050", "issue:45542|issue:46178", "issue:43452|issue:43577", - "issue:44484|issue:44704", - "issue:44936|issue:45823", + "issue:44493|issue:46050", + "issue:43819|issue:44811", "issue:30064|issue:33666", + "issue:44936|issue:45823", "issue:43295|issue:45356", + "issue:44484|issue:44704", "issue:43653|issue:44361", "issue:28282|issue:32090", "issue:44361|issue:44442", + "issue:41720|issue:43906", "issue:28282|issue:33357", "issue:44464|issue:44743", - "issue:41720|issue:43906", "issue:38175|issue:43295", "issue:28282|issue:35141", "issue:45491|issue:46002", "issue:43653|issue:44568", - "issue:43688|issue:45704", "issue:43577|issue:43697", + "issue:43688|issue:45704", "issue:44843|issue:44991", - "issue:43646|issue:43749", "issue:43688|issue:45440", + "issue:43646|issue:43749", "issue:29159|issue:33357", "issue:43688|issue:45701", "issue:43653|issue:43927", "issue:43531|issue:43742", "issue:44464|issue:44977", - "issue:43550|issue:43927", - "issue:43643|issue:44561" + "issue:43550|issue:43927" ], - "prepared_review_unit_hash": "aa1a9f59204fad9f466ea41c8ef87289b303689a2098e89764af9bb299e0d6f1", + "prepared_review_unit_hash": "81a19ac26b882e5ca85bd7b3a7bf754cd53b3116800a85c2ca28687eda58d247", "component_budget": { - "node_count": 427, - "item_count": 427, - "soft_pair_count": 1576, - "serialized_chars": 816182, - "estimated_input_tokens": 204046, - "estimated_eval_tokens": 408348 + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 30, - "serialized_chars": 36734, - "estimated_input_tokens": 9184, - "estimated_eval_tokens": 18624 + "serialized_chars": 36719, + "estimated_input_tokens": 9180, + "estimated_eval_tokens": 18616 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is heterogeneous: the soft edges mostly connect thematically similar but distinct issues (tokenizers, loading, training, generation, backend errors), so I reject all candidate duplicate links.", - "confidence": 0.91, - "canonical_issue_reason": "No single canonical issue stands out. The items split into several unrelated subgroups, so there is no safe duplicate master for the whole cluster.", + "summary": "This set is highly heterogeneous: it mixes unrelated tokenizer regressions, generation bugs, training errors, backend/config issues, and model-loading problems. I do not see any convincing duplicate pairs; most similarities are only at the subsystem level. Issue 45701 is the broadest umbrella-style issue in the set, but the cluster does not read like one single bug family.", + "confidence": 0.84, + "canonical_issue_reason": "Issue 45701 is the most umbrella-like and currently open item; it best matches the broad version-dependent tokenization regression theme that appears in a few nearby issues.", "canonical_pr_reason": null, - "best_issue_reason": "If one issue must serve as a loose anchor, issue 43653 is the most central tokenizer-related report, but it still only represents one subtheme and should not absorb the rest as duplicates.", + "best_issue_reason": "Issue 45701 is the best anchor for triage because it is broad, active, and representative of the only recurring theme that shows up across multiple items (tokenization behavior changing across versions).", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44466", "right": "issue:45005", "accept": false, - "reason": "Both mention tied weights, but one is device-dependent lm_head serialization and the other is a broader translation-model tied-weights issue; not the same bug." - }, - { - "left": "issue:44493", - "right": "issue:46050", - "accept": false, - "reason": "Unexpected position-id keys and quantized save/load breakage are different failure modes." + "reason": "Both mention tied weights, but one is a concrete serialization/device-dependent lm_head.weight issue while the other is a broader translation-model v5 regression." }, { "left": "issue:45542", "right": "issue:46178", "accept": false, - "reason": "Both are backend-related, but one is a TensorBoard-without-TF backend error and the other is a tensorflow_text backend mapping error." + "reason": "Both are backend/TensorFlow related, but the failure modes differ: missing TensorFlow with tensorboard vs tensorflow_text not being in BACKENDS_MAPPING." }, { "left": "issue:43452", "right": "issue:43577", "accept": false, - "reason": "GGUF loading failures and Blip2 dtype mismatches are unrelated code paths." + "reason": "Different subsystems: gguf tokenizer/model loading vs BLIP2 dtype propagation for qformer/model." }, { - "left": "issue:44484", - "right": "issue:44704", + "left": "issue:44493", + "right": "issue:46050", "accept": false, - "reason": "A default max_shard_size question and AutoProcessor kwarg propagation are different issues." + "reason": "Different bugs: unexpected position-id key handling versus quantized model save/load corruption after sinq." }, { - "left": "issue:44936", - "right": "issue:45823", + "left": "issue:43819", + "right": "issue:44811", "accept": false, - "reason": "trainer.evaluate() after train() and a Gemma4 auto device-map mismatch are unrelated bugs." + "reason": "DAC latent conversion with missing STE is unrelated to Whisper batch_decode ignoring skip_special_tokens." }, { "left": "issue:30064", "right": "issue:33666", "accept": false, - "reason": "Void segmentation-map processing and Qwen2-VL multi-GPU training are different subsystems and symptoms." + "reason": "Image segmentation map processing and Qwen2-VL multi-GPU training are unrelated." + }, + { + "left": "issue:44936", + "right": "issue:45823", + "accept": false, + "reason": "Trainer.evaluate() after train() is a training-loop bug, not the Gemma4 device-mismatch forward issue." }, { "left": "issue:43295", "right": "issue:45356", "accept": false, - "reason": "Both are tokenizer regressions, but one is processor/tokenizer API breakage and the other is codec handling plus a warning regression for Kimi-K2.5." + "reason": "Both are tokenizer/processor regressions, but one is processor.tokenizer/images passthrough and the other is Kimi codec/fix_mistral_regex behavior." + }, + { + "left": "issue:44484", + "right": "issue:44704", + "accept": false, + "reason": "Max shard size default sizing and AutoProcessor kwargs forwarding are unrelated." }, { "left": "issue:43653", "right": "issue:44361", "accept": false, - "reason": "BigBirdTokenizer special-token registration and MLukeTokenizer AttributeError are different tokenizer bugs." + "reason": "Both are tokenizer issues, but the concrete failures differ: BigBird special-token decode vs MLukeTokenizer AttributeError." }, { "left": "issue:28282", "right": "issue:32090", "accept": false, - "reason": "Missing PyTorch ImportError and NoneType broadcast errors in Trainer are unrelated." + "reason": "Missing PyTorch import error and Trainer _gpu_broadcast_one NoneType error are unrelated." }, { "left": "issue:44361", "right": "issue:44442", "accept": false, - "reason": "Both involve tokenizers, but the concrete failures differ: MLukeTokenizer AttributeError vs AutoTokenizer failing to load a specific tokenizer class." + "reason": "Both concern tokenizers, but they hit different tokenizer classes with different loading/runtime failures." + }, + { + "left": "issue:41720", + "right": "issue:43906", + "accept": false, + "reason": "Qwen3 auto device-map cudaErrorAssert is not the same as the separate isolated reproduction of issue 38071." }, { "left": "issue:28282", "right": "issue:33357", "accept": false, - "reason": "PyTorch-not-found ImportError and a MacOS bus error with CLIP are unrelated." + "reason": "PyTorch import failure and MacOS bus error with a CLIP model are unrelated." }, { "left": "issue:44464", "right": "issue:44743", "accept": false, - "reason": "Chunked generation with compiled forward and Qwen3 recurrent-state reset with cache are different generation-path bugs." - }, - { - "left": "issue:41720", - "right": "issue:43906", - "accept": false, - "reason": "Qwen3 auto device mapping cuda assert is unrelated to the isolated reproduction of another issue." + "reason": "Both involve generation/caching, but chunked generation with compiled forward is a different code path from recurrent-state reset with seq_len>1." }, { "left": "issue:38175", "right": "issue:43295", "accept": false, - "reason": "Zero probabilities in SigLIP2 and a processor.tokenizer regression are unrelated." + "reason": "Siglip2 zero probabilities are unrelated to the processor.tokenizer regression." }, { "left": "issue:28282", "right": "issue:35141", "accept": false, - "reason": "Import-time missing PyTorch and embedding reinitialization on resize are different issues." + "reason": "ImportError for missing PyTorch is unrelated to post_init reinitializing output embeddings after resize." }, { "left": "issue:45491", "right": "issue:46002", "accept": false, - "reason": "Both mention NaNs, but the former is a Gemma3 sliding-window GPU batching bug and the latter is a classification-head init/checkpoint issue." + "reason": "Both mention NaNs, but one is mixed-length sliding-window attention on Gemma3 and the other is classification-head NaNs after checkout/init." }, { "left": "issue:43653", "right": "issue:44568", "accept": false, - "reason": "Both are tokenizer special-token problems, but they affect different models and different behaviors." + "reason": "Both are special-token/tokenizer issues, but they are different models and different behaviors (mask token registration vs BOS/EOS insertion)." }, { - "left": "issue:43688", - "right": "issue:45704", + "left": "issue:43577", + "right": "issue:43697", "accept": false, - "reason": "Aux-loss normalization and a T5 RMSNorm memory leak are unrelated." + "reason": "Dtype staying float32 in BLIP2 is unrelated to RTDetrV2 output divergence in v5." }, { - "left": "issue:43577", - "right": "issue:43697", + "left": "issue:43688", + "right": "issue:45704", "accept": false, - "reason": "Blip2 dtype loading and RTDetrV2 output drift in v5 are different model behaviors." + "reason": "Auxiliary-loss normalization in MoE models is unrelated to T5 silently using apex FusedRMSNorm and leaking memory." }, { "left": "issue:44843", "right": "issue:44991", "accept": false, - "reason": "Offline model_info calls in tokenizer patching and a model-specific tokenizer load failure are not the same bug." + "reason": "Both are tokenizer-loading regressions, but one is an offline model_info call bug and the other is a specific tokenizer load failure for est-roberta." }, { - "left": "issue:43646", - "right": "issue:43749", + "left": "issue:43688", + "right": "issue:45440", "accept": false, - "reason": "Custom model initialization breakage and FSDP CPU RAM efficient loading are different code paths." + "reason": "Both touch MoE models, but auxiliary-loss normalization and DeepseekV3MoE divergence are not the same concrete bug." }, { - "left": "issue:43688", - "right": "issue:45440", + "left": "issue:43646", + "right": "issue:43749", "accept": false, - "reason": "Incorrect auxiliary-loss normalization is not the same as DeepseekV3MoE diverging from the remote implementation." + "reason": "Custom model initialization breakage and FSDP CPU RAM efficient loading are different failure modes." }, { "left": "issue:29159", "right": "issue:33357", "accept": false, - "reason": "Slow/fast tokenizer inconsistency and a MacOS bus error are unrelated." + "reason": "Slow/fast tokenizer inconsistency is unrelated to the MacOS bus error report." }, { "left": "issue:43688", "right": "issue:45701", "accept": false, - "reason": "Aux-loss normalization and a broad tokenization-change report are not the same underlying bug." + "reason": "MoE auxiliary-loss normalization is unrelated to version-dependent tokenization behavior." }, { "left": "issue:43653", "right": "issue:43927", "accept": false, - "reason": "A tokenizer special-token registration bug and DiaConfig losing custom token IDs are different failure modes." + "reason": "BigBird special-token decode and DiaConfig losing custom token IDs after save/load are different tokenizer/config bugs." }, { "left": "issue:43531", "right": "issue:43742", "accept": false, - "reason": "Qwen3-MoE sliding-window handling and a MobileLLM-125M load key error are unrelated." + "reason": "Qwen3-MoE sliding_window behavior and MobileLLM-125M loading key error are unrelated." }, { "left": "issue:44464", "right": "issue:44977", "accept": false, - "reason": "Both involve generation, but compiled-forward chunking and flash-attention on Qwen3.5 are different concrete problems." + "reason": "Both mention generation, but compiled-forward chunked generation and flash-attention failure on Qwen3.5 are different code-path problems." }, { "left": "issue:43550", "right": "issue:43927", "accept": false, - "reason": "torch.compile with SDPA on Bamba-9B-v2 and DiaConfig custom-token-ID loss are unrelated." - }, - { - "left": "issue:43643", - "right": "issue:44561", - "accept": false, - "reason": "Missing fields from trust_remote_code AutoConfig and removal of is_torch_fx_available break different parts of the remote-code path." + "reason": "torch.compile+SDPA failure in Bamba and DiaConfig custom token ID loss are unrelated." } ] }, @@ -6670,13 +6696,13 @@ "component_index": 1, "component_count": 17, "review_unit_index": 21, - "review_unit_count": 49, + "review_unit_count": 50, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-427", + "cluster_id": "cluster-10000-428", "nodes": [ "issue:28218", "issue:32090", @@ -6685,7 +6711,6 @@ "issue:34567", "issue:36010", "issue:36331", - "issue:39692", "issue:41720", "issue:43122", "issue:43262", @@ -6699,6 +6724,7 @@ "issue:43572", "issue:43577", "issue:43606", + "issue:43643", "issue:43688", "issue:43720", "issue:43742", @@ -6706,281 +6732,276 @@ "issue:43825", "issue:43906", "issue:43927", + "issue:43994", "issue:44361", "issue:44451", "issue:44464", "issue:44466", "issue:44479", "issue:44554", - "issue:44568", + "issue:44561", "issue:44610", "issue:44625", "issue:44821", "issue:44849", "issue:44913", - "issue:44964", "issue:45357", - "issue:45405", "issue:45588", - "issue:45812", "issue:45823", "issue:46002", - "issue:46050" + "issue:46050", + "issue:46192" ], "soft_pairs": [ + "issue:43643|issue:44561", "issue:44479|issue:44625", "issue:43550|issue:44361", "issue:41720|issue:44464", - "issue:32090|issue:34567", "issue:43122|issue:43906", - "issue:43577|issue:45588", "issue:43577|issue:44554", + "issue:43577|issue:45588", + "issue:32090|issue:34567", "issue:43504|issue:43550", "issue:43504|issue:43720", "issue:43504|issue:44610", "issue:44466|issue:46050", "issue:32090|issue:33666", "issue:33666|issue:36010", - "issue:45823|issue:46050", "issue:43742|issue:44821", + "issue:45823|issue:46050", + "issue:43452|issue:44451", "issue:43720|issue:43927", "issue:43606|issue:43927", - "issue:43452|issue:44451", - "issue:43262|issue:43295", "issue:43278|issue:43688", - "issue:43720|issue:44361", - "issue:33666|issue:36331", + "issue:43994|issue:46192", + "issue:43262|issue:43295", "issue:43525|issue:43572", "issue:44625|issue:46002", + "issue:33666|issue:36331", + "issue:43720|issue:44361", "issue:43756|issue:44913", "issue:44849|issue:45357", "issue:43493|issue:43825", - "issue:28218|issue:33357", - "issue:44568|issue:45812", - "issue:44964|issue:45405", - "issue:39692|issue:43493" + "issue:28218|issue:33357" ], - "prepared_review_unit_hash": "88a5d723b281d5cf241df04d3ff742b566894af5b75b727806fb0ed2bc759b82", + "prepared_review_unit_hash": "a361b2d36e60a78c7ce317badf55be84e18c6bf988ca247fcac96a93d046b106", "component_budget": { - "node_count": 427, - "item_count": 427, - "soft_pair_count": 1576, - "serialized_chars": 816182, - "estimated_input_tokens": 204046, - "estimated_eval_tokens": 408348 + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 }, "budget": { - "node_count": 48, - "item_count": 48, - "soft_pair_count": 31, - "serialized_chars": 37105, - "estimated_input_tokens": 9277, - "estimated_eval_tokens": 18810 + "node_count": 47, + "item_count": 47, + "soft_pair_count": 30, + "serialized_chars": 36168, + "estimated_input_tokens": 9042, + "estimated_eval_tokens": 18340 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The soft-similarity links are mostly false positives: they group together issues that only share broad vocabulary (tokenizer, loading, config, trainer, multimodal, serialization) but describe different code paths and different failures. No soft edge should be merged.", - "confidence": 0.94, - "canonical_issue_reason": "No single canonical duplicate exists here; the cluster is heterogeneous. If a representative is needed, issue 43504 is the most generic load-time regression, but it is not a duplicate anchor for the rest.", + "summary": "This set is mostly heterogeneous with many unrelated bug reports. The only credible duplicate-style overlap is the pair about v5 config losing pad-token fields on reload; the rest are distinct code paths and should stay separate.", + "confidence": 0.8, + "canonical_issue_reason": "issue:43525 is the clearest representative of the pad-token config regression, with the most generic title and a direct match to the shared underlying load/reload bug.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 43504 is the broadest and least model-specific item in the set (a generic pretrained-model loading failure), so it is the best cluster representative by theme, though not a duplicate of the others.", + "best_issue_reason": "issue:43525 best captures the recurring config-field-loss problem and is the strongest anchor for the one accepted duplicate pair.", "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:43643", + "right": "issue:44561", + "accept": false, + "reason": "Both involve trust_remote_code/v5 breakage, but they describe different root causes: missing config fields vs removal of a helper used by remote-code models." + }, { "left": "issue:44479", "right": "issue:44625", "accept": false, - "reason": "Different bugs: a video-input regression across Qwen VL models vs num_labels propagation in Qwen3.5 config." + "reason": "Video-input regression and num_labels propagation are unrelated model bugs." }, { "left": "issue:43550", "right": "issue:44361", "accept": false, - "reason": "Unrelated failures: torch.compile/SDPA in Bamba vs MLukeTokenizer attribute errors." + "reason": "Different subsystems and failures: SDPA/torch.compile on Bamba vs tokenizer AttributeError on tasks." }, { "left": "issue:41720", "right": "issue:44464", "accept": false, - "reason": "Different code paths: auto device-mapping CUDA assert vs chunked generation with compiled forward." + "reason": "Qwen3 auto device mapping cuda assert and chunked generation compiled-forward inconsistency are separate issues." }, { - "left": "issue:32090", - "right": "issue:34567", + "left": "issue:43122", + "right": "issue:43906", "accept": false, - "reason": "Trainer broadcast TypeError is unrelated to TrainerState.num_input_tokens_seen not updating." + "reason": "Tokenization-version change and an isolated reproduction of an older issue are not enough to show the same underlying bug." }, { - "left": "issue:43122", - "right": "issue:43906", + "left": "issue:43577", + "right": "issue:44554", "accept": false, - "reason": "Both mention tokenization, but one is a cross-version tokenization change and the other is an isolated reproduction of a different issue." + "reason": "Different problems: dtype handling in BLIP2 loading vs an MPS attention correctness issue." }, { "left": "issue:43577", "right": "issue:45588", "accept": false, - "reason": "Different subsystems: dtype handling in Blip2 vs flash_attention crash on s_aux=None." + "reason": "BLIP2 dtype mismatch and flash_attention s_aux=None crash are unrelated." }, { - "left": "issue:43577", - "right": "issue:44554", + "left": "issue:32090", + "right": "issue:34567", "accept": false, - "reason": "Blip2 dtype mismatch is not the same as an MPS attention correctness bug." + "reason": "Trainer broadcast NoneType error and num_input_tokens_seen not updating are distinct trainer bugs." }, { "left": "issue:43504", "right": "issue:43550", "accept": false, - "reason": "Pretrained model preset loading due to a legacy field is unrelated to Bamba torch.compile SDPA behavior." + "reason": "Legacy-field model loading failure vs torch.compile SDPA failure are different code paths." }, { "left": "issue:43504", "right": "issue:43720", "accept": false, - "reason": "Both are loading-related, but one is a legacy config field and the other is BitNet packed-weight unpacking during accelerate loading." + "reason": "Both are loading-related, but one is a legacy preset field issue and the other is about unpacking packed weights during accelerate loading." }, { "left": "issue:43504", "right": "issue:44610", "accept": false, - "reason": "Different failure modes: legacy field loading vs processor producing the wrong input size for OmDet-Turbo." + "reason": "Different model families and failures: legacy preset loading vs processor size mismatch." }, { "left": "issue:44466", "right": "issue:46050", "accept": false, - "reason": "One is tied-weight serialization varying by device; the other is quantized model save/load breakage after sinq." + "reason": "Both mention save/load behavior, but one is lm_head serialization inconsistency and the other is quantized sinq model restore breakage." }, { "left": "issue:32090", "right": "issue:33666", "accept": false, - "reason": "Trainer GPU broadcast TypeError does not match Qwen2-VL multi-GPU training behavior." + "reason": "Trainer GPU broadcast failure and Qwen2-VL multi-GPU training are not the same bug." }, { "left": "issue:33666", "right": "issue:36010", "accept": false, - "reason": "Multi-GPU training for Qwen2-VL is unrelated to GenerationMixin import failures." + "reason": "Multi-GPU training and GenerationMixin import failure are unrelated." + }, + { + "left": "issue:43742", + "right": "issue:44821", + "accept": false, + "reason": "KeyError loading one model and inability to load AutoImageProcessor from URL are different loading issues." }, { "left": "issue:45823", "right": "issue:46050", "accept": false, - "reason": "Device-map forward mismatch in Gemma4 PLE is not the same as quantized save/load corruption." + "reason": "Device-map forward mismatch and sinq save/load breakage are unrelated." }, { - "left": "issue:43742", - "right": "issue:44821", + "left": "issue:43452", + "right": "issue:44451", "accept": false, - "reason": "Different loading errors: MobileLLM key mismatch vs AutoImageProcessor loading from URL." + "reason": "Generic tokenizer/model loading failure and inability to load a specific model are too broad to treat as the same bug." }, { "left": "issue:43720", "right": "issue:43927", "accept": false, - "reason": "BitNet packed-weight loading is unrelated to DiaConfig losing custom token IDs on save/load." + "reason": "Packed-weight loading and DiaConfig token-id loss are different serialization problems." }, { "left": "issue:43606", "right": "issue:43927", "accept": false, - "reason": "CPU offload device mismatch is a different path from config field loss causing generation IndexError." + "reason": "CPU offload device mismatch and config token-id loss are unrelated." }, { - "left": "issue:43452", - "right": "issue:44451", + "left": "issue:43278", + "right": "issue:43688", + "accept": false, + "reason": "Embedding dtype drift in evaluate and auxiliary-loss normalization are separate issues." + }, + { + "left": "issue:43994", + "right": "issue:46192", "accept": false, - "reason": "GGUF tokenizer/model loading breakage is not the same as loading ScandiBERT." + "reason": "SigLIP2 output problems and return_tensors being ignored are unrelated API bugs." }, { "left": "issue:43262", "right": "issue:43295", "accept": false, - "reason": "Audio chat-template sampling-rate defaults are unrelated to processor.tokenizer access and image tokenization regression." + "reason": "Audio processor sampling-rate defaulting and processor/tokenizer image-handling regression are different behaviors." }, { - "left": "issue:43278", - "right": "issue:43688", - "accept": false, - "reason": "Embedding dtype drift and auxiliary-loss normalization are different training issues." + "left": "issue:43525", + "right": "issue:43572", + "accept": true, + "reason": "Both report v5 config reload losing pad-token fields; they appear to be the same underlying config-propagation regression with different field names/model classes." }, { - "left": "issue:43720", - "right": "issue:44361", + "left": "issue:44625", + "right": "issue:46002", "accept": false, - "reason": "Packed-weight accelerate loading is unrelated to tokenizer attribute errors on tasks." + "reason": "num_labels propagation and NaNs in classification heads are not the same defect." }, { "left": "issue:33666", "right": "issue:36331", "accept": false, - "reason": "Qwen2-VL multi-GPU training and CustomTrainer compute_loss signature mismatch are different bugs." - }, - { - "left": "issue:43525", - "right": "issue:43572", - "accept": false, - "reason": "Similar missing-pad-token themes, but different config classes and different missing attributes." + "reason": "Multi-GPU training and custom compute_loss signature mismatch are unrelated trainer issues." }, { - "left": "issue:44625", - "right": "issue:46002", + "left": "issue:43720", + "right": "issue:44361", "accept": false, - "reason": "num_labels propagation in Qwen3.5 is unrelated to NaNs in classification heads." + "reason": "Packed-weight loading and MLukeTokenizer task AttributeError are different failures." }, { "left": "issue:43756", "right": "issue:44913", "accept": false, - "reason": "Smollm3 RoPE-layer dropout is not the same issue as GPTNeoXConfig rotary_pct reload regression." + "reason": "Smollm3 RoPE-layer count mismatch and GPTNeoX rotary_pct reload regression are only loosely related to rotary settings, not the same bug." }, { "left": "issue:44849", "right": "issue:45357", "accept": false, - "reason": "Output_hidden_states behavior and save_pretrained visual-encoder key serialization are different regressions." + "reason": "Qwen3.5 hidden-states bug and incorrect visual encoder key serialization are different issues." }, { "left": "issue:43493", "right": "issue:43825", "accept": false, - "reason": "SigLIP2 implementation discrepancy is unrelated to a pipeline error message about translation tasks." + "reason": "SigLIP2 implementation discrepancy and pipeline translation-message regression are unrelated." }, { "left": "issue:28218", "right": "issue:33357", "accept": false, - "reason": "Tokenizer spacing after added tokens is unrelated to a MacOS bus error on a CLIP model." - }, - { - "left": "issue:44568", - "right": "issue:45812", - "accept": false, - "reason": "Missing BOS/EOS with add_special_tokens is a different tokenizer bug than Granite token-ID regression." - }, - { - "left": "issue:44964", - "right": "issue:45405", - "accept": false, - "reason": "A model-loading failure is unrelated to a minimum dependency version bump." - }, - { - "left": "issue:39692", - "right": "issue:43493", - "accept": false, - "reason": "Documentation example mistakes are not the same as the SigLIP2 HF-vs-JAX implementation discrepancy." + "reason": "Tokenizer extra-space behavior and a MacOS bus error on a CLIP model are unrelated." } ] }, - "evaluator_result": null, - "evaluator_used": false, + "evaluator_result": { + "accept": true, + "feedback": "The summary is well grounded in the packet, and the only accepted pair is the one soft edge that plausibly represents the same v5 config reload regression. The other verdicts remain conservative and do not overclaim duplicate status." + }, + "evaluator_used": true, "retried": false, - "accepted_nontrivial_soft_edge": false, + "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null }, @@ -6989,13 +7010,13 @@ "component_index": 1, "component_count": 17, "review_unit_index": 22, - "review_unit_count": 49, + "review_unit_count": 50, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-427", + "cluster_id": "cluster-10000-428", "nodes": [ "issue:33357", "issue:33666", @@ -7017,7 +7038,6 @@ "issue:43756", "issue:43824", "issue:43844", - "issue:43873", "issue:43927", "issue:43994", "issue:44112", @@ -7028,40 +7048,44 @@ "issue:44466", "issue:44488", "issue:44493", + "issue:44568", "issue:44589", "issue:44734", "issue:44792", "issue:44843", + "issue:44964", "issue:44977", "issue:45083", "issue:45200", "issue:45229", "issue:45405", + "issue:45406", "issue:45520", "issue:45584", "issue:45759", - "issue:45803", + "issue:45812", "issue:45823", - "issue:45987", - "issue:46050" + "issue:46121" ], "soft_pairs": [ + "issue:44568|issue:45812", + "issue:39692|issue:43493", + "issue:44964|issue:45405", "issue:43540|issue:44977", "issue:45200|issue:45229", - "issue:43493|issue:43844", "issue:44488|issue:44843", - "issue:43278|issue:43844", "issue:43334|issue:43572", - "issue:45803|issue:45987", - "issue:43756|issue:45083", + "issue:43493|issue:43844", + "issue:43278|issue:43844", "issue:43257|issue:43299", - "issue:39692|issue:43295", "issue:43334|issue:45759", + "issue:43756|issue:45083", + "issue:39692|issue:43295", "issue:45229|issue:45823", - "issue:43646|issue:44466", - "issue:33357|issue:33666", "issue:43824|issue:44589", + "issue:43646|issue:44466", "issue:43504|issue:43606", + "issue:33357|issue:33666", "issue:43927|issue:44361", "issue:44589|issue:45520", "issue:43572|issue:44291", @@ -7073,204 +7097,216 @@ "issue:44792|issue:45405", "issue:43994|issue:44493", "issue:44423|issue:44734", - "issue:43873|issue:46050" + "issue:45406|issue:46121" ], - "prepared_review_unit_hash": "ee50bcc08c4df45e4d1919e6dcbfcfc3adea2c17a06524a72c63d5750fcd3c1c", + "prepared_review_unit_hash": "d33fef169d33194cc24c0a32053a8914aae2cbfd5f5b92be6bb2799ecd217644", "component_budget": { - "node_count": 427, - "item_count": 427, - "soft_pair_count": 1576, - "serialized_chars": 816182, - "estimated_input_tokens": 204046, - "estimated_eval_tokens": 408348 + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 }, "budget": { - "node_count": 47, - "item_count": 47, - "soft_pair_count": 28, - "serialized_chars": 35680, - "estimated_input_tokens": 8920, - "estimated_eval_tokens": 18096 + "node_count": 48, + "item_count": 48, + "soft_pair_count": 30, + "serialized_chars": 36957, + "estimated_input_tokens": 9240, + "estimated_eval_tokens": 18736 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The cluster is mostly a set of unrelated Transformer/model-loading bugs that only superficially overlap by model family or subsystem (Qwen/Gemma/SigLIP2/offload/quantization). None of the soft-paired issues look like the same underlying bug or a single mergeable fix.", - "confidence": 0.95, - "canonical_issue_reason": null, + "summary": "This cluster is highly heterogeneous: it mixes unrelated tokenizer, multimodal, serving, training, config, and CI regressions. None of the soft pairs appears to describe the same concrete bug or change, so all soft edges should be rejected.", + "confidence": 0.97, + "canonical_issue_reason": "No single canonical issue stands out; the items span distinct models and failure modes, so any duplicate anchor would be arbitrary.", "canonical_pr_reason": null, - "best_issue_reason": null, + "best_issue_reason": "No globally best issue candidate is evident because the cluster is not centered on one underlying defect.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:43540", - "right": "issue:44977", + "left": "issue:44568", + "right": "issue:45812", "accept": false, - "reason": "Both involve Qwen3.5/Qwen3-related multimodal text, but one is a video-input processing ValueError and the other is a flash-attention generation issue; different code paths and failure modes." + "reason": "Different tokenizer regressions: missing BOS/EOS on mdeberta-v3 vs wrong token IDs on Granite." }, { - "left": "issue:45200", - "right": "issue:45229", + "left": "issue:39692", + "right": "issue:43493", "accept": false, - "reason": "Gemma 4 in both titles, but one is about missing mm_token_type_ids defaults for text-only fine-tuning and the other is a multi-GPU inference CUDA OOM; not the same bug." + "reason": "Docs/example errors in SigLIP2 vs an implementation discrepancy with the original JAX model." }, { - "left": "issue:43493", - "right": "issue:43844", + "left": "issue:44964", + "right": "issue:45405", "accept": false, - "reason": "SigLIP2 discrepancy vs a training instability under HfDeepSpeedConfig/ZeRO-3; same broad area only, no shared concrete failure or fix." + "reason": "Loading Phi-4 multimodal breaks vs a PEFT version bump; unrelated failure modes." }, { - "left": "issue:44488", - "right": "issue:44843", + "left": "issue:43540", + "right": "issue:44977", "accept": false, - "reason": "One is a specific model load failure for cjvt/sleng-bert, the other is an offline-mode regression in AutoTokenizer/Mistral regex patching; unrelated root causes." + "reason": "Video-input processing error in Qwen3OmniMoe vs flash-attention generation issues in Qwen3.5." }, { - "left": "issue:43278", - "right": "issue:43844", + "left": "issue:45200", + "right": "issue:45229", + "accept": false, + "reason": "Text-only mm_token_type_ids behavior in Gemma 4 vs multi-GPU CUDA OOM in Gemma4 31B." + }, + { + "left": "issue:44488", + "right": "issue:44843", "accept": false, - "reason": "Different problems: embedding dtype drift between train/eval versus gradient growth with ZeRO-3. No shared code-path bug." + "reason": "Model loading failure for sleng-bert vs offline-mode breakage from an unconditional hub call." }, { "left": "issue:43334", "right": "issue:43572", "accept": false, - "reason": "Both are missing-token-ID style config errors, but for different models and different fields (pad_token_id vs pad_token_idx); not mergeable as one fix." + "reason": "Missing pad_token_id in Qwen3-VL config vs missing pad_token_idx in StableLmConfig." }, { - "left": "issue:45803", - "right": "issue:45987", + "left": "issue:43493", + "right": "issue:43844", "accept": false, - "reason": "A bare-except bug in tensor conversion is unrelated to StaticCache.get_seq_length returning the wrong type; different components and fixes." + "reason": "SigLIP2 implementation mismatch vs exploding gradients under ZeRO-3; different bugs." }, { - "left": "issue:43756", - "right": "issue:45083", + "left": "issue:43278", + "right": "issue:43844", "accept": false, - "reason": "Smollm3 RoPE-layer mismatch and qwen3_omni_moe feature-length helper behavior are unrelated model-architecture issues." + "reason": "Dtype drift between train/eval vs gradient instability under DeepSpeed ZeRO-3." }, { "left": "issue:43257", "right": "issue:43299", "accept": false, - "reason": "Both mention Qwen3 MoE loading, but one is about accelerate/deepspeed weight conversion and the other about a dev-version load regression for Qwen3VL MoE; distinct failure points." - }, - { - "left": "issue:39692", - "right": "issue:43295", - "accept": false, - "reason": "One is a SigLIP2 docs example mismatch/quantization issue; the other is a regression in custom model code using processor.tokenizer and image passing. Different bugs." + "reason": "Accelerate/DeepSpeed weight conversion issue vs dev-version load failure for Qwen3VL MoE models." }, { "left": "issue:43334", "right": "issue:45759", "accept": false, - "reason": "Both touch Qwen multimodal config loading, but one is a missing pad_token_id attribute and the other is an AutoModelForCausalLM.from_config unwrap-text_config bug; not the same underlying defect." + "reason": "Pad-token config bug in Qwen3-VL vs composite config unwrapping for Qwen 3.5/3.6." }, { - "left": "issue:45229", - "right": "issue:45823", + "left": "issue:43756", + "right": "issue:45083", "accept": false, - "reason": "Same Gemma4 family, but one is multi-GPU OOM while the other is a PLE device-mismatch during forward; different code paths." + "reason": "RoPE layer dropout in Smollm3 vs a feat-extract helper bug in qwen3_omni_moe." }, { - "left": "issue:43646", - "right": "issue:44466", + "left": "issue:39692", + "right": "issue:43295", "accept": false, - "reason": "Transformers v5 breakage vs lm_head.weight serialization inconsistency; both version-related, but not the same concrete bug or fix." + "reason": "SigLIP2 doc example problems vs processor.tokenizer/image-tokenizer regression." }, { - "left": "issue:33357", - "right": "issue:33666", + "left": "issue:45229", + "right": "issue:45823", "accept": false, - "reason": "Completely different topics: MacOS bus error with CLIP vs Qwen2-VL multi-GPU training." + "reason": "OOM during Gemma4 inference vs device mismatch with device_map='auto'." }, { "left": "issue:43824", "right": "issue:44589", "accept": false, - "reason": "ImportError for a missing Qwen2_5_VL class is unrelated to a Float8 storage object TypeError." + "reason": "Import error for Qwen2_5_VLForConditionalGeneration vs Float8 storage lookup failure." + }, + { + "left": "issue:43646", + "right": "issue:44466", + "accept": false, + "reason": "Custom model init breakage in v5 vs device-dependent lm_head serialization inconsistency." }, { "left": "issue:43504", "right": "issue:43606", "accept": false, - "reason": "Beit legacy-field loading bug and Bark CPU-offload device mismatch are unrelated." + "reason": "Legacy field prevents BEiT segmentation load vs Bark-small CPU offload device mismatch." + }, + { + "left": "issue:33357", + "right": "issue:33666", + "accept": false, + "reason": "MacOS bus error on CLIP model load vs Qwen2-VL multi-GPU training; unrelated." }, { "left": "issue:43927", "right": "issue:44361", "accept": false, - "reason": "Both are tokenizer/config-related, but one is lost custom token IDs after save/load and the other is MLukeTokenizer AttributeError on tasks; not the same bug." + "reason": "DiaConfig save/load token-id loss vs MLukeTokenizer AttributeError on tasks." }, { "left": "issue:44589", "right": "issue:45520", "accept": false, - "reason": "Float8 storage lookup failure is unrelated to a flash_attn import_utils KeyError on Python 3.13." + "reason": "Float8 storage type lookup failure vs Python 3.13 flash_attn import key error." }, { "left": "issue:43572", "right": "issue:44291", "accept": false, - "reason": "Different 5.0-related regressions: StableLmConfig missing pad_token_idx versus init_empty_weights receiving an unexpected _is_hf_initialized argument." + "reason": "Missing StableLmConfig field vs init_empty_weights signature mismatch; distinct initialization bugs." }, { "left": "issue:44464", "right": "issue:45584", "accept": false, - "reason": "Chunked generation/compiled forward inconsistency and Whisper empty-transcription failure after align_special_tokens are unrelated." + "reason": "Compiled-forward chunked generation inconsistency vs Whisper empty-transcription failure." }, { "left": "issue:43504", "right": "issue:43653", "accept": false, - "reason": "Beit legacy-field loading and BigBirdTokenizer special-token registration are distinct bugs with no shared fix." + "reason": "BEiT legacy config load failure vs BigBirdTokenizer special-token registration issue." }, { "left": "issue:43606", "right": "issue:44112", "accept": false, - "reason": "One is a real CPU-offload device mismatch in Bark; the other is a stale GraniteSpeech CI test. Not the same underlying issue." + "reason": "Runtime device mismatch in Bark-small vs a stale GraniteSpeech CI test." }, { "left": "issue:43278", "right": "issue:43701", "accept": false, - "reason": "Embedding dtype drift and resume_from_checkpoint key mismatch are unrelated." + "reason": "Embedding dtype regression vs resume_from_checkpoint key mismatch." }, { "left": "issue:43575", "right": "issue:45405", "accept": false, - "reason": "Qwen2-57B-A14B-Instruct tensor-parallel OOM is unrelated to a PEFT version bump to an unreleased package." + "reason": "Tensor-parallel OOM on Qwen2-57B vs unreleased PEFT version bump." }, { "left": "issue:44792", "right": "issue:45405", "accept": false, - "reason": "A janus model test failure and a MIN_PEFT_VERSION packaging issue are unrelated." + "reason": "Janus image-generation test failure vs PEFT dependency versioning; unrelated." }, { "left": "issue:43994", "right": "issue:44493", "accept": false, - "reason": "SigLIP2 bad outputs with AutoModel/pipeline and unexpected position-id keys across many models are not the same failure." + "reason": "SigLIP2 nonsense outputs vs unexpected position-id key warnings across models." }, { "left": "issue:44423", "right": "issue:44734", "accept": false, - "reason": "Both are serving/runtime crashes, but one is a multimodal string-to-device error in continuous batching and the other is KV-cache continuation tensor indexing; different paths and fixes." + "reason": "Two different serve crashes: string .to() error vs wrong KV-cache indexing." }, { - "left": "issue:43873", - "right": "issue:46050", + "left": "issue:45406", + "right": "issue:46121", "accept": false, - "reason": "Quantization/offloading not working and a sinq save/load breakage are related only broadly by quantization, not by a single code-path bug." + "reason": "Gemma4Processor missing _tokenizer in serve vs rope-parameter validation TypeError." } ] }, @@ -7286,13 +7322,13 @@ "component_index": 1, "component_count": 17, "review_unit_index": 23, - "review_unit_count": 49, + "review_unit_count": 50, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-427", + "cluster_id": "cluster-10000-428", "nodes": [ "issue:15354", "issue:28218", @@ -7303,6 +7339,7 @@ "issue:34567", "issue:37428", "issue:38175", + "issue:39692", "issue:42490", "issue:43232", "issue:43295", @@ -7317,12 +7354,12 @@ "issue:43844", "issue:43873", "issue:43957", + "issue:44186", "issue:44315", "issue:44521", "issue:44554", "issue:44610", "issue:44849", - "issue:44898", "issue:44964", "issue:44977", "issue:44987", @@ -7333,18 +7370,16 @@ "issue:45290", "issue:45325", "issue:45356", - "issue:45406", - "issue:45412", "issue:45446", "issue:45520", "issue:45588", "issue:45823", "issue:46050", - "issue:46121", - "issue:46178" + "issue:46178", + "issue:46192" ], "soft_pairs": [ - "issue:45406|issue:46121", + "issue:43873|issue:46050", "issue:29942|issue:30333", "issue:45588|issue:46050", "issue:42490|issue:43696", @@ -7354,228 +7389,235 @@ "issue:29159|issue:32090", "issue:43582|issue:44977", "issue:44554|issue:45446", - "issue:43540|issue:45200", "issue:45084|issue:45520", + "issue:43540|issue:45200", "issue:15354|issue:38175", "issue:45200|issue:45823", "issue:45072|issue:45198", "issue:37428|issue:43957", "issue:43232|issue:44849", + "issue:44186|issue:46192", "issue:43493|issue:43701", - "issue:43295|issue:44610", + "issue:45823|issue:46192", "issue:44315|issue:44849", - "issue:43295|issue:43450", + "issue:43295|issue:44610", "issue:44554|issue:45356", + "issue:43295|issue:43450", "issue:43572|issue:44987", "issue:45200|issue:45325", + "issue:39692|issue:46192", "issue:43572|issue:43957", "issue:37428|issue:43824", - "issue:45520|issue:46178", - "issue:44898|issue:45412", - "issue:43295|issue:43873" + "issue:45520|issue:46178" ], - "prepared_review_unit_hash": "4ab08ddf9075e71e162a44d4f8abff242e6393a3928c14cb8e89ba1bd0a099eb", + "prepared_review_unit_hash": "b3e926f781aa5bf3c86469b60c31927d08d046b5e8c15ebe820e9fefe1f81be3", "component_budget": { - "node_count": 427, - "item_count": 427, - "soft_pair_count": 1576, - "serialized_chars": 816182, - "estimated_input_tokens": 204046, - "estimated_eval_tokens": 408348 + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 }, "budget": { - "node_count": 48, - "item_count": 48, - "soft_pair_count": 29, - "serialized_chars": 36484, - "estimated_input_tokens": 9121, - "estimated_eval_tokens": 18498 + "node_count": 47, + "item_count": 47, + "soft_pair_count": 30, + "serialized_chars": 36155, + "estimated_input_tokens": 9039, + "estimated_eval_tokens": 18334 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is highly heterogeneous: the paired issues touch unrelated bugs across tokenization, generation, loading, attention kernels, training, and utilities. None of the soft pairs look like the same underlying defect, so I would not merge any of them.", - "confidence": 0.98, - "canonical_issue_reason": "issue:43873 is the best representative issue in this set because it is open, broadly scoped, and has meaningful discussion/inbound references; the rest of the cluster is too disparate to form a true duplicate group.", + "summary": "This cluster is highly heterogeneous: most items are unrelated tokenizer, multimodal processor, flash-attention, loading, training, or quantization bugs. The soft edges are mostly embedding-level similarities rather than true duplicates, so none should be merged.", + "confidence": 0.62, + "canonical_issue_reason": "No single canonical issue is safe here because the cluster is over-broad. If one representative must be chosen, issue:43295 is the closest concrete, user-facing regression among the few multimodal/processor-related items, but it does not represent the whole cluster.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43873 is the strongest standalone issue to keep if one item must represent the cluster: it is active, user-facing, and describes a general framework behavior problem rather than a narrow edge case.", + "best_issue_reason": "Issue:43295 is the best representative of the densest coherent subtheme in this cluster: a regression around processor/tokenizer behavior in multimodal code paths. It is more concrete than the broader, noisier entries and has the strongest structural similarity to the few nearby processor-related issues.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:45406", - "right": "issue:46121", + "left": "issue:43873", + "right": "issue:46050", "accept": false, - "reason": "Different failures: Gemma4 processor token access vs rope-parameter list validation." + "reason": "Different bugs: quantization/offloading behavior versus save/load correctness for a quantized model. Same broad area, not same underlying defect." }, { "left": "issue:29942", "right": "issue:30333", "accept": false, - "reason": "Unrelated areas: Flash Attention 2 test failures vs MLflow reporting after failed jobs." + "reason": "Completely unrelated subsystems: Flash Attention tests versus MLFlow reporting on failed jobs." }, { "left": "issue:45588", "right": "issue:46050", "accept": false, - "reason": "Different code paths: flash_attention sinkless-model crash vs sinq save/load breakage." + "reason": "Different code paths: flash_attention sink-less crash versus sinq quantized model serialization/deserialization." }, { "left": "issue:42490", "right": "issue:43696", "accept": false, - "reason": "Import-path warning vs CUDA OOM; no shared underlying bug." + "reason": "Local-path import warning is unrelated to CUDA OOM during GPT-oss loading." }, { "left": "issue:44521", "right": "issue:45290", "accept": false, - "reason": "Both involve apply_chat_template, but one is multimodal assistant masks and the other is tool-call content handling; not the same defect." + "reason": "Both involve apply_chat_template, but one is multimodal assistant masks and the other is tool-call tokenization crash; these are distinct failures." }, { "left": "issue:43844", "right": "issue:44964", "accept": false, - "reason": "Different problems: ZeRO-3 gradient growth vs model-loading failure for Phi-4 multimodal." + "reason": "Unrelated: ZeRO-3 gradient anomaly versus loading failure for a specific multimodal model." }, { "left": "issue:28218", "right": "issue:34567", "accept": false, - "reason": "Tokenizer spacing bug is unrelated to TrainerState token counter updates." + "reason": "Tokenizer spacing bug and TrainerState token counter update are unrelated." }, { "left": "issue:29159", "right": "issue:32090", "accept": false, - "reason": "Tokenizer parity issue vs Trainer NoneType broadcast error." + "reason": "Slow/fast tokenizer inconsistency is unrelated to a Trainer GPU broadcast NoneType error." }, { "left": "issue:43582", "right": "issue:44977", "accept": false, - "reason": "Apple Silicon allocator TypeError is unrelated to Qwen3.5 flash-attention generation behavior." + "reason": "Apple Silicon allocator warmup TypeError and Qwen3.5 flash-attention generation bug are different problems." }, { "left": "issue:44554", "right": "issue:45446", "accept": false, - "reason": "MPS attention correctness bug vs PyTorch version-check import bug." + "reason": "Both touch attention/flex_attention, but one is an MPS correctness issue and the other is a version-check import bug." }, { - "left": "issue:43540", - "right": "issue:45200", + "left": "issue:45084", + "right": "issue:45520", "accept": false, - "reason": "Video-input processing error in Qwen3OmniMoe is unrelated to Gemma4 mm_token_type_ids defaults." + "reason": "Compiler/template-node error is unrelated to the Python 3.13 flash_attn import KeyError." }, { - "left": "issue:45084", - "right": "issue:45520", + "left": "issue:43540", + "right": "issue:45200", "accept": false, - "reason": "Non-template-node compilation error vs backend import KeyError for flash_attn." + "reason": "Both are multimodal, but one is video-input processing in Qwen3OmniMoe and the other is a Gemma4 token-type-id defaulting issue." }, { "left": "issue:15354", "right": "issue:38175", "accept": false, - "reason": "TorchScript GeneratorExp export issue is unrelated to SigLIP2 zero-probability outputs." + "reason": "TorchScript GeneratorExp export failure and SigLIP2 zero probabilities are unrelated model behaviors." }, { "left": "issue:45200", "right": "issue:45823", "accept": false, - "reason": "Different model families and failures: default mm_token_type_ids vs Gemma4 device-map mismatch." + "reason": "Same model family (Gemma4) but different bugs: required mm_token_type_ids versus device-map forward mismatch." }, { "left": "issue:45072", "right": "issue:45198", "accept": false, - "reason": "Dtype mismatch in inference is unrelated to Wav2Vec2 save_pretrained/tokenization failure." + "reason": "Different failures in different models: dtype mismatch in inference versus save_pretrained/tokenization breakage." }, { "left": "issue:37428", "right": "issue:43957", "accept": false, - "reason": "Both are loading/import issues, but one is flash-support import and the other is meta-device model loading; not the same root cause." + "reason": "Flash-attention import symbol error and meta-device model-loading breakage are distinct issues." }, { "left": "issue:43232", "right": "issue:44849", "accept": false, - "reason": "Generation kwargs update after sync_gpus is unrelated to Qwen3.5 hidden-states output bug." + "reason": "Generation kwargs sync issue and Qwen3.5 hidden-states bug are separate generation-path defects." + }, + { + "left": "issue:44186", + "right": "issue:46192", + "accept": false, + "reason": "NER/batched-padding tokenizer crash is not the same as silently ignored return_tensors when text_kwargs is passed." }, { "left": "issue:43493", "right": "issue:43701", "accept": false, - "reason": "SigLIP2 implementation discrepancy is unrelated to resume-from-checkpoint key mismatch." + "reason": "SigLIP2 implementation discrepancy and checkpoint resume key mismatch are unrelated." }, { - "left": "issue:43295", - "right": "issue:44610", + "left": "issue:45823", + "right": "issue:46192", "accept": false, - "reason": "Processor/tokenizer regression is different from an OmDet-Turbo image-size preprocessing bug." + "reason": "Device-map mismatch in Gemma4 and ignored return_tensors in multimodal kwargs are different bugs." }, { "left": "issue:44315", "right": "issue:44849", "accept": false, - "reason": "Liger Kernel application in model_init is unrelated to Qwen3.5 output_hidden_states behavior." + "reason": "Liger Kernel application with model_init is unrelated to Qwen3.5 output_hidden_states behavior." }, { "left": "issue:43295", - "right": "issue:43450", + "right": "issue:44610", "accept": false, - "reason": "Different processor bugs: tokenizer/image passing regression vs batched video shape error." + "reason": "Both involve multimodal processing, but the first is a processor.tokenizer regression and the second is a video processor output-shape bug; not the same underlying defect." }, { "left": "issue:44554", "right": "issue:45356", "accept": false, - "reason": "MPS attention correctness issue is unrelated to Kimi-K2.5 tokenizer codec/regex regression." + "reason": "MPS attention correctness and Kimi-K2.5 tokenizer codec regression are unrelated." + }, + { + "left": "issue:43295", + "right": "issue:43450", + "accept": false, + "reason": "Processor.tokenizer regression and batched video processor shape bug are different code paths and would not be merged into one fix." }, { "left": "issue:43572", "right": "issue:44987", "accept": false, - "reason": "StableLmConfig pad_token_idx regression is not the same as loading physical-intelligence/fast on transformers>=5.1.0." + "reason": "StableLmConfig pad_token_idx regression is unrelated to physical-intelligence/fast loading failure." }, { "left": "issue:45200", "right": "issue:45325", "accept": false, - "reason": "Different multimodal bugs: default mm_token_type_ids vs Qwen2.5-VL rope-index scaling." + "reason": "Both are multimodal model issues, but one is missing/default mm_token_type_ids and the other is a Qwen2.5-VL rope-index scaling bug." + }, + { + "left": "issue:39692", + "right": "issue:46192", + "accept": false, + "reason": "SigLIP2 documentation errors and ignored return_tensors are unrelated." }, { "left": "issue:43572", "right": "issue:43957", "accept": false, - "reason": "Config field regression vs meta-device loading failures; separate causes." + "reason": "Pad token config regression and meta-device model-loading failure are not the same bug." }, { "left": "issue:37428", "right": "issue:43824", "accept": false, - "reason": "Different import errors for different symbols/models; no shared bug." + "reason": "Flash-attention utility import error and Qwen2.5-VL class import error are separate import problems with different missing symbols." }, { "left": "issue:45520", "right": "issue:46178", "accept": false, - "reason": "Both are backend/import mapping issues, but flash_attn KeyError and tensorflow_text backend mapping are distinct problems." - }, - { - "left": "issue:44898", - "right": "issue:45412", - "accept": false, - "reason": "Perceiver non-default resolution failure is unrelated to RT-DETR memory release behavior." - }, - { - "left": "issue:43295", - "right": "issue:43873", - "accept": false, - "reason": "Processor/tokenizer regression and offloading-with-quantization are different subsystems and symptoms." + "reason": "Both are import/configuration errors, but flash_attn KeyError on Python 3.13 and tensorflow_text backend mapping are unrelated." } ] }, @@ -7591,13 +7633,13 @@ "component_index": 1, "component_count": 17, "review_unit_index": 24, - "review_unit_count": 49, + "review_unit_count": 50, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-427", + "cluster_id": "cluster-10000-428", "nodes": [ "issue:28218", "issue:35141", @@ -7607,6 +7649,7 @@ "issue:42915", "issue:43262", "issue:43278", + "issue:43295", "issue:43334", "issue:43335", "issue:43388", @@ -7624,6 +7667,7 @@ "issue:43827", "issue:43844", "issue:43856", + "issue:43873", "issue:43901", "issue:43975", "issue:44112", @@ -7635,33 +7679,35 @@ "issue:44610", "issue:44792", "issue:44857", + "issue:44898", "issue:45083", "issue:45127", "issue:45198", - "issue:45290", - "issue:45381", "issue:45412", "issue:45520", "issue:45563", "issue:45584", - "issue:45593", - "issue:46170" + "issue:46170", + "issue:46192" ], "soft_pairs": [ - "issue:43550|issue:43844", + "issue:44898|issue:45412", "issue:44466|issue:45127", + "issue:43295|issue:43873", + "issue:43550|issue:43844", "issue:43334|issue:43525", "issue:44792|issue:45563", "issue:43824|issue:43975", - "issue:43388|issue:43526", "issue:42915|issue:43824", + "issue:43388|issue:43526", "issue:43450|issue:43575", "issue:43278|issue:43856", "issue:42490|issue:43493", + "issue:43975|issue:45520", "issue:28218|issue:35141", "issue:43653|issue:44112", - "issue:43975|issue:45520", "issue:38175|issue:43493", + "issue:43526|issue:46192", "issue:44442|issue:45198", "issue:44514|issue:44521", "issue:45520|issue:46170", @@ -7672,214 +7718,218 @@ "issue:43819|issue:44112", "issue:44610|issue:44857", "issue:43819|issue:45083", - "issue:43335|issue:43550", - "issue:43606|issue:45083", "issue:37428|issue:43425", - "issue:44857|issue:45593", - "issue:45290|issue:45381" + "issue:43335|issue:43550", + "issue:43606|issue:45083" ], - "prepared_review_unit_hash": "404bf3588943a985eef0780d3fda15a7eb10cba995d44004f5b61ad704824c4e", + "prepared_review_unit_hash": "391218718bb8c247740b4efd9963e0a30109b5e7e3a07cf0cce28d249aca013a", "component_budget": { - "node_count": 427, - "item_count": 427, - "soft_pair_count": 1576, - "serialized_chars": 816182, - "estimated_input_tokens": 204046, - "estimated_eval_tokens": 408348 + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 }, "budget": { - "node_count": 47, - "item_count": 47, - "soft_pair_count": 29, - "serialized_chars": 36034, - "estimated_input_tokens": 9009, - "estimated_eval_tokens": 18274 + "node_count": 48, + "item_count": 48, + "soft_pair_count": 30, + "serialized_chars": 36907, + "estimated_input_tokens": 9227, + "estimated_eval_tokens": 18710 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The items are a loose topical cluster, not a set of true duplicates. Most pairs share a subsystem or API name at a high level, but they describe different bugs, different code paths, or different model-specific failures. No soft edge should be merged.", - "confidence": 0.93, - "canonical_issue_reason": "issue:44521 is the best anchor: it is open, active, has the most inbound references in this set, and represents a broad, user-visible multimodal `apply_chat_template` bug.", + "summary": "This cluster is mostly heterogeneous: the soft pairs are largely same-subsystem but not same-bug matches, so I reject them all. There is no strong duplicate canonical issue; if an anchor is needed, issue 44521 is the broadest and most active issue in the closest thematic area (multimodal apply_chat_template).", + "confidence": 0.86, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "issue:44521 looks like the strongest overall issue candidate because it has high discussion/inbound activity and a general API impact rather than a narrow model-specific failure.", + "best_issue_reason": "Issue 44521 has the broadest scope and the most traction among the only plausible subcluster (multimodal apply_chat_template regressions), so it is the best anchor even though the cluster is not a true duplicate set.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:43550", - "right": "issue:43844", + "left": "issue:44898", + "right": "issue:45412", "accept": false, - "reason": "Different failures: SDPA/torch.compile runtime bug vs DeepSpeed/ZeRO-3 gradient explosion during training." + "reason": "Different bugs: image-resolution handling vs memory not being released on deletion." }, { "left": "issue:44466", "right": "issue:45127", "accept": false, - "reason": "Both mention tied weights, but one is serialization/device-dependent `lm_head.weight` handling and the other is LoRA merge collapse after vocab extension." + "reason": "One is tied-weight serialization/device-dependent behavior; the other is LoRA merge collapse with extended vocab. Related theme, different failure." + }, + { + "left": "issue:43295", + "right": "issue:43873", + "accept": false, + "reason": "Processor API regression and quantization/offload behavior are unrelated code paths." + }, + { + "left": "issue:43550", + "right": "issue:43844", + "accept": false, + "reason": "Torch.compile/SDPA failure is not the same as a ZeRO-3 gradient growth issue." }, { "left": "issue:43334", "right": "issue:43525", "accept": false, - "reason": "Same symptom (`pad_token_id` missing) but on different config classes and models; not enough evidence of one underlying bug." + "reason": "Both mention missing pad_token_id, but they are model-specific config failures and not clearly one shared bug." }, { "left": "issue:44792", "right": "issue:45563", "accept": false, - "reason": "Unrelated generate-path issues: a Janus image-generation test failure vs a stale warning about `num_return_sequences`." + "reason": "Janus image-generation test failure and a paged-generate warning are distinct issues." }, { "left": "issue:43824", "right": "issue:43975", "accept": false, - "reason": "ImportError for a missing exported class vs a detokenization regression in a specific model." + "reason": "ImportError for a missing class and a detokenization regression are different failures." }, { - "left": "issue:43388", - "right": "issue:43526", + "left": "issue:42915", + "right": "issue:43824", "accept": false, - "reason": "Both concern label handling, but they hit different data pipelines (`gather_for_metrics` vs Beit image preprocessing) and different failure modes." + "reason": "FineGrainedFP8Config failure is unrelated to a missing Qwen2.5-VL import." }, { - "left": "issue:42915", - "right": "issue:43824", + "left": "issue:43388", + "right": "issue:43526", "accept": false, - "reason": "Completely different problems: Qwen3MoE FP8 config failure vs Qwen2.5-VL import/export issue." + "reason": "Different layers: metric gathering vs image-processor label reduction." }, { "left": "issue:43450", "right": "issue:43575", "accept": false, - "reason": "Different domains and symptoms: batched video processor shape bug vs tensor-parallel OOM on a large text model." + "reason": "Batched video shape bug and tensor-parallel OOM are not the same underlying problem." }, { "left": "issue:43278", "right": "issue:43856", "accept": false, - "reason": "Embedding dtype drift during eval is unrelated to Qwen3 MoE memory inefficiency." + "reason": "Training/eval dtype mismatch is unrelated to MoE memory usage." }, { "left": "issue:42490", "right": "issue:43493", "accept": false, - "reason": "Path-trailing-slash import warning and SigLIP2 implementation discrepancy are unrelated." + "reason": "Local-path import warning and SigLIP2 implementation discrepancy are unrelated." + }, + { + "left": "issue:43975", + "right": "issue:45520", + "accept": false, + "reason": "Detokenization regression and flash_attn import-time KeyError are unrelated." }, { "left": "issue:28218", "right": "issue:35141", "accept": false, - "reason": "Tokenizer spacing after added tokens is unrelated to embedding reinitialization when resizing vocab." + "reason": "Tokenizer spacing bug and embedding reinit after resize are different code paths." }, { "left": "issue:43653", "right": "issue:44112", "accept": false, - "reason": "Tokenizer special-token registration bug vs a GraniteSpeech CI stale device override test." + "reason": "Special-token registration bug and a stale CI test are not duplicates." }, { - "left": "issue:43975", - "right": "issue:45520", + "left": "issue:38175", + "right": "issue:43493", "accept": false, - "reason": "Model detokenization regression is unrelated to Python 3.13 `flash_attn` import metadata failure." + "reason": "Both involve SigLIP2, but zero probabilities vs JAX discrepancy are not clearly the same concrete bug." }, { - "left": "issue:38175", - "right": "issue:43493", + "left": "issue:43526", + "right": "issue:46192", "accept": false, - "reason": "Related model family, but one is zero-probability output and the other is a broader HF-vs-JAX discrepancy; not clearly the same defect." + "reason": "Label reduction behavior and text_kwargs/return_tensors handling are unrelated." }, { "left": "issue:44442", "right": "issue:45198", "accept": false, - "reason": "Different tokenizer/model families and different failure points (`AutoTokenizer` load vs save/tokenization failure)." + "reason": "Different tokenizer/processor failures in different model families." }, { "left": "issue:44514", "right": "issue:44521", "accept": false, - "reason": "Both are chat-template related, but one is a batching crash with `padding=False` and the other is incorrect assistant-mask generation for multimodal inputs." + "reason": "Same API area, but one is a batched padding=False crash and the other is all-zero assistant masks; not the same concrete failure." }, { "left": "issue:45520", "right": "issue:46170", "accept": false, - "reason": "Different environment/platform issues: `flash_attn` import lookup on Python 3.13 vs Windows ROCm/torch.distributed dependency errors." + "reason": "Different import/platform dependency problems on different stacks." }, { "left": "issue:38175", "right": "issue:45412", "accept": false, - "reason": "SigLIP2 probability issue and RT-DETR memory reclamation bug are unrelated." + "reason": "SigLIP2 probability issue and RT-DETR memory leak are unrelated." }, { "left": "issue:44206", "right": "issue:45584", "accept": false, - "reason": "LASR feature extractor regression is unrelated to Whisper empty-transcription failure after special-token alignment." + "reason": "Unsupported center argument in feature extraction is unrelated to empty-transcription failure after align_special_tokens." }, { "left": "issue:43827", "right": "issue:43901", "accept": false, - "reason": "Both are docs problems, but they concern different pipeline APIs and different deprecations." + "reason": "Both are docs issues, but they cover different APIs and deprecations." }, { "left": "issue:43262", "right": "issue:43450", "accept": false, - "reason": "Audio chat-template sampling-rate bug is unrelated to video processor batch-shape handling." + "reason": "Audio sampling-rate default bug and video batched-shape bug are unrelated." }, { "left": "issue:43819", "right": "issue:44112", "accept": false, - "reason": "DAC latent decoding mismatch is unrelated to a GraniteSpeech CI test flake." + "reason": "DAC latent reconstruction mismatch and a stale CI test are unrelated." }, { "left": "issue:44610", "right": "issue:44857", "accept": false, - "reason": "Different vision models and different bugs: input-size mismatch vs AMP/CUDA loss crash." + "reason": "Processor input-size mismatch and AMP/CUDA loss crash are different problems." }, { "left": "issue:43819", "right": "issue:45083", "accept": false, - "reason": "Unrelated code paths: DAC latent reconstruction vs Qwen3 Omni feature-length helper behavior." - }, - { - "left": "issue:43335", - "right": "issue:43550", - "accept": false, - "reason": "SwitchTransformers sparse-layer construction bug is unrelated to Bamba SDPA/torch.compile failure." - }, - { - "left": "issue:43606", - "right": "issue:45083", - "accept": false, - "reason": "CPU-offload device mismatch is unrelated to Qwen3 Omni feature-length computation." + "reason": "Latent decoding mismatch and feature-length helper behavior are not the same bug." }, { "left": "issue:37428", "right": "issue:43425", "accept": false, - "reason": "A specific missing import and a broad Torch 2.10 incompatibility are not the same concrete bug." + "reason": "Missing import for a model class and Torch version incompatibility are unrelated." }, { - "left": "issue:44857", - "right": "issue:45593", + "left": "issue:43335", + "right": "issue:43550", "accept": false, - "reason": "Different detection-model issues: AMP crash in one, auxiliary-loss logic in the other." + "reason": "Sparse-layer config bug and torch.compile/SDPA failure are different issues." }, { - "left": "issue:45290", - "right": "issue:45381", + "left": "issue:43606", + "right": "issue:45083", "accept": false, - "reason": "Different Qwen2.5-VL failures: tool-call/chat-template crash vs incorrect video `vision_position_ids`." + "reason": "CPU-offload device mismatch and feature-length helper behavior are unrelated." } ] }, @@ -7895,24 +7945,20 @@ "component_index": 1, "component_count": 17, "review_unit_index": 25, - "review_unit_count": 49, + "review_unit_count": 50, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-427", + "cluster_id": "cluster-10000-428", "nodes": [ - "issue:28218", "issue:29127", "issue:29942", "issue:32090", - "issue:34689", - "issue:35141", "issue:37428", "issue:38175", "issue:39692", - "issue:40444", "issue:42490", "issue:43278", "issue:43335", @@ -7922,7 +7968,6 @@ "issue:43606", "issue:43611", "issue:43653", - "issue:43701", "issue:43720", "issue:43819", "issue:43825", @@ -7946,290 +7991,268 @@ "issue:45200", "issue:45229", "issue:45245", + "issue:45290", "issue:45310", + "issue:45381", "issue:45478", "issue:45588", + "issue:45593", "issue:45823", - "issue:45859" + "issue:45859", + "issue:45920", + "issue:46192" ], "soft_pairs": [ + "issue:44857|issue:45593", + "issue:45290|issue:45381", "issue:44448|issue:44743", "issue:45127|issue:45245", "issue:29942|issue:32090", "issue:45200|issue:45588", "issue:43278|issue:43825", - "issue:43335|issue:43653", "issue:37428|issue:44351", "issue:44206|issue:45823", + "issue:43335|issue:43653", "issue:43611|issue:43931", "issue:43550|issue:44112", "issue:43819|issue:44493", - "issue:43335|issue:43504", "issue:44855|issue:45083", + "issue:43335|issue:43504", "issue:45588|issue:45859", "issue:29127|issue:32090", - "issue:39692|issue:43526", "issue:43335|issue:45083", + "issue:39692|issue:43526", "issue:45310|issue:45478", "issue:39692|issue:44493", "issue:43720|issue:44112", "issue:43335|issue:45072", + "issue:45920|issue:46192", "issue:43335|issue:44112", + "issue:44360|issue:44561", "issue:44448|issue:44779", "issue:42490|issue:43854", - "issue:44360|issue:44561", "issue:44977|issue:45229", "issue:38175|issue:44493", - "issue:43606|issue:44857", - "issue:34689|issue:35141", - "issue:44857|issue:45072", - "issue:44448|issue:44977", - "issue:40444|issue:43526", - "issue:28218|issue:32090", - "issue:43701|issue:45083" + "issue:43606|issue:44857" ], - "prepared_review_unit_hash": "17bfd9c1f9db20e875b0bdea439f76f5e3aca291d2e6dbb0b0c35b69a026e5ee", + "prepared_review_unit_hash": "85e61ef1745c23f4b1e4d9e5803f09b992e03a86c7403686d7e135e120e0de52", "component_budget": { - "node_count": 427, - "item_count": 427, - "soft_pair_count": 1576, - "serialized_chars": 816182, - "estimated_input_tokens": 204046, - "estimated_eval_tokens": 408348 + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 34, - "serialized_chars": 37982, - "estimated_input_tokens": 9496, - "estimated_eval_tokens": 19248 + "soft_pair_count": 31, + "serialized_chars": 37145, + "estimated_input_tokens": 9287, + "estimated_eval_tokens": 18830 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly heterogeneous and does not represent one duplicate family. The only clear duplicate pair is the two identical Qwen3.5 MoE from_pretrained reports; the rest are topic-adjacent at best but describe different code paths or different models.", - "confidence": 0.94, - "canonical_issue_reason": "Use issue:45310 as the canonical issue because it is the earlier of the only true duplicate pair and has the same underlying Qwen3.5 MoE from_pretrained failure as issue:45478.", + "summary": "Mostly unrelated singleton issues. There is one very similar pair (45310/45478) with matching titles, but the evidence is still soft, so the cluster should not be treated as a confirmed duplicate set.", + "confidence": 0.68, + "canonical_issue_reason": "Issue 45310 is the best representative for the shared Qwen3.5 Moe from_pretrained theme, but the evidence only supports a soft similarity, not a confirmed duplicate cluster.", "canonical_pr_reason": null, - "best_issue_reason": "Issue:45310 is the best representative of the cluster\u2019s only real duplicate because it is an exact title match to 45478; the other issues are unrelated or only loosely similar.", + "best_issue_reason": "Issue 45310 is the strongest issue representative because it is earlier and matches 45478 most closely, but it should be treated as a likely topic match rather than a proven duplicate.", "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:44857", + "right": "issue:45593", + "accept": false, + "reason": "Different failures: AMP/CUDA loss crash vs missing auxiliary losses." + }, + { + "left": "issue:45290", + "right": "issue:45381", + "accept": false, + "reason": "Tool-call chat-template crash and video vision_position_ids bug are unrelated." + }, { "left": "issue:44448", "right": "issue:44743", "accept": false, - "reason": "Both mention output regressions across versions, but they affect different models and different mechanisms: Pegasus generation vs Qwen3.5 recurrent state handling." + "reason": "Pegasus output drift and Qwen3 recurrent-state reset affect different code paths." }, { "left": "issue:45127", "right": "issue:45245", "accept": false, - "reason": "These are unrelated bugs: LoRA merge collapse with extended vocabulary versus a categorical-count limit runtime error." + "reason": "LoRA/vocab collapse and category-count limit are distinct bugs." }, { "left": "issue:29942", "right": "issue:32090", "accept": false, - "reason": "Flash Attention 2 test failures and a Trainer _gpu_broadcast_one NoneType error are distinct failures in different subsystems." + "reason": "Flash Attention test failures and NoneType GPU broadcast error are different problems." }, { "left": "issue:45200", "right": "issue:45588", "accept": false, - "reason": "Gemma4 text-only mm_token_type_ids handling and a flash_attention s_aux=None crash are different code paths and symptoms." + "reason": "Gemma4 token-type handling and flash_attention s_aux=None crash are unrelated." }, { "left": "issue:43278", "right": "issue:43825", "accept": false, - "reason": "Evaluate-time dtype drift for embeddings is unrelated to the pipeline translation-task warning message." - }, - { - "left": "issue:43335", - "right": "issue:43653", - "accept": false, - "reason": "SwitchTransformers sparse-layer construction and BigBirdTokenizer special-token registration are separate bugs in different components." + "reason": "BF16/FP32 dtype regression and pipeline task-support messaging do not share a code path." }, { "left": "issue:37428", "right": "issue:44351", "accept": false, - "reason": "These are both import-related, but they fail on different missing symbols and different dependency paths." + "reason": "Both are import errors, but they involve different missing symbols and likely different fixes." }, { "left": "issue:44206", "right": "issue:45823", "accept": false, - "reason": "Unsupported center arg in LasrFeatureExtractor and Gemma4 PLE device mismatch are unrelated failures." + "reason": "Unsupported center argument in feature extractor and Gemma4 device mismatch are unrelated." + }, + { + "left": "issue:43335", + "right": "issue:43653", + "accept": false, + "reason": "Sparse-layer config bug and tokenizer special-token registration bug affect different components." }, { "left": "issue:43611", "right": "issue:43931", "accept": false, - "reason": "Base-model-prefix loading regression and Qwen3-VL weight-shape mismatch are different model-loading issues." + "reason": "base_model_prefix loading breakage and Qwen3-VL shape mismatch are not the same defect." }, { "left": "issue:43550", "right": "issue:44112", "accept": false, - "reason": "torch.compile/SDPA failure in Bamba and a stale CI override test in GraniteSpeech do not describe the same bug." + "reason": "torch.compile SDPA failure and CI device override failure are different issues." }, { "left": "issue:43819", "right": "issue:44493", "accept": false, - "reason": "DAC.from_latents missing STE is unrelated to unexpected position-id keys during model loading." + "reason": "DAC from_latents mismatch and unexpected position-id warnings are unrelated model bugs." }, { - "left": "issue:43335", - "right": "issue:43504", + "left": "issue:44855", + "right": "issue:45083", "accept": false, - "reason": "Sparse-layer creation in SwitchTransformers and Beit legacy-field loading are different BEIT/SwitchTransformers issues." + "reason": "Python 3.13 import parsing failure and qwen3_omni_moe helper behavior are different problems." }, { - "left": "issue:44855", - "right": "issue:45083", + "left": "issue:43335", + "right": "issue:43504", "accept": false, - "reason": "Python 3.13 IndentationError in DebertaV2Model is unrelated to the qwen3_omni_moe feature-length helper behavior." + "reason": "Sparse layer creation bug and BEiT legacy-field loading bug are unrelated." }, { "left": "issue:45588", "right": "issue:45859", "accept": false, - "reason": "A flash_attention crash on sink-less models is a different problem from CPU-offload incompatibility in Qwen3_5MoeTextRotaryEmbedding." + "reason": "Both mention offload/flash attention, but the concrete failures differ." }, { "left": "issue:29127", "right": "issue:32090", "accept": false, - "reason": "LayoutLMv3 error-message clarity and Trainer broadcast NoneType support are not the same underlying bug." + "reason": "LayoutLMv3 error-message clarity and Trainer broadcast NoneType error are unrelated." }, { - "left": "issue:39692", - "right": "issue:43526", + "left": "issue:43335", + "right": "issue:45083", "accept": false, - "reason": "SigLIP2 docs/model-processor example problems are unrelated to BeitImageProcessorFast reduce_labels returning only one label." + "reason": "SwitchTransformers config bug and qwen3_omni_moe helper bug are unrelated." }, { - "left": "issue:43335", - "right": "issue:45083", + "left": "issue:39692", + "right": "issue:43526", "accept": false, - "reason": "SwitchTransformers sparse-layer creation and qwen3_omni_moe feature-length calculation are different model internals." + "reason": "SigLIP2 docs/example errors and BeitImageProcessorFast reduce_labels bug are different problems." }, { "left": "issue:45310", "right": "issue:45478", - "accept": true, - "reason": "Exact duplicate reports: same title, same transformers>=5.4.0 Qwen3.5 MoE from_pretrained failure, same underlying bug." + "accept": false, + "reason": "Titles are near-identical, but the packet only provides soft similarity and no strong duplicate evidence." }, { "left": "issue:39692", "right": "issue:44493", "accept": false, - "reason": "Documentation example failures for SigLIP2 and unexpected position-id keys are unrelated issues." + "reason": "SigLIP2 documentation/model mismatch and unexpected position-id keys are unrelated." }, { "left": "issue:43720", "right": "issue:44112", "accept": false, - "reason": "BitNet packed-weight loading and GraniteSpeech stale device override are distinct CI/model-loading problems." + "reason": "BitNet packed-weight loading and GraniteSpeech device-override CI issues are different." }, { "left": "issue:43335", "right": "issue:45072", "accept": false, - "reason": "SwitchTransformers config bug and bfloat16 dtype mismatch in inference are not the same failure." + "reason": "Sparse-layer construction and bfloat16 dtype mismatch are unrelated." }, { - "left": "issue:43335", - "right": "issue:44112", + "left": "issue:45920", + "right": "issue:46192", + "accept": false, + "reason": "Tokenizer wrong-token-IDs regression and return_tensors handling bug are different API issues." + }, + { + "left": "issue:44360", + "right": "issue:44561", "accept": false, - "reason": "SwitchTransformers sparse-layer construction and GraniteSpeech device-override CI failure are unrelated." + "reason": "DSA indexer discussion and removal of is_torch_fx_available are unrelated." }, { "left": "issue:44448", "right": "issue:44779", "accept": false, - "reason": "Both concern v4/v5 output differences, but they involve different models and different tokenizer/generation behavior." + "reason": "Pegasus output differences and DeepSeek tokenizer regression are different subsystems." }, { "left": "issue:42490", "right": "issue:43854", "accept": false, - "reason": "Local-path import warning and GLM-4.7-Flash unit-test loading failure are different loading scenarios." - }, - { - "left": "issue:44360", - "right": "issue:44561", - "accept": false, - "reason": "The DSA indexer ReLU discussion is unrelated to trust_remote_code breakage from removing is_torch_fx_available." + "reason": "Local-path import warning and GLM-4.7-Flash load failure are unrelated." }, { "left": "issue:44977", "right": "issue:45229", "accept": false, - "reason": "Flash-attention generation issues for Qwen3.5 and Gemma4 multi-GPU OOM are different problems." + "reason": "Qwen3.5 flash-attention generation bug and Gemma4 CUDA OOM are unrelated." }, { "left": "issue:38175", "right": "issue:44493", "accept": false, - "reason": "Unexpected zero probabilities in SigLIP2 and position-id key warnings in many models are not the same underlying defect." + "reason": "SigLIP2 zero-probabilities and position-id key warnings are unrelated." }, { "left": "issue:43606", "right": "issue:44857", "accept": false, - "reason": "CPU offload device mismatch in bark-small and float16 AMP crash in LwDetrImageLoss are separate device/precision bugs." - }, - { - "left": "issue:34689", - "right": "issue:35141", - "accept": false, - "reason": "Llama 3.2 Vision model loading regression and embedding reinitialization after resize are unrelated model-loading behaviors." - }, - { - "left": "issue:44857", - "right": "issue:45072", - "accept": false, - "reason": "AMP crash in LwDetrImageLoss and bfloat16 dtype mismatch in SwitchTransformers/TimmWrapperModel are different precision issues." - }, - { - "left": "issue:44448", - "right": "issue:44977", - "accept": false, - "reason": "Pegasus output differences and Qwen3.5 flash-attention generation failures are unrelated model-specific regressions." - }, - { - "left": "issue:40444", - "right": "issue:43526", - "accept": false, - "reason": "Qwen2.5-VL iterable-dataset multi-image finetuning failure and BeitImageProcessorFast label reduction are different image-model bugs." - }, - { - "left": "issue:28218", - "right": "issue:32090", - "accept": false, - "reason": "Tokenizer spacing after added tokens and Trainer broadcast NoneType errors are unrelated." - }, - { - "left": "issue:43701", - "right": "issue:45083", - "accept": false, - "reason": "resume_from_checkpoint key mismatch and qwen3_omni_moe feature-length behavior are different failure modes." + "reason": "CPU offload device mismatch in bark-small and AMP crash in LwDetrImageLoss are unrelated." } ] }, "evaluator_result": { - "accept": true, - "feedback": "The cluster summary is well grounded in the packet. The only true duplicate pair is the exact-title match between issue:45310 and issue:45478, and the soft-edge rejections stay conservative by relying on clearly different titles/models/subsystems. No mergeability concerns apply here since these are issues, not PRs." + "accept": false, + "feedback": "The summary is too strong for the packet. Although issue 45310 and 45478 have the highest similarity and matching titles, the packet marks that pair as deterministic_accept=false, so it does not support calling it a clear/true duplicate pair. There are no explicit links or stronger evidence establishing any duplicate cluster, so the claim of \"the only clear duplicate pair\" and the 0.97 confidence overstate the evidence. A more conservative read would be that the cluster is mostly singleton-like with one highly similar issue pair, not a confirmed duplicate." }, "evaluator_used": true, - "retried": false, - "accepted_nontrivial_soft_edge": true, + "retried": true, + "accepted_nontrivial_soft_edge": false, "error_kind": null, "error_message": null }, @@ -8238,50 +8261,51 @@ "component_index": 1, "component_count": 17, "review_unit_index": 26, - "review_unit_count": 49, + "review_unit_count": 50, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-427", + "cluster_id": "cluster-10000-428", "nodes": [ "issue:10000", "issue:15354", - "issue:25251", + "issue:28218", "issue:29942", "issue:32090", "issue:33357", + "issue:34689", "issue:35141", "issue:37428", + "issue:40444", "issue:41720", "issue:42490", "issue:43278", "issue:43329", "issue:43381", "issue:43412", + "issue:43526", "issue:43550", "issue:43575", - "issue:43653", + "issue:43701", "issue:43723", "issue:43873", "issue:43927", "issue:43976", "issue:44448", "issue:44462", - "issue:44466", "issue:44554", "issue:44561", "issue:44568", - "issue:44743", "issue:44779", "issue:44857", "issue:44977", "issue:44987", + "issue:45072", "issue:45083", "issue:45198", "issue:45229", - "issue:45245", "issue:45372", "issue:45406", "issue:45440", @@ -8295,231 +8319,245 @@ "issue:45823" ], "soft_pairs": [ + "issue:34689|issue:35141", + "issue:44857|issue:45072", + "issue:44448|issue:44977", + "issue:40444|issue:43526", + "issue:28218|issue:32090", + "issue:43701|issue:45083", "issue:43723|issue:44554", + "issue:29942|issue:35141", "issue:43550|issue:45083", "issue:32090|issue:33357", - "issue:29942|issue:35141", - "issue:44561|issue:44779", "issue:43873|issue:45229", + "issue:44561|issue:44779", "issue:43927|issue:45698", "issue:44462|issue:44561", "issue:41720|issue:45229", - "issue:44977|issue:45563", "issue:37428|issue:44987", + "issue:44977|issue:45563", "issue:43278|issue:43329", - "issue:42490|issue:43575", "issue:44568|issue:45491", - "issue:10000|issue:15354", + "issue:42490|issue:43575", "issue:45372|issue:45588", + "issue:10000|issue:15354", "issue:45198|issue:45803", "issue:43412|issue:45588", - "issue:44857|issue:45588", "issue:45372|issue:45406", + "issue:44857|issue:45588", "issue:43976|issue:45440", "issue:45459|issue:45712", "issue:45588|issue:45823", - "issue:43278|issue:43381", - "issue:44448|issue:44466", - "issue:45245|issue:45563", - "issue:43653|issue:44743", - "issue:25251|issue:32090" + "issue:43278|issue:43381" ], - "prepared_review_unit_hash": "d160c0bc3070f8930ae10d529e24e3dae1a792fedd79a965385c5337404e8ebf", + "prepared_review_unit_hash": "dc37b8bed06cac499fef760da2a986dddc02b2f31e1755541a2fcf30e6442f0e", "component_budget": { - "node_count": 427, - "item_count": 427, - "soft_pair_count": 1576, - "serialized_chars": 816182, - "estimated_input_tokens": 204046, - "estimated_eval_tokens": 408348 + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 }, "budget": { - "node_count": 47, - "item_count": 47, - "soft_pair_count": 28, - "serialized_chars": 35788, - "estimated_input_tokens": 8947, - "estimated_eval_tokens": 18150 + "node_count": 48, + "item_count": 48, + "soft_pair_count": 30, + "serialized_chars": 36869, + "estimated_input_tokens": 9218, + "estimated_eval_tokens": 18692 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The set is mostly a grab bag of unrelated transformer bugs and regressions; none of the soft pairs look like the same concrete underlying issue, so I found no convincing duplicate cluster or single canonical item.", - "confidence": 0.91, - "canonical_issue_reason": null, + "summary": "This cluster is highly heterogeneous: the pairs share broad themes like loading, tokenization, attention, or dtype issues, but they do not appear to be the same underlying bug/change. I would not merge any of the soft edges.", + "confidence": 0.94, + "canonical_issue_reason": "If a single representative is needed, issue:45698 is the broadest and most centrally discussed loading-related bug in the set (custom module resolution after save/load).", "canonical_pr_reason": null, - "best_issue_reason": null, + "best_issue_reason": "issue:45698 is the best single issue candidate because it has the strongest discussion/inbound-reference signal and a general save_pretrained/from_pretrained failure mode that is easier to use as an umbrella reference than the other, more specialized reports.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:43723", - "right": "issue:44554", + "left": "issue:34689", + "right": "issue:35141", "accept": false, - "reason": "Tokenizer-loading v5 regression and MPS attention correctness are different subsystems and failure modes." + "reason": "Different bugs: model loading for Llama 3.2 Vision vs embedding reinitialization after resizing tokens." }, { - "left": "issue:43550", + "left": "issue:44857", + "right": "issue:45072", + "accept": false, + "reason": "Both involve dtype/AMP, but they hit different models and failure modes: LwDetr float16 CUDA crash vs bfloat16 inference mismatch." + }, + { + "left": "issue:44448", + "right": "issue:44977", + "accept": false, + "reason": "Different models and code paths: Pegasus v4/v5 output drift vs Qwen3.5 flash-attention generation failure." + }, + { + "left": "issue:40444", + "right": "issue:43526", + "accept": false, + "reason": "Different layers: Qwen2.5-VL iterable-dataset multi-image failure vs BeitImageProcessorFast reduce_labels bug." + }, + { + "left": "issue:28218", + "right": "issue:32090", + "accept": false, + "reason": "Tokenizer added-space behavior is unrelated to the Trainer _gpu_broadcast_one NoneType error." + }, + { + "left": "issue:43701", "right": "issue:45083", "accept": false, - "reason": "torch.compile/SDPA failure in Bamba is unrelated to a helper output-length bug in qwen3_omni_moe." + "reason": "Checkpoint resume key mismatch is unrelated to the qwen3_omni_moe helper-length bug." }, { - "left": "issue:32090", - "right": "issue:33357", + "left": "issue:43723", + "right": "issue:44554", "accept": false, - "reason": "Trainer broadcast NoneType error and MacOS CLIP bus error are unrelated runtime failures." + "reason": "AutoTokenizer v5 loading regression is unrelated to the MPS attention head-dimension correctness issue." }, { "left": "issue:29942", "right": "issue:35141", "accept": false, - "reason": "Flash Attention 2 test failures do not match the token-embedding post_init reinitialization bug." + "reason": "Flash Attention 2 test failures are unrelated to post_init reinitializing resized embeddings." }, { - "left": "issue:44561", - "right": "issue:44779", + "left": "issue:43550", + "right": "issue:45083", + "accept": false, + "reason": "torch.compile/SDPA failure on Bamba-9B-v2 is not the same as a qwen3_omni_moe feature-length helper bug." + }, + { + "left": "issue:32090", + "right": "issue:33357", "accept": false, - "reason": "Removal of is_torch_fx_available and Deepseek tokenizer regressions affect different code paths." + "reason": "Trainer broadcast type error and MacOS bus error on a CLIP model are different failures." }, { "left": "issue:43873", "right": "issue:45229", "accept": false, - "reason": "Quantization offloading behavior and multi-GPU OOM are different problems, not the same bug." + "reason": "Quantization/offloading behavior and Gemma4 multi-GPU CUDA OOM are distinct problems." + }, + { + "left": "issue:44561", + "right": "issue:44779", + "accept": false, + "reason": "Removal of is_torch_fx_available breaking trust_remote_code is unrelated to DeepSeek tokenizer correctness." }, { "left": "issue:43927", "right": "issue:45698", "accept": false, - "reason": "Both involve save/load flows, but one loses custom token IDs and the other loads the wrong custom module." + "reason": "Both involve save/load behavior, but one is custom token IDs in config and the other is wrong custom module resolution; different root causes." }, { "left": "issue:44462", "right": "issue:44561", "accept": false, - "reason": "AutoTokenizer ignoring tokenizer.json is unrelated to the trust_remote_code breakage from removed API." + "reason": "AutoTokenizer ignoring tokenizer.json is unrelated to the v5 removal of is_torch_fx_available." }, { "left": "issue:41720", "right": "issue:45229", "accept": false, - "reason": "Qwen3 device-mapping assertion failure is not the same as Gemma4 multi-GPU OOM." - }, - { - "left": "issue:44977", - "right": "issue:45563", - "accept": false, - "reason": "Flash-attention generation failure and stale paging warning are different generation issues." + "reason": "Qwen3 auto device-mapping cuda assert is not the same bug as Gemma4 multi-GPU OOM." }, { "left": "issue:37428", "right": "issue:44987", "accept": false, - "reason": "Both are loading/import regressions, but they point to different missing dependencies and model-entry failures." + "reason": "Missing _flash_supports_window_size import and loading physical-intelligence/fast on 5.1.0 are not clearly the same concrete failure." }, { - "left": "issue:43278", - "right": "issue:43329", + "left": "issue:44977", + "right": "issue:45563", "accept": false, - "reason": "Embedding dtype drift and undefined video-branch variables are unrelated bugs." + "reason": "Flash-attention generation corruption and a stale num_return_sequences warning are unrelated." }, { - "left": "issue:42490", - "right": "issue:43575", + "left": "issue:43278", + "right": "issue:43329", "accept": false, - "reason": "Local-path import warning and tp-related OOM do not share the same underlying cause." + "reason": "Embedding dtype drift in eval mode is unrelated to the undefined helper variables in the video token-count path." }, { "left": "issue:44568", "right": "issue:45491", "accept": false, - "reason": "Tokenizer special-token insertion bug is unrelated to Gemma3 NaN embeddings in attention." + "reason": "Tokenizer add_special_tokens regression is unrelated to Gemma3 NaN embeddings from sliding-window attention." }, { - "left": "issue:10000", - "right": "issue:15354", + "left": "issue:42490", + "right": "issue:43575", "accept": false, - "reason": "German DistilBertModel issue and ViT torch.jit GeneratorExp export error are unrelated legacy reports." + "reason": "Import warning for trailing-slash local paths and tp-related OOM on Qwen2-57B are different issues." }, { "left": "issue:45372", "right": "issue:45588", "accept": false, - "reason": "Gemma4 processor import failure and flash_attention s_aux=None crash are separate Gemma4 code paths." + "reason": "Gemma4 processor import failure from mistral_common is unrelated to flash_attention crashing on s_aux=None." + }, + { + "left": "issue:10000", + "right": "issue:15354", + "accept": false, + "reason": "German DistilBertModel issue and GeneratorExp/jit.script export failure are unrelated." }, { "left": "issue:45198", "right": "issue:45803", "accept": false, - "reason": "Wav2Vec2 save/tokenization failure and bare-except tensor conversion bug are different defects." + "reason": "Wav2Vec2 save_pretrained/tokenization failure is unrelated to the bare except swallowing errors in FuyuBatchFeature.convert_to_tensors()." }, { "left": "issue:43412", "right": "issue:45588", "accept": false, - "reason": "Executorch export problems for gemma3n do not match the sink-less flash_attention crash." + "reason": "Gemma3n executorch export problems are unrelated to the flash_attention s_aux=None crash." }, { - "left": "issue:44857", - "right": "issue:45588", + "left": "issue:45372", + "right": "issue:45406", "accept": false, - "reason": "CUDA AMP loss crash in LwDetrImageLoss is unrelated to flash_attention's s_aux=None AttributeError." + "reason": "Both are Gemma4-related, but one is a processor import error and the other is a missing _tokenizer attribute in serve; different concrete bugs." }, { - "left": "issue:45372", - "right": "issue:45406", + "left": "issue:44857", + "right": "issue:45588", "accept": false, - "reason": "Processor import dependency failure and missing _tokenizer in serve are both Gemma4-related but not the same bug." + "reason": "LwDetr float16 CUDA crash is unrelated to the flash_attention s_aux=None AttributeError." }, { "left": "issue:43976", "right": "issue:45440", "accept": false, - "reason": "Python-version compatibility regression and DeepSeekV3MoE divergence are unrelated." + "reason": "Python-version compatibility breakage is unrelated to DeepseekV3MoE diverging from remote implementation." }, { "left": "issue:45459", "right": "issue:45712", "accept": false, - "reason": "Tokenizer exception masking and dummy class cleanup are separate maintenance issues." + "reason": "Hidden tokenizer errors from protobuf handling are unrelated to leftover dummy classes in dummy_pt_objects.py." }, { "left": "issue:45588", "right": "issue:45823", "accept": false, - "reason": "flash_attention s_aux=None crash and Gemma4 PLE device mismatch are different forward-pass failures." + "reason": "flash_attention sink-less-model crash and Gemma4 PLE device mismatch under auto device map are different code paths." }, { "left": "issue:43278", "right": "issue:43381", "accept": false, - "reason": "BF16/FP32 embedding drift is not the same as gradient-checkpointing eval-mode rejection." - }, - { - "left": "issue:44448", - "right": "issue:44466", - "accept": false, - "reason": "Pegasus v4/v5 output drift and lm_head.weight serialization inconsistency are both regressions, but not one concrete bug." - }, - { - "left": "issue:45245", - "right": "issue:45563", - "accept": false, - "reason": "Category-limit runtime error and stale num_return_sequences warning are unrelated." - }, - { - "left": "issue:43653", - "right": "issue:44743", - "accept": false, - "reason": "BigBirdTokenizer special-token registration and Qwen3.5 recurrent-state reset are different models and failures." - }, - { - "left": "issue:25251", - "right": "issue:32090", - "accept": false, - "reason": "Pipeline top_k nesting behavior and Trainer broadcast NoneType error are unrelated." + "reason": "Embedding dtype regression in eval mode is unrelated to gradient checkpointing being disallowed in eval mode." } ] }, @@ -8535,14 +8573,15 @@ "component_index": 1, "component_count": 17, "review_unit_index": 27, - "review_unit_count": 49, + "review_unit_count": 50, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-427", + "cluster_id": "cluster-10000-428", "nodes": [ + "issue:25251", "issue:29127", "issue:29942", "issue:30064", @@ -8550,7 +8589,6 @@ "issue:33290", "issue:33357", "issue:34567", - "issue:36010", "issue:39692", "issue:42490", "issue:42915", @@ -8561,12 +8599,10 @@ "issue:43278", "issue:43381", "issue:43421", - "issue:43526", "issue:43611", "issue:43653", "issue:43701", "issue:43756", - "issue:43761", "issue:43824", "issue:43873", "issue:43994", @@ -8576,23 +8612,29 @@ "issue:44279", "issue:44361", "issue:44442", + "issue:44448", "issue:44462", + "issue:44466", "issue:44479", "issue:44610", + "issue:44743", "issue:44871", "issue:44977", "issue:45198", + "issue:45245", "issue:45372", "issue:45412", "issue:45440", "issue:45459", "issue:45491", - "issue:45593", - "issue:45834", - "issue:45859", + "issue:45563", "issue:46121" ], "soft_pairs": [ + "issue:44448|issue:44466", + "issue:45245|issue:45563", + "issue:43653|issue:44743", + "issue:25251|issue:32090", "issue:43065|issue:43994", "issue:44206|issue:44479", "issue:45372|issue:45459", @@ -8601,12 +8643,12 @@ "issue:43611|issue:46121", "issue:44871|issue:44977", "issue:44361|issue:45198", - "issue:32090|issue:33290", + "issue:42915|issue:42994", "issue:42490|issue:43873", + "issue:32090|issue:33290", "issue:33357|issue:39692", - "issue:42915|issue:42994", - "issue:29127|issue:34567", "issue:29942|issue:30064", + "issue:29127|issue:34567", "issue:43994|issue:45412", "issue:44265|issue:45198", "issue:43066|issue:44462", @@ -8616,222 +8658,204 @@ "issue:43653|issue:44442", "issue:43381|issue:43701", "issue:44610|issue:45412", - "issue:43824|issue:44279", - "issue:45440|issue:45593", - "issue:29127|issue:33357", - "issue:43761|issue:45859", - "issue:29942|issue:36010", - "issue:43526|issue:45834", - "issue:43526|issue:44186" + "issue:43824|issue:44279" ], - "prepared_review_unit_hash": "61899b4e721b937adc1a3540ead6a39356d9ea6257243b2c88a4d92260a8dc65", + "prepared_review_unit_hash": "25107c1da88ee4c6c20c70e063f7659e4b7fb1f3a40fbc575cfdced30ecfe85b", "component_budget": { - "node_count": 427, - "item_count": 427, - "soft_pair_count": 1576, - "serialized_chars": 816182, - "estimated_input_tokens": 204046, - "estimated_eval_tokens": 408348 + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 30, - "serialized_chars": 36632, - "estimated_input_tokens": 9158, - "estimated_eval_tokens": 18572 + "soft_pair_count": 28, + "serialized_chars": 36089, + "estimated_input_tokens": 9023, + "estimated_eval_tokens": 18302 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The cluster is a heterogeneous mix of unrelated bug reports across tokenization, model loading, quantization, training, and vision/audio pipelines. None of the soft pairs look like the same underlying defect, so all soft edges are rejected.", - "confidence": 0.92, - "canonical_issue_reason": "issue:44462 is the broadest infrastructure-level bug in the set: AutoTokenizer ignoring the repository\u2019s tokenizer.json. It is more central and reusable as a representative issue than the many model-specific regressions, though this cluster does not form a true duplicate group.", + "summary": "The cluster is a grab bag of unrelated Transformers issues; none of the soft pairs look like true duplicates or the same fix. I\u2019d use the core tokenizer-loading bug as the representative issue, but with low confidence because the set is highly heterogeneous.", + "confidence": 0.44, + "canonical_issue_reason": "issue:44462 is the clearest, most generally useful concrete bug report here: a core AutoTokenizer loading problem with a specific failure mode and broad relevance across models.", "canonical_pr_reason": null, - "best_issue_reason": "issue:44462 is the best single representative because it concerns core tokenizer resolution/loading and has wider impact than the mostly model- or feature-specific issues elsewhere in the cluster.", + "best_issue_reason": "issue:44462 is the best representative issue because it is specific, reproducible, and centered on a core library code path rather than a model-specific edge case or an ancillary regression.", "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:44448", + "right": "issue:44466", + "accept": false, + "reason": "Different problems: output divergence for a model checkpoint vs inconsistent lm_head.weight serialization/device-dependent saving." + }, + { + "left": "issue:45245", + "right": "issue:45563", + "accept": false, + "reason": "Unrelated failures: category cardinality error in preprocessing vs a stale warning in paged generate()." + }, + { + "left": "issue:43653", + "right": "issue:44743", + "accept": false, + "reason": "Different subsystems: BigBirdTokenizer special-token registration vs Qwen3.5 recurrent-state caching behavior." + }, + { + "left": "issue:25251", + "right": "issue:32090", + "accept": false, + "reason": "Top-k pipeline nesting bug is unrelated to Trainer broadcast errors involving NoneType." + }, { "left": "issue:43065", "right": "issue:43994", "accept": false, - "reason": "Different problems: dummy Conv2d in Sam3PixelDecoder vs SigLIP2 model/processor mismatch and quantization failure." + "reason": "Sam3PixelDecoder dummy Conv2d issue is not the same as SigLIP2 bad outputs from AutoModel/pipeline loading." }, { "left": "issue:44206", "right": "issue:44479", "accept": false, - "reason": "Unrelated regressions: unsupported center arg in LasrFeatureExtractor vs video-input regression in Qwen multimodal models." + "reason": "Feature extractor center-arg crash and Qwen VL video regression are different code paths and failure modes." }, { "left": "issue:45372", "right": "issue:45459", "accept": false, - "reason": "Both are import-related, but one is a Gemma 4 processor dependency failure and the other is tokenizer error masking when protobuf is missing." + "reason": "Gemma 4 loading import error from mistral_common is distinct from tokenizer error masking when protobuf is missing." }, { "left": "issue:44186", "right": "issue:45198", "accept": false, - "reason": "Both involve tokenization, but one is LayoutLMv2 NER/padding crashes and the other is Wav2Vec2 save_pretrained/tokenization failure." + "reason": "LayoutLMv2Tokenizer NER/padding crash is unrelated to Wav2Vec2 save_pretrained/tokenization failure." }, { "left": "issue:43065", "right": "issue:45412", "accept": false, - "reason": "Unrelated: dummy Conv2d in Sam3PixelDecoder vs RT-DETR memory not being released on deletion." + "reason": "A dummy layer in Sam3PixelDecoder is not the same as RT-DETR models leaking memory after deletion." }, { "left": "issue:43611", "right": "issue:46121", "accept": false, - "reason": "Different scope: base_model_prefix loading breakage in v5 vs rope-parameter conversion TypeError on JSON-loaded lists." + "reason": "Base-model-prefix loading breakage in v5 is unrelated to convert_rope_params_to_dict rejecting a JSON-loaded list." }, { "left": "issue:44871", "right": "issue:44977", "accept": false, - "reason": "Both are Gemma/Qwen generation issues, but one is eos_token_id config mismatch and the other is flash-attention generation failure." + "reason": "Gemma eos_token_id mismatch and Qwen3.5 flash-attention generation issues are separate model/config problems." }, { "left": "issue:44361", "right": "issue:45198", "accept": false, - "reason": "Different tokenization failures: MLukeTokenizer task AttributeError vs Wav2Vec2 save_pretrained/tokenization problems." + "reason": "MLukeTokenizer AttributeError on tasks is not the same as Wav2Vec2 save/tokenization failure." }, { - "left": "issue:32090", - "right": "issue:33290", + "left": "issue:42915", + "right": "issue:42994", "accept": false, - "reason": "Trainer NoneType broadcast error is a distributed-training bug; adafactor/deepspeed OOM is a memory usage issue." + "reason": "FineGrainedFP8Config conversion failure and quantized model saving are different quantization workflows." }, { "left": "issue:42490", "right": "issue:43873", "accept": false, - "reason": "Local-path import warning with trailing slash is unrelated to offloading not working with quantization." + "reason": "Trailing-slash local model import warning is unrelated to offloading behavior with quantization." + }, + { + "left": "issue:32090", + "right": "issue:33290", + "accept": false, + "reason": "NoneType broadcast error in Trainer is unrelated to Adafactor/deepspeed OOM." }, { "left": "issue:33357", "right": "issue:39692", "accept": false, - "reason": "Different failures: MacOS bus error for CLIP vs SigLIP2 docs example/model-processor mismatch and quantization error." + "reason": "MacOS bus error for a pretrained CLIP model is not the same as a documentation example with model/processor mismatch and quantization issues." }, { - "left": "issue:42915", - "right": "issue:42994", + "left": "issue:29942", + "right": "issue:30064", "accept": false, - "reason": "Both mention quantization, but one is Qwen3 MoE failing with FineGrainedFP8Config and the other is quantized model saving failure." + "reason": "Flash Attention 2 test failures and void segmentation-map processing are different failures in different areas." }, { "left": "issue:29127", "right": "issue:34567", "accept": false, - "reason": "Different Trainer-adjacent issues: LayoutLMv3 error messaging vs TrainerState num_input_tokens_seen not updating." - }, - { - "left": "issue:29942", - "right": "issue:30064", - "accept": false, - "reason": "Flash Attention 2 test failures are unrelated to void segmentation map processing in the image processor." + "reason": "LayoutLMv3 error-message clarity is unrelated to TrainerState num_input_tokens_seen not updating." }, { "left": "issue:43994", "right": "issue:45412", "accept": false, - "reason": "SigLIP2 inference/processor mismatch is unrelated to RT-DETR memory leak on deletion." + "reason": "SigLIP2 nonsense outputs and RT-DETR memory retention are unrelated bugs." }, { "left": "issue:44265", "right": "issue:45198", "accept": false, - "reason": "torch.export failure with torch_compilable_check is a compilation/export bug, not a Wav2Vec2 save/tokenization issue." + "reason": "torch.export exportability with torch_compilable_check is not the same as Wav2Vec2 tokenization/save_pretrained failure." }, { "left": "issue:43066", "right": "issue:44462", "accept": false, - "reason": "Tokenizer decoder type in v5 is not the same as AutoTokenizer ignoring tokenizer.json from the repository." + "reason": "Wrong tokenizer decoder type in v5 and ignoring tokenizer.json from the repo are both tokenizer-related but not the same underlying defect." }, { "left": "issue:43756", "right": "issue:45440", "accept": false, - "reason": "Smollm3 RoPE-layer count bug is unrelated to DeepseekV3MoE divergence from the remote implementation." + "reason": "Smollm3 RoPE-layer dropping and DeepseekV3MoE divergence are separate architecture/implementation mismatches." }, { "left": "issue:43257", "right": "issue:43278", "accept": false, - "reason": "Both involve model state handling, but one is Qwen3 MoE weights not converted under accelerate+deepspeed and the other is embedding dtype drift in eval." + "reason": "Qwen3 MOE weight conversion with accelerate/deepspeed is unrelated to embedding dtype changing in evaluation." }, { "left": "issue:43421", "right": "issue:45491", "accept": false, - "reason": "Tokenizer special-token post-processor update bug is unrelated to Gemma3 NaN embeddings with sliding-window attention." + "reason": "Runtime special-token post-processor staleness and Gemma3 NaN embeddings in mixed-length batches are different bugs." }, { "left": "issue:43653", "right": "issue:44442", "accept": false, - "reason": "BigBirdTokenizer special-token registration issue is not the same as AutoTokenizer failing to load a FastSpeech2ConformerTokenizer." + "reason": "BigBirdTokenizer decode-output bug and AutoTokenizer failing to load FastSpeech2ConformerTokenizer are distinct tokenizer issues." }, { "left": "issue:43381", "right": "issue:43701", "accept": false, - "reason": "Gradient checkpointing in eval mode and resume_from_checkpoint key mismatch are different training/resume bugs." + "reason": "Gradient checkpointing in eval mode and resume_from_checkpoint key mismatch do not share the same failure mechanism." }, { "left": "issue:44610", "right": "issue:45412", "accept": false, - "reason": "OmDet-Turbo processor size mismatch is unrelated to RT-DETR memory release behavior." + "reason": "Processor/model image-size mismatch is unrelated to RT-DETR memory not being released." }, { "left": "issue:43824", "right": "issue:44279", "accept": false, - "reason": "ImportError for Qwen2_5_VLForConditionalGeneration is not the same as a generic dependency issue with transformers." - }, - { - "left": "issue:45440", - "right": "issue:45593", - "accept": false, - "reason": "DeepseekV3MoE implementation divergence is unrelated to D-FINE not using auxiliary losses when denoising is off." - }, - { - "left": "issue:29127", - "right": "issue:33357", - "accept": false, - "reason": "LayoutLMv3 error-message clarity and CLIP MacOS bus error are entirely different issues." - }, - { - "left": "issue:43761", - "right": "issue:45859", - "accept": false, - "reason": "CLIPVisionModel hidden_states regression is unrelated to Qwen3_5Moe CPU offload incompatibility." - }, - { - "left": "issue:29942", - "right": "issue:36010", - "accept": false, - "reason": "Flash Attention 2 test failures do not match the GenerationMixin import error." - }, - { - "left": "issue:43526", - "right": "issue:45834", - "accept": false, - "reason": "BeitImageProcessorFast reduce_labels bug is unrelated to Kosmos2.5 OCR index errors." - }, - { - "left": "issue:43526", - "right": "issue:44186", - "accept": false, - "reason": "Both are processor/tokenization adjacent, but they concern different processors and failure modes with no shared code-path indicated." + "reason": "Missing Qwen2_5_VLForConditionalGeneration import is a specific API availability bug, not the generic dependency issue described by the other report." } ] }, @@ -8847,26 +8871,24 @@ "component_index": 1, "component_count": 17, "review_unit_index": 28, - "review_unit_count": 49, + "review_unit_count": 50, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-427", + "cluster_id": "cluster-10000-428", "nodes": [ "issue:29127", - "issue:33290", - "issue:34567", - "issue:34689", + "issue:29942", + "issue:33357", "issue:35141", + "issue:36010", "issue:42490", "issue:42915", - "issue:43232", "issue:43262", "issue:43334", "issue:43381", - "issue:43388", "issue:43441", "issue:43526", "issue:43531", @@ -8875,11 +8897,11 @@ "issue:43701", "issue:43723", "issue:43761", - "issue:43873", "issue:43931", "issue:44062", "issue:44077", "issue:44079", + "issue:44186", "issue:44206", "issue:44387", "issue:44442", @@ -8889,7 +8911,6 @@ "issue:44743", "issue:44811", "issue:44857", - "issue:44898", "issue:44977", "issue:45003", "issue:45072", @@ -8902,9 +8923,18 @@ "issue:45491", "issue:45584", "issue:45588", + "issue:45593", + "issue:45834", + "issue:45859", "issue:45907" ], "soft_pairs": [ + "issue:45440|issue:45593", + "issue:29127|issue:33357", + "issue:29942|issue:36010", + "issue:43761|issue:45859", + "issue:43526|issue:45834", + "issue:43526|issue:44186", "issue:45103|issue:45372", "issue:44625|issue:45584", "issue:29127|issue:35141", @@ -8913,9 +8943,9 @@ "issue:44977|issue:45245", "issue:42915|issue:43931", "issue:43526|issue:44857", + "issue:44464|issue:45491", "issue:42490|issue:43381", "issue:43575|issue:43701", - "issue:44464|issue:45491", "issue:43262|issue:44206", "issue:43575|issue:44387", "issue:44743|issue:44811", @@ -8925,211 +8955,198 @@ "issue:43761|issue:44079", "issue:45003|issue:45440", "issue:43723|issue:45479", - "issue:44479|issue:45290", - "issue:44442|issue:44898", - "issue:43232|issue:43388", - "issue:33290|issue:34689", - "issue:33290|issue:34567", - "issue:33290|issue:35141", - "issue:42490|issue:44743", - "issue:43381|issue:43873" + "issue:44479|issue:45290" ], - "prepared_review_unit_hash": "f531783f82b918b49a6132bc0c4f1d12ff9db0c189952121312b919e439fb0bf", + "prepared_review_unit_hash": "37613528d34c6cf8bf92e37cf50bff71ee9a8a0dbd169f9dcfc33492c5d57937", "component_budget": { - "node_count": 427, - "item_count": 427, - "soft_pair_count": 1576, - "serialized_chars": 816182, - "estimated_input_tokens": 204046, - "estimated_eval_tokens": 408348 + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 28, - "serialized_chars": 36258, - "estimated_input_tokens": 9065, - "estimated_eval_tokens": 18386 + "soft_pair_count": 27, + "serialized_chars": 35861, + "estimated_input_tokens": 8966, + "estimated_eval_tokens": 18188 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is heterogeneous: the soft-similarity pairs are mostly false positives across unrelated bugs, model families, and subsystems. No duplicate pair looks safe to merge, and there is no true PR cluster here.", - "confidence": 0.97, - "canonical_issue_reason": "No strong canonical duplicate exists; if one issue must anchor the cluster, #43873 is the most active and referenced, but it is not actually connected to the rest of the items.", + "summary": "All proposed soft edges look like false positives: the issues span unrelated models, tokenizers, processors, generation paths, and utility bugs rather than the same underlying defect.", + "confidence": 0.96, + "canonical_issue_reason": "No canonical issue stands out because this is not a true duplicate cluster; the items cover many unrelated subsystems and symptoms.", "canonical_pr_reason": null, - "best_issue_reason": "#43873 is the strongest standalone issue in the set by activity and inbound references, but it still does not represent the rest of the cluster.", + "best_issue_reason": "If forced to pick one representative issue, issue 43761 is a clear, well-scoped regression report, but it is not representative of the rest of the heterogeneous set.", "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:45440", + "right": "issue:45593", + "accept": false, + "reason": "Both are MoE-related, but one is a model-implementation divergence and the other is an auxiliary-loss behavior change; different bugs." + }, + { + "left": "issue:29127", + "right": "issue:33357", + "accept": false, + "reason": "LayoutLMv3 error messaging vs. a CLIP Mac bus error are unrelated issues in different models and failure modes." + }, + { + "left": "issue:29942", + "right": "issue:36010", + "accept": false, + "reason": "FlashAttention 2 test failures and a GenerationMixin import error are different code paths and symptoms." + }, + { + "left": "issue:43761", + "right": "issue:45859", + "accept": false, + "reason": "Both are regressions, but CLIPVision hidden_states handling and Qwen3_5Moe CPU-offload rotary embedding are separate bugs." + }, + { + "left": "issue:43526", + "right": "issue:45834", + "accept": false, + "reason": "BEiT reduce_labels behavior and Kosmos2.5 OCR index errors affect different processors and failure mechanisms." + }, + { + "left": "issue:43526", + "right": "issue:44186", + "accept": false, + "reason": "These are different image/text processor bugs; they do not point to the same underlying defect." + }, { "left": "issue:45103", "right": "issue:45372", "accept": false, - "reason": "Different failures in different areas: auto-docstring annotation parsing vs Gemma 4 processor import error." + "reason": "Auto-docstring crashes with future annotations is unrelated to the Gemma 4/ReasoningEffort import failure." }, { "left": "issue:44625", "right": "issue:45584", "accept": false, - "reason": "Unrelated bugs: Qwen3.5 config propagation vs Whisper empty-transcription generation after token alignment." + "reason": "Qwen3.5 num_labels propagation and Whisper empty-transcription generation are unrelated bugs." }, { "left": "issue:29127", "right": "issue:35141", "accept": false, - "reason": "Different subsystems entirely: LayoutLMV3 error messaging vs embedding reinitialization on resize." + "reason": "One is about clearer validation errors in LayoutLMv3; the other is output embedding reinitialization after resize." }, { "left": "issue:44442", "right": "issue:45072", "accept": false, - "reason": "Tokenizer loading failure and bf16 dtype mismatch are unrelated code paths." + "reason": "Tokenizer loading failure and bfloat16 dtype mismatch are different issues with no shared concrete code path." }, { "left": "issue:43334", "right": "issue:43531", "accept": false, - "reason": "Same model family, but different issues: missing pad_token_id on load vs sliding-window behavior in Qwen3-MoE." + "reason": "Both mention Qwen models, but one is a pad_token_id/config loading bug and the other is a sliding_window runtime issue." }, { "left": "issue:44977", "right": "issue:45245", "accept": false, - "reason": "Qwen3.5 flash-attention generation bug is unrelated to category-count limits." + "reason": "Flash-attention generation regression and the 2^24 category limit error are unrelated." }, { "left": "issue:42915", "right": "issue:43931", "accept": false, - "reason": "Different models and symptoms: FineGrainedFP8Config failure vs Qwen3-VL weight-shape mismatch." + "reason": "FineGrainedFP8Config failure and Qwen3-VL weight-shape mismatch are separate loading problems." }, { "left": "issue:43526", "right": "issue:44857", "accept": false, - "reason": "ImageProcessor label reduction bug is unrelated to LwDetrImageLoss AMP crashes." + "reason": "BEiT label reduction and LwDetrImageLoss AMP crash are unrelated vision bugs." + }, + { + "left": "issue:44464", + "right": "issue:45491", + "accept": false, + "reason": "Compiled-forward chunked generation inconsistency and Gemma3 NaNs with sliding-window attention are different problems." }, { "left": "issue:42490", "right": "issue:43381", "accept": false, - "reason": "Local-path import warning has nothing to do with gradient checkpointing in eval mode." + "reason": "Local-path import warnings and gradient checkpointing being disallowed in eval mode are unrelated." }, { "left": "issue:43575", "right": "issue:43701", "accept": false, - "reason": "OOM on tensor-parallel model loading is unrelated to checkpoint key mismatch." - }, - { - "left": "issue:44464", - "right": "issue:45491", - "accept": false, - "reason": "Both involve generation, but the concrete bugs differ: compiled forward inconsistency vs Gemma3 sliding-window NaNs." + "reason": "Tensor-parallel OOM and resume_from_checkpoint key mismatch are unrelated training/loading bugs." }, { "left": "issue:43262", "right": "issue:44206", "accept": false, - "reason": "Audio chat-template sampling-rate defaulting and LasrFeatureExtractor center-arg handling are different bugs." + "reason": "Audio chat-template sample-rate defaulting and an unsupported feature-extractor arg are different APIs and bugs." }, { "left": "issue:43575", "right": "issue:44387", "accept": false, - "reason": "Both are memory-related, but one is tensor-parallel load OOM and the other is int4 reserved-memory growth." + "reason": "Both involve memory, but one is load-time OOM with tensor parallelism and the other is int4 CUDA reserved-memory growth." }, { "left": "issue:44743", "right": "issue:44811", "accept": false, - "reason": "Recurrent-state reset in Qwen3.5 is unrelated to Whisper batch_decode skip_special_tokens handling." + "reason": "Qwen3.5 recurrent-state reset and Whisper batch_decode skip_special_tokens handling are unrelated." }, { "left": "issue:43646", "right": "issue:44077", "accept": false, - "reason": "Generic custom-model init regression and patchtsmixer post_init policy are not the same code-path problem." + "reason": "Custom model initialization breakage and patchtsmixer post_init policy are not the same issue." }, { "left": "issue:43441", "right": "issue:45588", "accept": false, - "reason": "Different flash-attention failures in different models and different stack traces." + "reason": "Ministral-3 FlashAttention failure and sink-less-model s_aux=None crash are different flash-attention bugs." }, { "left": "issue:44062", "right": "issue:45907", "accept": false, - "reason": "Tokenizer AddedToken keyword duplication has no relation to list_repo_files revision forwarding." + "reason": "AddedToken special-argument duplication and list_repo_files revision forwarding are unrelated utility issues." }, { "left": "issue:43761", "right": "issue:44079", "accept": false, - "reason": "One is CLIPVisionModel output behavior; the other is generic ModelOutput key assignment semantics." + "reason": "CLIPVision hidden_states regression and ModelOutput key assignment are different output-handling bugs." }, { "left": "issue:45003", "right": "issue:45440", "accept": false, - "reason": "Unsafe sys.modules access and DeepSeekV3 MoE divergence are unrelated." + "reason": "Unsafe sys.modules access in modeling_utils and DeepseekV3MoE divergence are unrelated." }, { "left": "issue:43723", "right": "issue:45479", "accept": false, - "reason": "Tokenizer loading regression and zero-loss sequence-classification bug are unrelated." + "reason": "AutoTokenizer loading failure and zero-loss single-label classification are different bug classes." }, { "left": "issue:44479", "right": "issue:45290", "accept": false, - "reason": "Video-input regression and assistant tool-call chat-template crash are different generation/preprocessing bugs." - }, - { - "left": "issue:44442", - "right": "issue:44898", - "accept": false, - "reason": "Tokenizer loading failure is unrelated to Perceiver interpolation at non-default image resolution." - }, - { - "left": "issue:43232", - "right": "issue:43388", - "accept": false, - "reason": "Generation kwargs sync issue and metrics-gathering label truncation are different distributed-training bugs." - }, - { - "left": "issue:33290", - "right": "issue:34689", - "accept": false, - "reason": "Adafactor+DeepSpeed OOM is unrelated to Llama 3.2 Vision model loading breakage." - }, - { - "left": "issue:33290", - "right": "issue:34567", - "accept": false, - "reason": "OOM in optimizer/deepspeed has no overlap with TrainerState token counter updates." - }, - { - "left": "issue:33290", - "right": "issue:35141", - "accept": false, - "reason": "These concern unrelated training-memory and embedding-resize behaviors." - }, - { - "left": "issue:42490", - "right": "issue:44743", - "accept": false, - "reason": "Trailing-slash import warning and recurrent-state reset are unrelated." - }, - { - "left": "issue:43381", - "right": "issue:43873", - "accept": false, - "reason": "Gradient checkpointing eval-mode restriction and quantization offloading problems are not the same bug." + "reason": "Qwen video-input regression and apply_chat_template tool-call crashes are separate code paths." } ] }, @@ -9145,33 +9162,35 @@ "component_index": 1, "component_count": 17, "review_unit_index": 29, - "review_unit_count": 49, + "review_unit_count": 50, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-427", + "cluster_id": "cluster-10000-428", "nodes": [ "issue:25251", "issue:28282", "issue:33290", + "issue:34567", "issue:34689", + "issue:35141", "issue:36010", "issue:36331", + "issue:42490", "issue:42907", "issue:43065", + "issue:43232", "issue:43295", "issue:43316", + "issue:43381", + "issue:43388", "issue:43493", "issue:43525", "issue:43540", - "issue:43550", - "issue:43606", "issue:43611", "issue:43643", - "issue:43701", - "issue:43716", "issue:43720", "issue:43749", "issue:43761", @@ -9180,19 +9199,17 @@ "issue:43873", "issue:43931", "issue:44077", - "issue:44162", "issue:44188", "issue:44190", "issue:44442", "issue:44479", "issue:44496", "issue:44514", - "issue:44534", "issue:44617", - "issue:44792", + "issue:44743", "issue:44855", "issue:44871", - "issue:44877", + "issue:44898", "issue:44913", "issue:44987", "issue:45137", @@ -9203,19 +9220,26 @@ "issue:46178" ], "soft_pairs": [ + "issue:43232|issue:43388", + "issue:44442|issue:44898", + "issue:33290|issue:34567", + "issue:33290|issue:34689", "issue:43316|issue:43525", + "issue:33290|issue:35141", + "issue:42490|issue:44743", + "issue:43381|issue:43873", "issue:43761|issue:44188", "issue:44442|issue:45803", "issue:34689|issue:43611", + "issue:44987|issue:46178", "issue:43720|issue:44442", "issue:44077|issue:44479", - "issue:44987|issue:46178", "issue:43931|issue:45216", - "issue:43493|issue:43827", "issue:42907|issue:45698", + "issue:43493|issue:43827", "issue:34689|issue:36010", - "issue:33290|issue:36331", "issue:43316|issue:45759", + "issue:33290|issue:36331", "issue:43749|issue:43931", "issue:43873|issue:44871", "issue:43643|issue:44913", @@ -9223,211 +9247,211 @@ "issue:44855|issue:45137", "issue:43540|issue:44514", "issue:25251|issue:28282", - "issue:43065|issue:44617", "issue:43295|issue:43825", - "issue:43550|issue:43701", - "issue:43316|issue:43716", - "issue:44162|issue:44792", - "issue:43316|issue:44877", - "issue:43606|issue:43701", - "issue:44077|issue:44534" + "issue:43065|issue:44617" ], - "prepared_review_unit_hash": "931b89f902217f75c64c354b1da10f6234fad0a8e8d6a61bc41bf64da53d74fb", + "prepared_review_unit_hash": "759edfabeb50972a369751d5e60860d6b63ae084478b931481be73a1703c8538", "component_budget": { - "node_count": 427, - "item_count": 427, - "soft_pair_count": 1576, - "serialized_chars": 816182, - "estimated_input_tokens": 204046, - "estimated_eval_tokens": 408348 + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 28, - "serialized_chars": 36067, - "estimated_input_tokens": 9017, - "estimated_eval_tokens": 18290 + "soft_pair_count": 29, + "serialized_chars": 36622, + "estimated_input_tokens": 9156, + "estimated_eval_tokens": 18568 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The set is highly heterogeneous, and every proposed soft pair appears to be a false duplicate: they usually share a broad subsystem (configs, loading, multimodal inputs, offloading) but not the same concrete bug or code path. No PRs are present to compare.", - "confidence": 0.91, - "canonical_issue_reason": "issue:43873 is the broadest and most actively discussed open bug in the set, so it is the best representative issue if one must be chosen.", + "summary": "This cluster is highly heterogeneous: the items span unrelated bugs in loading, generation, training, quantization, vision/video, and docs. None of the soft pairs looks like the same concrete underlying issue, so I would not merge any of them.", + "confidence": 0.98, + "canonical_issue_reason": "issue:43611 is the broadest representative of the cluster\u2019s dominant loading/regression theme, but even it is not a true duplicate of the others; the rest are mostly separate bugs across different subsystems.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43873 has the strongest triage signal overall: open status, multiple inbound references, and a broad user-impacting loading/offloading regression.", + "best_issue_reason": "issue:43611 is the best single representative because it targets a general model-loading regression with wider applicability than the more model-specific or peripheral reports.", "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:43232", + "right": "issue:43388", + "accept": false, + "reason": "Generation kwargs after sync_gpus is unrelated to metric gathering dropping tuple labels." + }, + { + "left": "issue:44442", + "right": "issue:44898", + "accept": false, + "reason": "Tokenizer loading failure and Perceiver interpolation failure are different code paths." + }, + { + "left": "issue:33290", + "right": "issue:34567", + "accept": false, + "reason": "Deepspeed/Adafactor OOM is not the same as TrainerState token-count bookkeeping." + }, + { + "left": "issue:33290", + "right": "issue:34689", + "accept": false, + "reason": "Optimizer OOM and model-loading regression are different failure modes." + }, { "left": "issue:43316", "right": "issue:43525", "accept": false, - "reason": "Both are config-related, but one is a Gemma3TextConfig API mismatch and the other is a missing pad_token_id on Llama4Config; different bugs." + "reason": "Gemma3 config API mismatch and missing Llama4 pad_token_id are distinct config bugs." + }, + { + "left": "issue:33290", + "right": "issue:35141", + "accept": false, + "reason": "Training OOM does not match embedding reinitialization in post_init." + }, + { + "left": "issue:42490", + "right": "issue:44743", + "accept": false, + "reason": "Local-path import warning is unrelated to Qwen3.5 recurrent-state reset behavior." + }, + { + "left": "issue:43381", + "right": "issue:43873", + "accept": false, + "reason": "Gradient checkpointing in eval mode is separate from quantization offloading behavior." }, { "left": "issue:43761", "right": "issue:44188", "accept": false, - "reason": "A CLIPVisionModel hidden_states regression and a torch.compile attention-kernel divergence are unrelated code paths." + "reason": "CLIP hidden_states regression and attention-kernel divergence are different issues." }, { "left": "issue:44442", "right": "issue:45803", "accept": false, - "reason": "Tokenizer load failure vs. a bare-except bug in Fuyu tensor conversion; different models and failure modes." + "reason": "Tokenizer loading failure is not the same as a bare-except bug in tensor conversion." }, { "left": "issue:34689", "right": "issue:43611", "accept": false, - "reason": "Both affect model loading, but one is a Llama 3.2 vision regression and the other is a base_model_prefix loading issue." + "reason": "Both are loading-related, but they concern different regressions and different triggering conditions/models." + }, + { + "left": "issue:44987", + "right": "issue:46178", + "accept": false, + "reason": "Model loading failure for a specific repo is unrelated to backend mapping missing tensorflow_text." }, { "left": "issue:43720", "right": "issue:44442", "accept": false, - "reason": "Accelerate packed-weight loading in BitNet vs. AutoTokenizer failing on FastSpeech2ConformerTokenizer; unrelated." + "reason": "Packed-weight unpacking in BitNet is unrelated to AutoTokenizer failing for FastSpeech2ConformerTokenizer." }, { "left": "issue:44077", "right": "issue:44479", "accept": false, - "reason": "PatchTSMixer post_init validation and Qwen video-input regression are different bugs." - }, - { - "left": "issue:44987", - "right": "issue:46178", - "accept": false, - "reason": "A model-loading regression for physical-intelligence/fast is not the same as the BACKENDS_MAPPING tensorflow_text error." + "reason": "Optional post_init enforcement and video-input regression are separate changes." }, { "left": "issue:43931", "right": "issue:45216", "accept": false, - "reason": "One is a Qwen3-VL weight-shape mismatch on load; the other is a Qwen3.5 save_pretrained checkpoint bug." + "reason": "Weight-shape mismatch at load time is different from save_pretrained producing a bad checkpoint." }, { - "left": "issue:43493", - "right": "issue:43827", + "left": "issue:42907", + "right": "issue:45698", "accept": false, - "reason": "SigLIP2 implementation discrepancy and stale docs mentioning pipeline() are unrelated." + "reason": "Saving dequantized models and loading the wrong custom module after save_pretrained are distinct bugs." }, { - "left": "issue:42907", - "right": "issue:45698", + "left": "issue:43493", + "right": "issue:43827", "accept": false, - "reason": "Both involve save/load behavior, but one is about failing to save dequantized models and the other about loading the wrong custom module after save_pretrained." + "reason": "Implementation discrepancy in SigLIP2 is unrelated to stale docs referencing pipeline()." }, { "left": "issue:34689", "right": "issue:36010", "accept": false, - "reason": "A broken Llama 3.2 model load and a GenerationMixin ImportError are different issues." + "reason": "Model-loading regression and GenerationMixin import error are different API breakages." }, { - "left": "issue:33290", - "right": "issue:36331", + "left": "issue:43316", + "right": "issue:45759", "accept": false, - "reason": "AdaFactor OOM under DeepSpeed and a CustomTrainer signature mismatch are unrelated training bugs." + "reason": "Config API discrepancy is not the same as failing to unwrap text_config in from_config." }, { - "left": "issue:43316", - "right": "issue:45759", + "left": "issue:33290", + "right": "issue:36331", "accept": false, - "reason": "Gemma3TextConfig API inconsistency and Qwen composite config unwrapping are different config-handling bugs." + "reason": "OOM in DeepSpeed/Adafactor does not match a compute_loss signature break." }, { "left": "issue:43749", "right": "issue:43931", "accept": false, - "reason": "FSDP CPU RAM efficient loading and Qwen3-VL weight-shape mismatch both involve loading, but not the same code-path problem." + "reason": "FSDP CPU-efficient loading and Qwen3-VL weight mismatch are both loading-related but not the same bug." }, { "left": "issue:43873", "right": "issue:44871", "accept": false, - "reason": "Quantization/offloading behavior and an eos_token_id config inconsistency are not duplicates." + "reason": "Quantization offloading is unrelated to eos_token_id configuration inconsistency." }, { "left": "issue:43643", "right": "issue:44913", "accept": false, - "reason": "trust_remote_code dropping fields and GPTNeoXConfig rotary_pct resetting on reload are separate config persistence bugs." + "reason": "trust_remote_code missing fields and GPTNeoX rotary_pct reload behavior are separate config issues." }, { "left": "issue:44190", "right": "issue:44496", "accept": false, - "reason": "Local dataset loading in an example script and an unrecognized custom model config are different problems." + "reason": "Dataset loading in an example script is unrelated to an unrecognized model config." }, { "left": "issue:44855", "right": "issue:45137", "accept": false, - "reason": "Python 3.13 torch.jit parsing failure and a DeepSpeed ZeRO3 deque IndexError are unrelated." + "reason": "Python 3.13 TorchScript parsing and ZeRO3 deque underflow are different runtime failures." }, { "left": "issue:43540", "right": "issue:44514", "accept": false, - "reason": "Qwen3OmniMoe video processing and Qwen2_5_VL batched chat-template padding are both multimodal, but the concrete bugs differ." + "reason": "Video-input processing in Qwen3OmniMoe is unrelated to batched chat-template padding crash." }, { "left": "issue:25251", "right": "issue:28282", "accept": false, - "reason": "Pipeline top_k output-shape change and missing PyTorch ImportError have nothing in common." - }, - { - "left": "issue:43065", - "right": "issue:44617", - "accept": false, - "reason": "Sam3PixelDecoder dummy Conv2d and Sam3Video CUDA OOM are different model issues." + "reason": "Pipeline top_k nesting bug and missing PyTorch import are entirely unrelated." }, { "left": "issue:43295", "right": "issue:43825", "accept": false, - "reason": "Processor/tokenizer regression and an incorrect pipeline() error message are not the same bug." - }, - { - "left": "issue:43550", - "right": "issue:43701", - "accept": false, - "reason": "torch.compile SDPA failure in Bamba-9B-v2 and a resume_from_checkpoint key mismatch are unrelated." + "reason": "processor.tokenizer regression is not the same as an error-message/doc wording bug about translation tasks." }, { - "left": "issue:43316", - "right": "issue:43716", + "left": "issue:43065", + "right": "issue:44617", "accept": false, - "reason": "Gemma3TextConfig API mismatch and Mistral-3 image-preprocessor dtype mismatch are different issues." - }, - { - "left": "issue:44162", - "right": "issue:44792", - "accept": false, - "reason": "ESM2 broad breakage and a janus test failure for image generation do not share a concrete code path." - }, - { - "left": "issue:43316", - "right": "issue:44877", - "accept": false, - "reason": "Gemma3TextConfig API discrepancy and strict config blocking granite_speech are distinct config-loading behaviors." - }, - { - "left": "issue:43606", - "right": "issue:43701", - "accept": false, - "reason": "CPU offload device mismatch and resume_from_checkpoint key mismatch are unrelated." - }, - { - "left": "issue:44077", - "right": "issue:44534", - "accept": false, - "reason": "PatchTSMixer post_init policy and non-persistent buffers filled with junk are separate v5 issues." + "reason": "Dummy Conv2d in Sam3PixelDecoder and Sam3Video OOM are different defects." } ] }, @@ -9443,13 +9467,13 @@ "component_index": 1, "component_count": 17, "review_unit_index": 30, - "review_unit_count": 49, + "review_unit_count": 50, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-427", + "cluster_id": "cluster-10000-428", "nodes": [ "issue:16998", "issue:22355", @@ -9458,39 +9482,39 @@ "issue:41950", "issue:42617", "issue:43295", + "issue:43316", "issue:43388", "issue:43404", "issue:43441", "issue:43452", "issue:43475", "issue:43504", - "issue:43525", - "issue:43531", + "issue:43550", "issue:43572", "issue:43575", + "issue:43606", "issue:43646", "issue:43673", "issue:43688", + "issue:43701", + "issue:43716", "issue:43742", "issue:43761", - "issue:43931", - "issue:44060", + "issue:44077", + "issue:44162", "issue:44242", - "issue:44464", "issue:44479", "issue:44496", - "issue:44521", "issue:44534", "issue:44617", "issue:44734", "issue:44743", "issue:44792", "issue:44871", + "issue:44877", "issue:44913", - "issue:44977", "issue:45127", "issue:45200", - "issue:45381", "issue:45406", "issue:45563", "issue:45584", @@ -9501,238 +9525,231 @@ "issue:46178" ], "soft_pairs": [ + "issue:43316|issue:43716", + "issue:43316|issue:44877", + "issue:44162|issue:44792", + "issue:43550|issue:43701", + "issue:44077|issue:44534", + "issue:43606|issue:43701", "issue:45803|issue:46170", "issue:43404|issue:43441", - "issue:42617|issue:43575", - "issue:44871|issue:45584", "issue:16998|issue:22355", + "issue:44871|issue:45584", + "issue:42617|issue:43575", "issue:43673|issue:44743", "issue:43575|issue:43742", - "issue:43295|issue:44871", "issue:43388|issue:43688", - "issue:43761|issue:44479", "issue:41950|issue:43475", + "issue:43761|issue:44479", + "issue:43295|issue:44871", "issue:43452|issue:44496", - "issue:44871|issue:46050", "issue:45127|issue:45200", + "issue:44871|issue:46050", "issue:43646|issue:44913", "issue:43572|issue:46178", "issue:43504|issue:45698", - "issue:32090|issue:35141", "issue:44734|issue:45406", + "issue:32090|issue:35141", "issue:44617|issue:45584", "issue:45563|issue:45584", - "issue:43761|issue:44242", - "issue:43525|issue:43531", - "issue:44792|issue:44977", - "issue:43931|issue:44060", - "issue:44479|issue:44734", - "issue:44534|issue:44913", - "issue:44464|issue:45563", - "issue:44521|issue:45381" + "issue:43761|issue:44242" ], - "prepared_review_unit_hash": "195bbe47f55afe7652464c47d5c7e15c49806e22529855bf8dfc48f4697ba447", + "prepared_review_unit_hash": "21a990489c997cac759d47a7a392da335992600365c86b1872f5798ff2212148", "component_budget": { - "node_count": 427, - "item_count": 427, - "soft_pair_count": 1576, - "serialized_chars": 816182, - "estimated_input_tokens": 204046, - "estimated_eval_tokens": 408348 + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 29, - "serialized_chars": 36486, - "estimated_input_tokens": 9122, - "estimated_eval_tokens": 18500 + "soft_pair_count": 28, + "serialized_chars": 36076, + "estimated_input_tokens": 9019, + "estimated_eval_tokens": 18294 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly a noisy mix of unrelated Transformers issues spanning loading, generation, multimodal, distributed, and config regressions. None of the soft pairs look like true duplicates or mergeable PR-equivalents.", - "confidence": 0.84, - "canonical_issue_reason": "Issue 45698 is the best anchor: it is open, has substantial discussion, and is a clear save/load regression with broad enough relevance to represent the cluster better than the narrower reports.", + "summary": "This cluster is heterogeneous: the issues span unrelated loading/config regressions, model-specific bugs, and serving/training edge cases. I don\u2019t see a single underlying duplicate bug or change set tying them together.", + "confidence": 0.96, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "Issue 45698 is the most useful representative item here because it is active, well-scoped, and centered on a reproducible from_pretrained/save_pretrained bug.", + "best_issue_reason": null, "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:43316", + "right": "issue:43716", + "accept": false, + "reason": "Both are Gemma-related API complaints, but one is a config discrepancy and the other is an image/text dtype mismatch; different code paths." + }, + { + "left": "issue:43316", + "right": "issue:44877", + "accept": false, + "reason": "Config/API mismatch in Gemma3TextConfig vs strict config loading for granite_speech are unrelated failures." + }, + { + "left": "issue:44162", + "right": "issue:44792", + "accept": false, + "reason": "ESM2 training/runtime breakage and a Janus image-generation test failure are different model pipelines and bugs." + }, + { + "left": "issue:43550", + "right": "issue:43701", + "accept": false, + "reason": "torch.compile/SDPA failure in Bamba is unrelated to resume_from_checkpoint key mismatches." + }, + { + "left": "issue:44077", + "right": "issue:44534", + "accept": false, + "reason": "Optional post_init enforcement in patchtsmixer and non-persistent buffer junk in v5 are distinct initialization issues." + }, + { + "left": "issue:43606", + "right": "issue:43701", + "accept": false, + "reason": "CPU offload device mismatch and checkpoint key mismatch are separate runtime/loading problems." + }, { "left": "issue:45803", "right": "issue:46170", "accept": false, - "reason": "Different failure modes: tensor conversion exception handling vs Windows/ROCm distributed import/type error." + "reason": "Bare except swallowing errors in FuyuBatchFeature is unrelated to Windows ROCm torch.distributed import/type issues." }, { "left": "issue:43404", "right": "issue:43441", "accept": false, - "reason": "Same model family, but one is tied-weights behavior and the other is a FlashAttention runtime failure; not the same bug." + "reason": "LM head tying bug in Mistral3 and FlashAttention failure in Ministral-3 are different model-specific regressions." }, { - "left": "issue:42617", - "right": "issue:43575", + "left": "issue:16998", + "right": "issue:22355", "accept": false, - "reason": "A script/runtime failure and a tensor-parallel OOM on a specific model are distinct issues." + "reason": "model_max_length discussion for DeBERTa-V3 is unrelated to the missing transformers.onnx module import error." }, { "left": "issue:44871", "right": "issue:45584", "accept": false, - "reason": "One is an eos_token_id config mismatch, the other is Whisper empty-transcription generation failure after align_special_tokens." + "reason": "Gemma eos_token_id config inconsistency and Whisper empty-transcription failure are unrelated generation bugs." }, { - "left": "issue:16998", - "right": "issue:22355", + "left": "issue:42617", + "right": "issue:43575", "accept": false, - "reason": "Model_max_length question and missing transformers.onnx import are unrelated." + "reason": "3d_parallel.py execution failure and Qwen tensor-parallel OOM are different symptoms with different causes." }, { "left": "issue:43673", "right": "issue:44743", "accept": false, - "reason": "Both involve cache, but one is missing GenerationMixin cache in chunked prefill and the other is recurrent state reset in modular_qwen3_5." + "reason": "Chunked-prefill cache missing and recurrent state reset in modular_qwen3_5 are different cache-related defects." }, { "left": "issue:43575", "right": "issue:43742", "accept": false, - "reason": "OOM when loading a large model with TP is unrelated to a key error loading MobileLLM-125M." + "reason": "Tensor-parallel OOM for a large Qwen model is unrelated to a MobileLLM config loading KeyError." }, { - "left": "issue:43295", - "right": "issue:44871", + "left": "issue:43388", + "right": "issue:43688", "accept": false, - "reason": "Processor/tokenizer regression and Gemma eos_token_id inconsistency affect different code paths." + "reason": "Metric gathering label truncation and auxiliary-loss normalization are separate training/eval bugs." }, { - "left": "issue:43388", - "right": "issue:43688", + "left": "issue:41950", + "right": "issue:43475", "accept": false, - "reason": "Metric label truncation and auxiliary-loss normalization are different training bugs." + "reason": "Video-classification processor lookup and SAM3 video missing attribute are different video-model issues." }, { "left": "issue:43761", "right": "issue:44479", "accept": false, - "reason": "CLIP vision hidden_states regression and Qwen VL video-input regression are separate model-path failures." + "reason": "CLIPVision hidden_states regression and Qwen video-input regression affect different architectures and code paths." }, { - "left": "issue:41950", - "right": "issue:43475", + "left": "issue:43295", + "right": "issue:44871", "accept": false, - "reason": "Pipeline processor selection and SAM3 video output attributes are unrelated." + "reason": "processor.tokenizer regression in custom code is unrelated to Gemma eos_token_id list-vs-scalar mismatch." }, { "left": "issue:43452", "right": "issue:44496", "accept": false, - "reason": "GGUF loading path breakage is not the same as an unrecognized model missing model_type." + "reason": "GGUF loading breakage and unrecognized-model/config model_type errors are both loading-related but not the same bug." }, { - "left": "issue:44871", - "right": "issue:46050", + "left": "issue:45127", + "right": "issue:45200", "accept": false, - "reason": "Config eos_token_id mismatch and quantized save/load corruption are different serialization bugs." + "reason": "LoRA merge collapse with tied embeddings and missing mm_token_type_ids default are distinct model fine-tuning problems." }, { - "left": "issue:45127", - "right": "issue:45200", + "left": "issue:44871", + "right": "issue:46050", "accept": false, - "reason": "LoRA/extended-vocab weight tying collapse and Gemma4 mm_token_type_ids defaults are unrelated." + "reason": "Gemma eos_token configuration mismatch and quantized save/load breakage are unrelated persistence issues." }, { "left": "issue:43646", "right": "issue:44913", "accept": false, - "reason": "Custom model initialization breakage in v5 is not the same as GPTNeoX rotary_pct resetting on reload." + "reason": "Custom model initialization breakage in v5 and rotary_pct reload regression are different config/init bugs." }, { "left": "issue:43572", "right": "issue:46178", "accept": false, - "reason": "Missing pad_token_idx and a backend-mapping error for tensorflow_text are unrelated." + "reason": "StableLmConfig pad_token_idx regression and tensorflow_text backend mapping error are unrelated." }, { "left": "issue:43504", "right": "issue:45698", "accept": false, - "reason": "Legacy preset-field loading and wrong custom-module loading after save_pretrained are different import/load bugs." + "reason": "Legacy field preventing Beit preset load and wrong custom module after save/load are different serialization issues." }, { - "left": "issue:32090", - "right": "issue:35141", + "left": "issue:44734", + "right": "issue:45406", "accept": false, - "reason": "Distributed broadcast of NoneType and embedding reinitialization after resizing are separate issues." + "reason": "KV-cache continuation tensor indexing crash and Gemma4Processor missing _tokenizer are separate serve bugs." }, { - "left": "issue:44734", - "right": "issue:45406", + "left": "issue:32090", + "right": "issue:35141", "accept": false, - "reason": "Serve KV-cache tensor indexing crash and Gemma4Processor missing _tokenizer are unrelated failures." + "reason": "GPU broadcast NoneType error in Trainer and output embedding reinit after resize are unrelated training/model-init issues." }, { "left": "issue:44617", "right": "issue:45584", "accept": false, - "reason": "CUDA out-of-memory in SAM3Video is not the same as Whisper empty-transcription generation failure." + "reason": "SAM3 video OOM and Whisper empty-transcription regression do not share a root cause." }, { "left": "issue:45563", "right": "issue:45584", "accept": false, - "reason": "A stale warning and an actual generation failure are different bugs." + "reason": "A stale warning in paged generate() is different from Whisper failing on empty transcription after special-token alignment." }, { "left": "issue:43761", "right": "issue:44242", "accept": false, - "reason": "Vision hidden_states regression and missing load-balancing loss are unrelated model behaviors." - }, - { - "left": "issue:43525", - "right": "issue:43531", - "accept": false, - "reason": "Missing pad_token_id and sliding_window behavior in Qwen3-MoE are distinct configuration issues." - }, - { - "left": "issue:44792", - "right": "issue:44977", - "accept": false, - "reason": "Janus image-generation test failure and Qwen3.5 flash-attention generation bug do not share the same code path." - }, - { - "left": "issue:43931", - "right": "issue:44060", - "accept": false, - "reason": "Qwen3-VL weight-shape mismatch and wrong tied-weights warning in Qwen3-Next are different model-loading problems." - }, - { - "left": "issue:44479", - "right": "issue:44734", - "accept": false, - "reason": "Video-input regression and serve KV-cache continuation indexing crash are separate defects." - }, - { - "left": "issue:44534", - "right": "issue:44913", - "accept": false, - "reason": "Non-persistent buffer junk and config reload resetting rotary_pct are not the same underlying bug." - }, - { - "left": "issue:44464", - "right": "issue:45563", - "accept": false, - "reason": "Compiled-forward output inconsistency and a stale generate warning are unrelated." - }, - { - "left": "issue:44521", - "right": "issue:45381", - "accept": false, - "reason": "All-zero assistant masks in multimodal chat templating and wrong vision_position_ids in Qwen2.5-VL video input are different multimodal bugs." + "reason": "CLIPVision forward regression and load-balancing loss omission are separate model behaviors." } ] }, @@ -9748,38 +9765,39 @@ "component_index": 1, "component_count": 17, "review_unit_index": 31, - "review_unit_count": 49, + "review_unit_count": 50, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-427", + "cluster_id": "cluster-10000-428", "nodes": [ "issue:35141", - "issue:36246", - "issue:39692", - "issue:41628", "issue:41720", "issue:42490", "issue:43065", - "issue:43295", "issue:43335", - "issue:43479", + "issue:43525", + "issue:43531", "issue:43540", "issue:43550", "issue:43644", - "issue:43650", "issue:43827", "issue:43906", + "issue:43931", "issue:43976", "issue:43994", + "issue:44060", "issue:44077", "issue:44079", "issue:44112", "issue:44190", - "issue:44220", "issue:44462", + "issue:44464", + "issue:44479", + "issue:44521", + "issue:44534", "issue:44610", "issue:44734", "issue:44737", @@ -9787,257 +9805,256 @@ "issue:44855", "issue:44913", "issue:44933", - "issue:44938", "issue:44964", + "issue:44977", "issue:45127", "issue:45245", - "issue:45276", "issue:45290", "issue:45357", "issue:45362", - "issue:45375", - "issue:45542", - "issue:45584", + "issue:45381", + "issue:45563", "issue:45588", "issue:45600", "issue:45803", "issue:45907", + "issue:45987", "issue:46082", "issue:46121" ], "soft_pairs": [ - "issue:44190|issue:44964", + "issue:43525|issue:43531", + "issue:44792|issue:44977", + "issue:43931|issue:44060", + "issue:44479|issue:44734", + "issue:44464|issue:45563", + "issue:44534|issue:44913", "issue:43906|issue:45290", + "issue:44190|issue:44964", + "issue:44521|issue:45381", "issue:43827|issue:43994", "issue:44792|issue:45600", "issue:41720|issue:45245", + "issue:45803|issue:45987", "issue:43065|issue:44610", "issue:44855|issue:45588", "issue:35141|issue:42490", "issue:44933|issue:45362", "issue:43335|issue:44855", "issue:44077|issue:44079", - "issue:43644|issue:44913", - "issue:43550|issue:44190", "issue:43976|issue:44855", + "issue:43550|issue:44190", + "issue:43644|issue:44913", "issue:43540|issue:44734", - "issue:44737|issue:45803", "issue:45127|issue:45357", - "issue:44112|issue:45600", + "issue:44737|issue:45803", "issue:44462|issue:45907", + "issue:44112|issue:45600", "issue:43335|issue:44190", - "issue:46082|issue:46121", - "issue:44938|issue:45542", - "issue:39692|issue:44190", - "issue:43479|issue:45375", - "issue:45276|issue:45584", - "issue:43295|issue:45276", - "issue:36246|issue:41628", - "issue:43650|issue:44220", - "issue:45245|issue:45584" + "issue:46082|issue:46121" ], - "prepared_review_unit_hash": "262cbfff895f833259aa7f9f4fd562fe4e6088a61ee428e749c3518e90685b94", + "prepared_review_unit_hash": "7752f49269afdf16ecf863625f2503bfa6d44d1c938c35ca213b57ef31df7eee", "component_budget": { - "node_count": 427, - "item_count": 427, - "soft_pair_count": 1576, - "serialized_chars": 816182, - "estimated_input_tokens": 204046, - "estimated_eval_tokens": 408348 + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 29, - "serialized_chars": 36577, - "estimated_input_tokens": 9145, - "estimated_eval_tokens": 18546 + "serialized_chars": 36463, + "estimated_input_tokens": 9116, + "estimated_eval_tokens": 18488 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is not a single duplicate set; the soft pairs span unrelated bugs across imports, config validation, multimodal processors, CI failures, and runtime errors. I would not merge any of the proposed edges.", - "confidence": 0.97, - "canonical_issue_reason": "No clear canonical issue: the items are heterogeneous and do not share one underlying bug or code path.", - "canonical_pr_reason": "No PRs are present in this cluster.", - "best_issue_reason": "No single issue is a good global representative because the cluster is a mix of unrelated topics rather than one duplicate family.", - "best_pr_reason": "No PRs are present, so there is no global best PR candidate.", + "summary": "The set is a grab bag of unrelated issues; none of the soft pairs look like the same underlying bug/change, so all soft edges should be rejected. There are no PRs in the cluster.", + "confidence": 0.98, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:44190", - "right": "issue:44964", + "left": "issue:43525", + "right": "issue:43531", "accept": false, - "reason": "Both are loading-related, but one is a local dataset script failure and the other is a model import/load regression; different code paths." + "reason": "Different model/config bugs: missing pad_token_id vs sliding_window handling." }, { - "left": "issue:43906", - "right": "issue:45290", + "left": "issue:44792", + "right": "issue:44977", "accept": false, - "reason": "One is an isolated reproduction issue and the other is a chat template crash with tool calls; not the same bug." + "reason": "Unrelated failures: janus test failure vs Qwen3.5 flash-attention generation." }, { - "left": "issue:43827", - "right": "issue:43994", + "left": "issue:43931", + "right": "issue:44060", "accept": false, - "reason": "Docs still referencing pipeline() and SigLIP2 incorrect outputs are distinct problems, not one underlying defect." + "reason": "Both Qwen-related, but different problems: load-time weight shape mismatch vs tied-weights warning." }, { - "left": "issue:44792", - "right": "issue:45600", + "left": "issue:44479", + "right": "issue:44734", "accept": false, - "reason": "Both touch CI, but they concern different breakages in different files and subsystems." + "reason": "Different code paths: multimodal video regression vs serve KV-cache indexing." }, { - "left": "issue:41720", - "right": "issue:45245", + "left": "issue:44464", + "right": "issue:45563", "accept": false, - "reason": "Qwen3 device-mapping CUDA assert and a category-limit runtime error are unrelated." + "reason": "Chunked generation inconsistency is not the same as a stale warning about num_return_sequences." }, { - "left": "issue:43065", - "right": "issue:44610", + "left": "issue:44534", + "right": "issue:44913", "accept": false, - "reason": "Different model families and failures: dummy Conv2d in Sam3PixelDecoder vs processor/model input-size mismatch." + "reason": "Different bugs: non-persistent buffer init vs GPTNeoX config reload behavior." }, { - "left": "issue:44855", - "right": "issue:45588", + "left": "issue:43906", + "right": "issue:45290", "accept": false, - "reason": "Python 3.13 import parsing failure is unrelated to flash_attention handling of s_aux=None." + "reason": "Insufficient overlap; one is an isolated reproduction of another issue, the other is a chat template tool-call crash." }, { - "left": "issue:35141", - "right": "issue:42490", + "left": "issue:44190", + "right": "issue:44964", "accept": false, - "reason": "Token embedding reinitialization and trailing-slash local model path warning are different issues." + "reason": "Dataset-loading example failure vs model-loading failure; different subsystems." }, { - "left": "issue:44933", - "right": "issue:45362", + "left": "issue:44521", + "right": "issue:45381", "accept": false, - "reason": "Nonexistent import from image_utils and Qwen3.5 chat crash are unrelated." + "reason": "Both multimodal, but assistant mask generation and vision_position_ids are different bugs." }, { - "left": "issue:43335", - "right": "issue:44855", + "left": "issue:43827", + "right": "issue:43994", "accept": false, - "reason": "SwitchTransformers sparse-layer config bug and DebertaV2 Python 3.13 import error do not share a code path." + "reason": "Docs referencing removed pipeline APIs is unrelated to SigLIP inference quality." }, { - "left": "issue:44077", - "right": "issue:44079", + "left": "issue:44792", + "right": "issue:45600", "accept": false, - "reason": "Optional post_init handling and ModelOutput key assignment are different framework behaviors." + "reason": "Test failure in a model path vs CI break from removed auto_mappings entries." }, { - "left": "issue:43644", - "right": "issue:44913", + "left": "issue:41720", + "right": "issue:45245", "accept": false, - "reason": "Non-persistent buffer initialization and GPTNeoX rotary_pct reload persistence are separate config/runtime issues." + "reason": "CUDA assert on Qwen3 auto device mapping is unrelated to the 2^24 category limit error." }, { - "left": "issue:43550", - "right": "issue:44190", + "left": "issue:45803", + "right": "issue:45987", "accept": false, - "reason": "torch.compile SDPA failure in Bamba and local dataset loading in a training script are unrelated." + "reason": "Different behaviors: exception handling in Fuyu batching vs cache length return type." }, { - "left": "issue:43976", - "right": "issue:44855", + "left": "issue:43065", + "right": "issue:44610", "accept": false, - "reason": "Python version support regression and a specific DebertaV2 parse failure are not the same bug." + "reason": "Dummy Conv2d in Sam3PixelDecoder is unrelated to OmDet-Turbo input size mismatch." }, { - "left": "issue:43540", - "right": "issue:44734", + "left": "issue:44855", + "right": "issue:45588", "accept": false, - "reason": "Video-input validation in Qwen3OmniMoe and serve KV-cache indexing are different code paths." + "reason": "Python 3.13 import/parse error is unrelated to flash_attention s_aux=None crash." }, { - "left": "issue:44737", - "right": "issue:45803", + "left": "issue:35141", + "right": "issue:42490", "accept": false, - "reason": "CPU device placement in XLNet and bare except in Fuyu tensor conversion are unrelated defects." + "reason": "Embedding reinit on resize is unrelated to path-trailing-slash import warning." }, { - "left": "issue:45127", - "right": "issue:45357", + "left": "issue:44933", + "right": "issue:45362", "accept": false, - "reason": "LoRA merge with extended vocab and incorrect visual encoder keys in save_pretrained are distinct bugs." + "reason": "Missing import from image_utils is unrelated to Qwen3.5 chat crash." }, { - "left": "issue:44112", - "right": "issue:45600", + "left": "issue:43335", + "right": "issue:44855", "accept": false, - "reason": "Both are CI failures, but caused by different tests and different removed config references." + "reason": "SwitchTransformers sparse-layer creation bug is unrelated to Python 3.13 indentation/parsing." }, { - "left": "issue:44462", - "right": "issue:45907", + "left": "issue:44077", + "right": "issue:44079", "accept": false, - "reason": "Tokenizer.json resolution and list_repo_files revision forwarding are unrelated repository/hub behaviors." + "reason": "Different object-model issues: optional post_init vs ModelOutput key assignment." }, { - "left": "issue:43335", - "right": "issue:44190", + "left": "issue:43976", + "right": "issue:44855", "accept": false, - "reason": "Different subsystems: SwitchTransformers config creation vs dataset loading in an example script." + "reason": "Python version incompatibility at 5.1.0 is a different issue than a decorator/comment parsing failure." }, { - "left": "issue:46082", - "right": "issue:46121", + "left": "issue:43550", + "right": "issue:44190", "accept": false, - "reason": "LlamaConfig head_dim validation and rope-parameter JSON-list handling are separate validation paths." + "reason": "torch.compile/SDPA model failure is unrelated to a local dataset loading problem." }, { - "left": "issue:44938", - "right": "issue:45542", + "left": "issue:43644", + "right": "issue:44913", "accept": false, - "reason": "Python 3.14 loading failure and missing TensorFlow backend wiring are unrelated environment issues." + "reason": "Non-persistent buffer initialization bug is unrelated to GPTNeoX config persistence." }, { - "left": "issue:39692", - "right": "issue:44190", + "left": "issue:43540", + "right": "issue:44734", "accept": false, - "reason": "SigLIP2 docs/model mismatch and local dataset loading failure do not describe the same problem." + "reason": "Video-input ValueError in Qwen3OmniMoe is unrelated to serve KV-cache tensor indexing." }, { - "left": "issue:43479", - "right": "issue:45375", + "left": "issue:45127", + "right": "issue:45357", "accept": false, - "reason": "Both are config-field issues, but they involve different models and different missing/default fields." + "reason": "LoRA merge with extended vocab is unrelated to incorrect visual encoder keys on save_pretrained." }, { - "left": "issue:45276", - "right": "issue:45584", + "left": "issue:44737", + "right": "issue:45803", "accept": false, - "reason": "Gemma4 resize_token_embeddings behavior and Whisper empty-transcription generation failure are unrelated." + "reason": "XLNet device placement bug is unrelated to broad exception swallowing in Fuyu batching." }, { - "left": "issue:43295", - "right": "issue:45276", + "left": "issue:44462", + "right": "issue:45907", "accept": false, - "reason": "Processor.tokenizer regression and Gemma4 embedding-resize behavior are different breakages." + "reason": "AutoTokenizer ignoring tokenizer.json is unrelated to list_repo_files not forwarding revision." }, { - "left": "issue:36246", - "right": "issue:41628", + "left": "issue:44112", + "right": "issue:45600", "accept": false, - "reason": "Both are import errors, but for different APIs and different modules; not the same underlying bug." + "reason": "Both are CI-related, but one is a stale GraniteSpeech test and the other is a broken auto_mappings reference." }, { - "left": "issue:43650", - "right": "issue:44220", + "left": "issue:43335", + "right": "issue:44190", "accept": false, - "reason": "Placeholder issue title and fbank feature extraction error are unrelated." + "reason": "SwitchTransformers config bug is unrelated to dataset loading in the training script." }, { - "left": "issue:45245", - "right": "issue:45584", + "left": "issue:46082", + "right": "issue:46121", "accept": false, - "reason": "Category cardinality runtime error and Whisper empty transcription are different failures." + "reason": "Llama head_dim validation and rope-param JSON list handling are different configuration validation bugs." } ] }, @@ -10053,317 +10070,310 @@ "component_index": 1, "component_count": 17, "review_unit_index": 32, - "review_unit_count": 49, + "review_unit_count": 50, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-427", + "cluster_id": "cluster-10000-428", "nodes": [ "issue:36246", "issue:38175", "issue:38617", + "issue:39692", + "issue:41628", "issue:41950", "issue:42175", "issue:42490", "issue:42907", "issue:43065", "issue:43295", - "issue:43299", - "issue:43335", - "issue:43381", "issue:43404", "issue:43441", + "issue:43479", "issue:43526", "issue:43531", "issue:43643", - "issue:43673", - "issue:43824", + "issue:43650", "issue:43854", "issue:43901", "issue:43976", "issue:44016", "issue:44062", - "issue:44075", "issue:44112", "issue:44186", + "issue:44190", "issue:44220", "issue:44246", "issue:44279", - "issue:44336", - "issue:44368", "issue:44496", "issue:44561", "issue:44683", "issue:44734", - "issue:44933", + "issue:44938", "issue:44964", "issue:44987", "issue:45003", "issue:45042", "issue:45216", + "issue:45245", + "issue:45276", "issue:45335", + "issue:45375", "issue:45446", - "issue:45593", + "issue:45542", + "issue:45584", "issue:45600", - "issue:45632", - "issue:45759" + "issue:45632" ], "soft_pairs": [ - "issue:44186|issue:44734", + "issue:44938|issue:45542", + "issue:39692|issue:44190", + "issue:43479|issue:45375", + "issue:45276|issue:45584", + "issue:43295|issue:45276", + "issue:36246|issue:41628", + "issue:43650|issue:44220", + "issue:45245|issue:45584", "issue:41950|issue:44016", - "issue:42175|issue:44279", + "issue:44186|issue:44734", "issue:36246|issue:38617", - "issue:44246|issue:45003", + "issue:42175|issue:44279", "issue:44683|issue:45446", + "issue:44246|issue:45003", "issue:43404|issue:45216", - "issue:38175|issue:43065", "issue:44016|issue:44220", + "issue:38175|issue:43065", "issue:43643|issue:44496", "issue:44734|issue:45632", - "issue:44964|issue:45600", "issue:43854|issue:44964", - "issue:44112|issue:44964", + "issue:44964|issue:45600", "issue:44561|issue:45632", + "issue:44112|issue:44964", "issue:43976|issue:44987", - "issue:43531|issue:45600", "issue:42907|issue:43441", "issue:42490|issue:45632", + "issue:43531|issue:45600", "issue:43295|issue:45335", "issue:43526|issue:43901", "issue:44496|issue:45042", - "issue:43673|issue:44336", - "issue:44016|issue:44062", - "issue:43335|issue:45593", - "issue:43824|issue:44933", - "issue:44075|issue:44368", - "issue:43299|issue:45216", - "issue:43381|issue:45593", - "issue:43976|issue:44279", - "issue:43901|issue:44496", - "issue:44496|issue:45759" + "issue:44016|issue:44062" ], - "prepared_review_unit_hash": "2630200bf7f6b1dce437710986990c46d463b56ce1011a551776e67fb7424a1c", + "prepared_review_unit_hash": "5ae5dc67ee6c80ff190e2d8f064a55eb7be17438d1b67f1902c63307126a99da", "component_budget": { - "node_count": 427, - "item_count": 427, - "soft_pair_count": 1576, - "serialized_chars": 816182, - "estimated_input_tokens": 204046, - "estimated_eval_tokens": 408348 + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 32, - "serialized_chars": 37322, - "estimated_input_tokens": 9331, - "estimated_eval_tokens": 18918 + "soft_pair_count": 31, + "serialized_chars": 36923, + "estimated_input_tokens": 9231, + "estimated_eval_tokens": 18718 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is heterogeneous: it mixes unrelated import/export regressions, model-loading bugs, pipeline issues, docs/tests, and performance reports. None of the soft pairs look like the same underlying bug or change, so all should be rejected as duplicates.", - "confidence": 0.76, - "canonical_issue_reason": "No single clear canonical issue exists because the cluster spans unrelated subsystems; if one must be chosen, issue 43824 is the closest representative as a concrete import regression with inbound references.", + "summary": "This cluster is heterogeneous: most items are unrelated model-loading/import/config regressions across different models and subsystems. I did not find any soft pair that looks like the same concrete bug/change, so all soft edges are rejected.", + "confidence": 0.94, + "canonical_issue_reason": "Issue 44561 is the broadest, most cross-cutting regression-style report in the set: a v5 compatibility break impacting trust_remote_code models. It is a better anchor than the many model-specific import/load/config complaints here.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 43824 is the strongest representative among the set because it is a concrete, user-facing import failure and has inbound references, but it still does not explain the rest of the cluster.", + "best_issue_reason": "Issue 44561 is the best representative of the cluster\u2019s main theme (Transformers v5 regressions affecting remote-code/model-loading paths) and has the clearest general duplicate-target potential.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:44186", - "right": "issue:44734", + "left": "issue:44938", + "right": "issue:45542", "accept": false, - "reason": "Different bugs: tokenizer NER/padding crash vs serve-side KV cache continuation indexing." + "reason": "Different failure domains: Python 3.14 import/load issue vs missing TensorFlow backend when only tensorboard is installed." }, { - "left": "issue:41950", - "right": "issue:44016", + "left": "issue:39692", + "right": "issue:44190", "accept": false, - "reason": "Different areas: video-classification pipeline image-processor lookup vs a notebook syntax error." + "reason": "Different problems: SigLIP2 docs/example errors vs local dataset loading in a training script." }, { - "left": "issue:42175", - "right": "issue:44279", + "left": "issue:43479", + "right": "issue:45375", "accept": false, - "reason": "One is missing TensorFlow in a torch-only install, the other is a generic dependency problem." + "reason": "Both are config-related, but they affect different models and different fields; not the same underlying bug." }, { - "left": "issue:36246", - "right": "issue:38617", + "left": "issue:45276", + "right": "issue:45584", "accept": false, - "reason": "Both are import errors, but for unrelated symbols and code paths." + "reason": "Unrelated code paths: token embedding resize behavior in Gemma4 vs Whisper generation after special-token alignment." }, { - "left": "issue:44246", - "right": "issue:45003", + "left": "issue:43295", + "right": "issue:45276", "accept": false, - "reason": "Import slowness from package initialization is not the same as unsafe sys.modules access in modeling_utils." + "reason": "Different regressions: processor/tokenizer API behavior vs embedding resize propagation." }, { - "left": "issue:44683", - "right": "issue:45446", + "left": "issue:36246", + "right": "issue:41628", "accept": false, - "reason": "Both mention flex_attention, but they concern different version checks and failure modes." + "reason": "Both are import errors, but for different symbols and different modules; not the same bug." }, { - "left": "issue:43404", - "right": "issue:45216", + "left": "issue:43650", + "right": "issue:44220", "accept": false, - "reason": "Different model-specific regressions: untied lm_head vs incorrect save_pretrained checkpoint state." + "reason": "One is a placeholder/data request issue; the other is a specific audio feature extraction bug." }, { - "left": "issue:38175", - "right": "issue:43065", + "left": "issue:45245", + "right": "issue:45584", "accept": false, - "reason": "Unrelated model behavior: zero probabilities in SigLIP2 vs a dummy Conv2d in Sam3PixelDecoder." + "reason": "Category-count runtime error is unrelated to Whisper empty-transcription generation failure." }, { - "left": "issue:44016", - "right": "issue:44220", + "left": "issue:41950", + "right": "issue:44016", "accept": false, - "reason": "Notebook syntax error and audio feature extraction failure are unrelated." + "reason": "Video-classification processor lookup bug vs notebook syntax error; no shared code-path." }, { - "left": "issue:43643", - "right": "issue:44496", + "left": "issue:44186", + "right": "issue:44734", "accept": false, - "reason": "Different config-loading bugs: missing fields under trust_remote_code vs unrecognized model_type." + "reason": "Tokenizer/padding crash and server KV-cache indexing crash are different components and failures." }, { - "left": "issue:44734", - "right": "issue:45632", + "left": "issue:36246", + "right": "issue:38617", "accept": false, - "reason": "Serve KV-cache indexing and trust_remote_code cache path collision are unrelated." + "reason": "Both are import-related but for different missing names and different areas of the package." }, { - "left": "issue:44964", - "right": "issue:45600", + "left": "issue:42175", + "right": "issue:44279", "accept": false, - "reason": "Both affect loading, but one is a model load failure and the other is stale auto-mapping references in CI." + "reason": "Packaging/backends dependency complaint vs a generic dependency issue; too vague and not the same concrete problem." }, { - "left": "issue:43854", - "right": "issue:44964", + "left": "issue:44683", + "right": "issue:45446", "accept": false, - "reason": "Different model-specific load failures for different architectures and causes." + "reason": "Both touch flex_attention, but one is a compiled-kernel failure on newer torch and the other is a version-check/import guard bug." }, { - "left": "issue:44112", - "right": "issue:44964", + "left": "issue:44246", + "right": "issue:45003", "accept": false, - "reason": "CI test failure in GraniteSpeech is not the same as a Phi-4 multimodal loading regression." + "reason": "Import performance regression and unsafe sys.modules access are not the same concrete bug." }, { - "left": "issue:44561", - "right": "issue:45632", + "left": "issue:43404", + "right": "issue:45216", "accept": false, - "reason": "Removal of a helper breaking trust_remote_code models is unrelated to cache-path collisions." + "reason": "Different models and failures: untied lm_head weights vs incorrect save_pretrained checkpoint serialization." }, { - "left": "issue:43976", - "right": "issue:44987", + "left": "issue:44016", + "right": "issue:44220", "accept": false, - "reason": "Python-version compatibility and a specific model-loading failure are different issues." + "reason": "Notebook syntax error is unrelated to audio feature extraction code." }, { - "left": "issue:43531", - "right": "issue:45600", + "left": "issue:38175", + "right": "issue:43065", "accept": false, - "reason": "Qwen3-MoE sliding_window behavior is unrelated to removed Sam3LiteText config references." + "reason": "Different model families and different issues: SigLIP2 zero probabilities vs SAM3 decoder dummy layer." }, { - "left": "issue:42907", - "right": "issue:43441", + "left": "issue:43643", + "right": "issue:44496", "accept": false, - "reason": "Different model families and failure modes: save/dequantize vs FlashAttention loading." + "reason": "Both involve config/model loading, but one is missing fields with trust_remote_code and the other is an unrecognized model_type error; distinct root causes." }, { - "left": "issue:42490", + "left": "issue:44734", "right": "issue:45632", "accept": false, - "reason": "Trailing-slash local-path warning and trust_remote_code cache collisions are distinct path-handling bugs." + "reason": "Server tensor indexing crash vs local trust_remote_code cache-path collision are different failure modes." }, { - "left": "issue:43295", - "right": "issue:45335", + "left": "issue:43854", + "right": "issue:44964", "accept": false, - "reason": "Processor/tokenizer regression is unrelated to resize_token_embeddings not reaching decoder.embed_tokens." + "reason": "Different models and symptoms: GLM unit-test loading issue vs Phi-4 multimodal loading issue." }, { - "left": "issue:43526", - "right": "issue:43901", + "left": "issue:44964", + "right": "issue:45600", "accept": false, - "reason": "An image-processor label-reduction bug is unrelated to stale documentation wording." + "reason": "Model loading failure for Phi-4 multimodal is unrelated to stale auto_mappings references and CI breakage." }, { - "left": "issue:44496", - "right": "issue:45042", + "left": "issue:44561", + "right": "issue:45632", "accept": false, - "reason": "Unrecognized model config and unnecessary torchvision dependency in PIL processors are different bugs." + "reason": "Both mention trust_remote_code, but one is a v5 API removal regression and the other is a cache-path collision; not the same bug." }, { - "left": "issue:43673", - "right": "issue:44336", + "left": "issue:44112", + "right": "issue:44964", "accept": false, - "reason": "Generation cache regression and ANSI escaping in loading_report are unrelated." + "reason": "CI test flake in GraniteSpeech is unrelated to Phi-4 multimodal model loading." }, { - "left": "issue:44016", - "right": "issue:44062", + "left": "issue:43976", + "right": "issue:44987", "accept": false, - "reason": "Notebook syntax error is unrelated to AddedToken keyword argument handling." + "reason": "General Python-version compatibility complaint vs a specific model-loading failure; not the same underlying issue." }, { - "left": "issue:43335", - "right": "issue:45593", + "left": "issue:42907", + "right": "issue:43441", "accept": false, - "reason": "SwitchTransformers sparse-layer creation and D-FINE auxiliary-loss behavior are different model bugs." + "reason": "Different Ministral issues: saving dequantized models vs FlashAttention failure." }, { - "left": "issue:43824", - "right": "issue:44933", + "left": "issue:42490", + "right": "issue:45632", "accept": false, - "reason": "Both involve imports, but they are missing different symbols from different modules." + "reason": "Both are local-path related, but one is a warning on trailing slashes and the other is a cache collision for same leaf names." }, { - "left": "issue:44075", - "right": "issue:44368", + "left": "issue:43531", + "right": "issue:45600", "accept": false, - "reason": "Optimizer argument usage and a tie_word_embeddings warning are unrelated." + "reason": "Qwen3-MoE sliding-window behavior vs Sam3LiteText auto-mapping removal are unrelated." }, { - "left": "issue:43299", - "right": "issue:45216", + "left": "issue:43295", + "right": "issue:45335", "accept": false, - "reason": "Different Qwen regressions: MoE model loading vs incorrect checkpoint saving." + "reason": "Both involve resize_token_embeddings, but on different architectures and different propagation bugs." }, { - "left": "issue:43381", - "right": "issue:45593", + "left": "issue:43526", + "right": "issue:43901", "accept": false, - "reason": "Gradient checkpointing in eval mode is unrelated to D-FINE auxiliary-loss suppression." + "reason": "Image processor label reduction bug vs text-classification docs/behavior mismatch are unrelated." }, { - "left": "issue:43976", - "right": "issue:44279", + "left": "issue:44496", + "right": "issue:45042", "accept": false, - "reason": "Python support mismatch and a generic dependency issue are not the same problem." + "reason": "Unrecognized model config loading vs PIL image processors requiring torchvision are different code paths." }, { - "left": "issue:43901", - "right": "issue:44496", + "left": "issue:44016", + "right": "issue:44062", "accept": false, - "reason": "Documentation mismatch and unrecognized model config are unrelated." - }, - { - "left": "issue:44496", - "right": "issue:45759", - "accept": false, - "reason": "Unrecognized model config is unrelated to missing text_config unwrapping in composite multimodal configs." + "reason": "Notebook syntax error and AddedToken keyword collision are unrelated." } ] }, @@ -10379,47 +10389,49 @@ "component_index": 1, "component_count": 17, "review_unit_index": 33, - "review_unit_count": 49, + "review_unit_count": 50, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-427", + "cluster_id": "cluster-10000-428", "nodes": [ "issue:38175", "issue:39692", "issue:41950", - "issue:42994", "issue:43295", "issue:43299", - "issue:43317", - "issue:43441", + "issue:43335", + "issue:43381", "issue:43450", "issue:43479", "issue:43525", "issue:43577", "issue:43650", + "issue:43673", + "issue:43824", "issue:43854", "issue:43901", "issue:43976", - "issue:44016", + "issue:44075", "issue:44242", "issue:44279", - "issue:44291", "issue:44297", + "issue:44336", "issue:44351", + "issue:44368", "issue:44462", - "issue:44466", + "issue:44496", "issue:44521", "issue:44617", "issue:44683", "issue:44779", + "issue:44933", "issue:45020", "issue:45081", "issue:45200", "issue:45216", - "issue:45276", "issue:45278", "issue:45325", "issue:45335", @@ -10427,15 +10439,22 @@ "issue:45405", "issue:45542", "issue:45584", - "issue:45600", - "issue:45712", + "issue:45593", + "issue:45759", "issue:45812", "issue:45823", - "issue:45834", - "issue:45907", "issue:46132" ], "soft_pairs": [ + "issue:43673|issue:44336", + "issue:43335|issue:45593", + "issue:43824|issue:44933", + "issue:44075|issue:44368", + "issue:43299|issue:45216", + "issue:43976|issue:44279", + "issue:43381|issue:45593", + "issue:43901|issue:44496", + "issue:44496|issue:45759", "issue:45335|issue:45584", "issue:43479|issue:43525", "issue:44521|issue:45325", @@ -10443,12 +10462,12 @@ "issue:43299|issue:45357", "issue:44351|issue:45278", "issue:43577|issue:45216", - "issue:38175|issue:44617", "issue:41950|issue:45020", "issue:43976|issue:45405", + "issue:38175|issue:44617", + "issue:41950|issue:45823", "issue:44779|issue:45812", "issue:39692|issue:43901", - "issue:41950|issue:45823", "issue:44279|issue:45405", "issue:44683|issue:45542", "issue:43295|issue:44779", @@ -10456,211 +10475,216 @@ "issue:43650|issue:46132", "issue:44462|issue:45081", "issue:43450|issue:44279", - "issue:44297|issue:44462", - "issue:44016|issue:44291", - "issue:45081|issue:45907", - "issue:41950|issue:43650", - "issue:43441|issue:45834", - "issue:44466|issue:45276", - "issue:45600|issue:45712", - "issue:42994|issue:43317" + "issue:44297|issue:44462" ], - "prepared_review_unit_hash": "3e90676754a866ff8210bada21acb3bf3c3a0a1f836430e704481ee20decd389", + "prepared_review_unit_hash": "2b57409df080534f53c133ed5ca0b285ae4fdfa64e2a3c591fd3072a1d1b6e6a", "component_budget": { - "node_count": 427, - "item_count": 427, - "soft_pair_count": 1576, - "serialized_chars": 816182, - "estimated_input_tokens": 204046, - "estimated_eval_tokens": 408348 + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 }, "budget": { "node_count": 47, "item_count": 47, - "soft_pair_count": 28, - "serialized_chars": 35526, - "estimated_input_tokens": 8882, - "estimated_eval_tokens": 18020 + "soft_pair_count": 30, + "serialized_chars": 36129, + "estimated_input_tokens": 9033, + "estimated_eval_tokens": 18322 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The cluster is highly heterogeneous: most pairs share only broad subsystems (tokenizers, multimodal, quantization, docs) but not the same concrete bug. I rejected all soft-edge candidates; none look like safe duplicate merges.", - "confidence": 0.92, + "summary": "This cluster is highly heterogeneous: it mixes unrelated bugs in tokenizer handling, multimodal configs, model loading/saving, pipelines, docs, and dependency/version compatibility. I found no true duplicate pairs; all soft edges are false positives.", + "confidence": 0.93, "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "issue:44521 is the strongest standalone issue in the set by activity and references, but the overall collection is too mixed to treat it as a true canonical for the cluster.", + "best_issue_reason": "issue:43299 is the closest representative because it is a concrete, high-signal regression in a major model-loading path, but it still does not serve as a true duplicate hub for this mixed cluster.", "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:43673", + "right": "issue:44336", + "accept": false, + "reason": "Different subsystems and symptoms: chunked_prefill cache handling vs ANSI formatting in loading_report." + }, + { + "left": "issue:43335", + "right": "issue:45593", + "accept": false, + "reason": "Unrelated model logic: SwitchTransformers sparse-layer initialization vs D-FINE auxiliary-loss behavior." + }, + { + "left": "issue:43824", + "right": "issue:44933", + "accept": false, + "reason": "Both are import problems, but for different missing symbols in different modules and with different causes." + }, + { + "left": "issue:44075", + "right": "issue:44368", + "accept": false, + "reason": "Optimizer argument handling is unrelated to the tie_word_embeddings warning in a Qwen LoRA setup." + }, + { + "left": "issue:43299", + "right": "issue:45216", + "accept": false, + "reason": "One is a model-loading failure; the other is a save_pretrained checkpoint serialization regression." + }, + { + "left": "issue:43976", + "right": "issue:44279", + "accept": false, + "reason": "Python-version compatibility and a generic dependency issue are too broad to be the same bug." + }, + { + "left": "issue:43381", + "right": "issue:45593", + "accept": false, + "reason": "Gradient-checkpointing eval-mode behavior is unrelated to D-FINE denoising/auxiliary losses." + }, + { + "left": "issue:43901", + "right": "issue:44496", + "accept": false, + "reason": "Docs mismatch vs a model/config loading error; different failure modes and code paths." + }, + { + "left": "issue:44496", + "right": "issue:45759", + "accept": false, + "reason": "Both involve model configuration, but one is unrecognized-model config metadata and the other is from_config unwrapping for composite Qwen configs." + }, { "left": "issue:45335", "right": "issue:45584", "accept": false, - "reason": "Different bugs: one is about resize_token_embeddings not updating Gemma/t5Gemma embeddings, the other is Whisper generation failing after align_special_tokens." + "reason": "T5Gemma embedding resizing and Whisper empty-transcription generation are unrelated bugs." }, { "left": "issue:43479", "right": "issue:43525", "accept": false, - "reason": "Different models and failure modes: Phi4MultimodalConfig default init vs Llama4Config missing pad_token_id." + "reason": "Different config fields and failures: missing default vision/audio configs vs missing pad_token_id." }, { "left": "issue:44521", "right": "issue:45325", "accept": false, - "reason": "Both are multimodal-related, but one is assistant mask generation and the other is rope/position id scaling; not the same bug." + "reason": "Both are multimodal, but assistant mask generation and rope-index scaling are separate code paths." }, { "left": "issue:44242", "right": "issue:45200", "accept": false, - "reason": "Unrelated concrete issues: MoE load balancing loss vs Gemma 4 text-only mm_token_type_ids defaults." + "reason": "MoE load-balancing loss logic is unrelated to Gemma4 mm_token_type_ids defaults." }, { "left": "issue:43299", "right": "issue:45357", "accept": false, - "reason": "Different Qwen variants and different code paths: model loading vs save_pretrained serialization keys." + "reason": "Loading regression vs save_pretrained writing incorrect visual-encoder keys; not the same concrete bug." }, { "left": "issue:44351", "right": "issue:45278", "accept": false, - "reason": "One is a specific HybridCache import failure; the other is a broad post-upgrade import breakage. Too generic to merge." + "reason": "A specific HybridCache import error is not the same as broad import errors after version upgrades." }, { "left": "issue:43577", "right": "issue:45216", "accept": false, - "reason": "Different regressions: BLIP2 dtype propagation vs Qwen3.5 save_pretrained checkpoint correctness." - }, - { - "left": "issue:38175", - "right": "issue:44617", - "accept": false, - "reason": "Completely different problems: SigLIP2 zero probabilities vs Sam3Video CUDA OOM." + "reason": "Blip2 dtype propagation and Qwen3.5 checkpoint saving are different subsystems and symptoms." }, { "left": "issue:41950", "right": "issue:45020", "accept": false, - "reason": "Different failures: video-classification processor lookup vs remote_code model breakage." + "reason": "Video-classification processor selection and remote_code model breakage are unrelated." }, { "left": "issue:43976", "right": "issue:45405", "accept": false, - "reason": "Python compatibility issue vs a dependency version bump; not the same underlying bug." + "reason": "Python support/versioning and a PEFT minimum-version bump are not the same bug." + }, + { + "left": "issue:38175", + "right": "issue:44617", + "accept": false, + "reason": "SigLIP2 zero probabilities and Sam3Video OOM are completely different problems." + }, + { + "left": "issue:41950", + "right": "issue:45823", + "accept": false, + "reason": "Pipeline processor lookup vs Gemma4 device placement mismatch are unrelated failures." }, { "left": "issue:44779", "right": "issue:45812", "accept": false, - "reason": "Both are tokenizer regressions, but they affect different model families and likely different causes." + "reason": "Both are tokenizer regressions, but they affect different model families and likely different tokenizer code paths." }, { "left": "issue:39692", "right": "issue:43901", "accept": false, - "reason": "Both mention docs, but one is a SigLIP2 example error and the other is a TextClassificationPipeline parameter mismatch." - }, - { - "left": "issue:41950", - "right": "issue:45823", - "accept": false, - "reason": "No shared concrete code path: video-classification pipeline processor bug vs Gemma4 PLE device mismatch." + "reason": "SigLIP2 docs example errors and a TextClassificationPipeline docs note are unrelated documentation issues." }, { "left": "issue:44279", "right": "issue:45405", "accept": false, - "reason": "Dependency issue and MIN_PEFT_VERSION bump are unrelated." + "reason": "Generic dependency trouble and a specific unreleased PEFT version bump are different issues." }, { "left": "issue:44683", "right": "issue:45542", "accept": false, - "reason": "Compiled flex_attention torch-compatibility bug vs TensorBoard/TensorFlow backend import issue." + "reason": "torch flex_attention compilation failure is unrelated to the TensorFlow/tensorboard backend error." }, { "left": "issue:43295", "right": "issue:44779", "accept": false, - "reason": "Both affect tokenizer-related behavior, but one is processor/tokenizer API regression and the other is Deepseek tokenization correctness." + "reason": "Custom model processor/tokenizer regression is different from Deepseek tokenizer output corruption." }, { "left": "issue:43299", "right": "issue:43854", "accept": false, - "reason": "Different problems: Qwen3VL MoE model loading vs GLM-4.7-Flash unit test loading." + "reason": "They are both model-loading reports, but for different models and different failure points." }, { "left": "issue:43650", "right": "issue:46132", "accept": false, - "reason": "The first is an invalid issue title; the second is an AttentionInterface behavior change. No duplicate signal." + "reason": "Noise/placeholder issue title vs an AttentionInterface.register behavior change; no shared bug." }, { "left": "issue:44462", "right": "issue:45081", "accept": false, - "reason": "Both involve tokenizer code, but one is AutoTokenizer ignoring tokenizer.json and the other is a Mistral regex patch crash." + "reason": "AutoTokenizer ignoring tokenizer.json and a Mistral regex patch crash are separate tokenizer-loading bugs." }, { "left": "issue:43450", "right": "issue:44279", "accept": false, - "reason": "Video processor batching shape bug vs generic dependency issue." + "reason": "Batched video processor output shape and a dependency issue have no concrete overlap." }, { "left": "issue:44297", "right": "issue:44462", "accept": false, - "reason": "Related area, but one is tokenizer_class serialization mismatch and the other is AutoTokenizer loading behavior; not the same bug." - }, - { - "left": "issue:44016", - "right": "issue:44291", - "accept": false, - "reason": "Notebook syntax error is unrelated to init_empty_weights argument handling." - }, - { - "left": "issue:45081", - "right": "issue:45907", - "accept": false, - "reason": "Tokenizer regex crash vs repo file listing revision forwarding; no shared bug." - }, - { - "left": "issue:41950", - "right": "issue:43650", - "accept": false, - "reason": "The items are unrelated; one is a real pipeline bug, the other is just a malformed issue title." - }, - { - "left": "issue:43441", - "right": "issue:45834", - "accept": false, - "reason": "Ministral FlashAttention bug and Kosmos2.5 OCR index error are different model/path failures." - }, - { - "left": "issue:44466", - "right": "issue:45276", - "accept": false, - "reason": "Both concern weights/embeddings, but one is serialization inconsistency and the other is resize_token_embeddings not updating Gemma4 tensors." - }, - { - "left": "issue:45600", - "right": "issue:45712", - "accept": false, - "reason": "Different CI hygiene issues: stale auto_mappings entries vs leftover dummy classes." - }, - { - "left": "issue:42994", - "right": "issue:43317", - "accept": false, - "reason": "Different quantization scenarios: saving a quantized model vs loading a dequantized model with device_map=auto offload." + "reason": "Saving the wrong tokenizer_class and ignoring tokenizer.json are different serialization/loading defects." } ] }, @@ -10676,23 +10700,23 @@ "component_index": 1, "component_count": 17, "review_unit_index": 34, - "review_unit_count": 49, + "review_unit_count": 50, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-427", + "cluster_id": "cluster-10000-428", "nodes": [ "issue:33290", - "issue:41628", + "issue:41950", "issue:42175", + "issue:42994", "issue:43065", "issue:43262", + "issue:43317", "issue:43441", - "issue:43450", "issue:43525", - "issue:43643", "issue:43650", "issue:43653", "issue:43673", @@ -10706,251 +10730,251 @@ "issue:43994", "issue:44016", "issue:44162", - "issue:44188", - "issue:44246", "issue:44261", "issue:44263", + "issue:44291", "issue:44297", - "issue:44336", "issue:44361", + "issue:44466", "issue:44521", - "issue:44589", - "issue:44617", "issue:44623", "issue:44671", "issue:44683", "issue:44749", "issue:44779", "issue:44898", + "issue:45081", + "issue:45276", "issue:45290", - "issue:45362", "issue:45397", "issue:45542", - "issue:45561", - "issue:45563", "issue:45593", + "issue:45600", "issue:45656", - "issue:45823" + "issue:45712", + "issue:45823", + "issue:45834", + "issue:45907" ], "soft_pairs": [ + "issue:44016|issue:44291", + "issue:45081|issue:45907", + "issue:41950|issue:43650", + "issue:43441|issue:45834", + "issue:44466|issue:45276", + "issue:45600|issue:45712", + "issue:42994|issue:43317", "issue:33290|issue:45656", - "issue:43653|issue:45593", "issue:43650|issue:45290", - "issue:44623|issue:45397", "issue:44623|issue:44749", + "issue:44623|issue:45397", + "issue:43653|issue:45593", "issue:43065|issue:44898", "issue:44261|issue:44263", - "issue:43262|issue:44521", "issue:43441|issue:44683", - "issue:44361|issue:45593", "issue:44016|issue:45823", + "issue:43262|issue:44521", + "issue:44361|issue:45593", + "issue:42175|issue:43976", "issue:43723|issue:44162", "issue:44671|issue:44779", - "issue:42175|issue:43976", - "issue:43673|issue:43761", "issue:43901|issue:43994", + "issue:43673|issue:43761", "issue:43784|issue:45542", "issue:43824|issue:43976", - "issue:43450|issue:44617", "issue:44297|issue:45397", - "issue:43525|issue:43881", - "issue:43901|issue:44246", - "issue:43525|issue:43643", - "issue:44188|issue:44336", - "issue:44589|issue:45362", - "issue:41628|issue:43976", - "issue:45561|issue:45563" + "issue:43525|issue:43881" ], - "prepared_review_unit_hash": "e6bf5fa6d8e98c09f7d9e42388d2ca27e0b7b7abccf8c6171d67fc8bec35134d", + "prepared_review_unit_hash": "fe8d52f42a9911139d324c5f7fcde60063736d008beb26e3fa11ae2d96915bfc", "component_budget": { - "node_count": 427, - "item_count": 427, - "soft_pair_count": 1576, - "serialized_chars": 816182, - "estimated_input_tokens": 204046, - "estimated_eval_tokens": 408348 + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 }, "budget": { "node_count": 47, "item_count": 47, "soft_pair_count": 27, - "serialized_chars": 34890, - "estimated_input_tokens": 8723, - "estimated_eval_tokens": 17702 + "serialized_chars": 35215, + "estimated_input_tokens": 8804, + "estimated_eval_tokens": 17864 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "These issues are mostly unrelated one-off bugs or regressions; none of the soft pairs look like the same underlying defect closely enough to merge. The closest themes are chat-template/processor bugs and a few model-loading regressions, but each pair still targets a different code path or symptom.", - "confidence": 0.93, - "canonical_issue_reason": "issue:44521 is the best cluster anchor: it is still open, has substantial discussion/inbound references, and is a concrete, reproducible regression with broad processor/chat-template impact.", + "summary": "This cluster is highly heterogeneous: the soft-linked issues cover unrelated tokenizer, model, pipeline, packaging, CI, and notebook problems. None of the candidate soft edges look like true duplicates or a single shared fix.", + "confidence": 0.98, + "canonical_issue_reason": "No clear canonical issue: the cluster mixes unrelated bugs across many subsystems, so there is no single issue that represents the underlying problem.", "canonical_pr_reason": null, - "best_issue_reason": "issue:44521 is the strongest standalone triage candidate because it has a clear repro, active attention, and a user-visible failure affecting multimodal chat-template handling.", + "best_issue_reason": "No strong best issue candidate for deduplication purposes; the issues are too diverse to safely treat one as the representative fix target.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:33290", - "right": "issue:45656", + "left": "issue:44016", + "right": "issue:44291", "accept": false, - "reason": "Both mention deepspeed/optimizer behavior, but one is OOM in Adafactor while the other is double optimizer stepping; different failure modes and fixes." + "reason": "Notebook syntax error vs. init_empty_weights argument mismatch; different failure modes and code paths." }, { - "left": "issue:43653", - "right": "issue:45593", + "left": "issue:45081", + "right": "issue:45907", "accept": false, - "reason": "BigBirdTokenizer special-token registration and D-FINE auxiliary-loss behavior are unrelated bugs in different subsystems." + "reason": "Mistral tokenizer regex crash vs. list_repo_files revision forwarding; unrelated areas." }, { - "left": "issue:43650", - "right": "issue:45290", + "left": "issue:41950", + "right": "issue:43650", "accept": false, - "reason": "Both touch chat-template paths, but one is a vague data request and the other is a specific tool-call/content crash; not the same defect." + "reason": "Video-classification processor bug vs. an unrelated issue titled \"ADD THE DATA\"; no shared bug." }, { - "left": "issue:44623", - "right": "issue:45397", + "left": "issue:43441", + "right": "issue:45834", "accept": false, - "reason": "Processor save_pretrained missing files and gemma-4 zero3 from_pretrained are different loading/saving problems with different code paths." + "reason": "Ministral FlashAttention regression vs. Kosmos2.5 OCR index error; different models and symptoms." }, { - "left": "issue:44623", - "right": "issue:44749", + "left": "issue:44466", + "right": "issue:45276", "accept": false, - "reason": "Missing processor files and slow Chinese data filtering are entirely unrelated." + "reason": "lm_head.weight serialization inconsistency vs. Gemma4 embedding resize behavior; not the same bug." }, { - "left": "issue:43065", - "right": "issue:44898", + "left": "issue:45600", + "right": "issue:45712", "accept": false, - "reason": "Both are vision-model issues, but one is a dummy Conv2d in Sam3PixelDecoder and the other is a Perceiver resolution/interpolation failure." + "reason": "Sam3LiteText auto-mapping CI break vs. leftover dummy classes in dummy_pt_objects; separate cleanup issues." }, { - "left": "issue:44261", - "right": "issue:44263", + "left": "issue:42994", + "right": "issue:43317", "accept": false, - "reason": "Different model internals and different symptoms: RMS norm epsilon precision mismatch vs torch.split return values." + "reason": "Quantized model saving failure vs. dequantized model loading/offload failure; opposite workflows, different bugs." }, { - "left": "issue:43262", - "right": "issue:44521", + "left": "issue:33290", + "right": "issue:45656", "accept": false, - "reason": "Both involve apply_chat_template, but one is an audio sampling-rate default bug while the other is multimodal assistant_mask generation; not the same issue." + "reason": "Adafactor OOM under deepspeed vs. optimizer.step called twice under deepspeed; related subsystem but different concrete defects." }, { - "left": "issue:43441", - "right": "issue:44683", + "left": "issue:43650", + "right": "issue:45290", "accept": false, - "reason": "Different attention-kernel failures: FlashAttention breakage for Ministral-3 vs compiled flex_attention failing on newer torch." + "reason": "Unrelated issue text vs. apply_chat_template crash on tool-call assistant messages." }, { - "left": "issue:44361", - "right": "issue:45593", + "left": "issue:44623", + "right": "issue:44749", "accept": false, - "reason": "MLukeTokenizer task AttributeError and D-FINE auxiliary-loss behavior are unrelated." + "reason": "processor.save_pretrained missing files vs. Chinese-language performance regression report; unrelated." }, { - "left": "issue:44016", - "right": "issue:45823", + "left": "issue:44623", + "right": "issue:45397", "accept": false, - "reason": "Notebook syntax error and Gemma4 device-map mismatch are unrelated." + "reason": "Missing processor files on save vs. gemma-4 zero3 from_pretrained failure; different code paths." }, { - "left": "issue:43723", - "right": "issue:44162", + "left": "issue:43653", + "right": "issue:45593", "accept": false, - "reason": "Tokenizer loading regression and ESM2 model breakage are different problems." + "reason": "BigBirdTokenizer special-token registration bug vs. D-FINE auxiliary-loss behavior; unrelated models/features." }, { - "left": "issue:44671", - "right": "issue:44779", + "left": "issue:43065", + "right": "issue:44898", "accept": false, - "reason": "Both are v5 regressions, but one is CamemBERT MLM prediction drift and the other is Deepseek tokenizer output corruption; different models and code paths." + "reason": "Sam3PixelDecoder dummy Conv2d vs. Perceiver non-default-resolution interpolation bug; different components." }, { - "left": "issue:42175", - "right": "issue:43976", + "left": "issue:44261", + "right": "issue:44263", "accept": false, - "reason": "TensorFlow backend packaging and Python version incompatibility are unrelated." + "reason": "MLA q_a_layernorm precision/config issue vs. torch.split return values in GlmMoeDsaIndexer; unrelated defects." }, { - "left": "issue:43673", - "right": "issue:43761", + "left": "issue:43441", + "right": "issue:44683", "accept": false, - "reason": "Generation cache missing during chunked_prefill and CLIPVisionModel hidden_states=None are separate regressions." + "reason": "FlashAttention failure for Ministral vs. compiled flex_attention failure on torch>=2.9; both attention-related but not the same bug." }, { - "left": "issue:43901", - "right": "issue:43994", + "left": "issue:44016", + "right": "issue:45823", "accept": false, - "reason": "Docs mismatch for return_all_scores and incorrect SigLIP pipeline outputs are not the same bug." + "reason": "Notebook syntax error vs. Gemma4 PLE device mismatch; no shared underlying issue." }, { - "left": "issue:43784", - "right": "issue:45542", + "left": "issue:43262", + "right": "issue:44521", "accept": false, - "reason": "Import NameError in sentence-transformers integration and undefined tf backend from tensorboard-only install are different import-time issues." + "reason": "Audio apply_chat_template sampling-rate default vs. multimodal assistant_masks bug; same API surface, different bug." }, { - "left": "issue:43824", - "right": "issue:43976", + "left": "issue:44361", + "right": "issue:45593", "accept": false, - "reason": "Importing Qwen2_5_VLForConditionalGeneration and Python 3.9+ compatibility are unrelated." + "reason": "MLukeTokenizer AttributeError vs. D-FINE loss configuration issue; unrelated." }, { - "left": "issue:43450", - "right": "issue:44617", + "left": "issue:42175", + "right": "issue:43976", "accept": false, - "reason": "Batched video processor shape bug and Sam3Video CUDA OOM are different failures." + "reason": "Missing TensorFlow in torch-only install vs. Python version compatibility failure; different packaging problems." }, { - "left": "issue:44297", - "right": "issue:45397", + "left": "issue:43723", + "right": "issue:44162", "accept": false, - "reason": "Tokenizer save_pretrained metadata mismatch and gemma-4 zero3 loading are unrelated." + "reason": "AutoTokenizer loading problem in v5 vs. ESM2 broken inference; distinct models and symptoms." }, { - "left": "issue:43525", - "right": "issue:43881", + "left": "issue:44671", + "right": "issue:44779", "accept": false, - "reason": "Missing pad_token_id on Llama4Config and glm-4v-9b loading failure are not the same concrete bug." + "reason": "CamemBERT masked-LM regression vs. Deepseek tokenizer regression; both v5 regressions, but different code paths." }, { "left": "issue:43901", - "right": "issue:44246", + "right": "issue:43994", "accept": false, - "reason": "Pipeline docs mismatch and import-time slowness are unrelated." + "reason": "Documentation mismatch for return_all_scores vs. SigLIP2 bad model/pipeline outputs; unrelated." }, { - "left": "issue:43525", - "right": "issue:43643", + "left": "issue:43673", + "right": "issue:43761", "accept": false, - "reason": "Both involve config fields, but one is a specific missing pad_token_id attribute and the other is trust_remote_code returning incomplete objects; different issues." + "reason": "GenerationMixin chunked_prefill cache issue vs. CLIPVisionModel hidden_states bug; unrelated model internals." }, { - "left": "issue:44188", - "right": "issue:44336", + "left": "issue:43784", + "right": "issue:45542", "accept": false, - "reason": "Attention-kernel divergence and ANSI codes in loading_report are unrelated." + "reason": "Missing nn import in sentence-transformers integration vs. TensorBoard/TensorFlow backend confusion; different dependency issues." }, { - "left": "issue:44589", - "right": "issue:45362", + "left": "issue:43824", + "right": "issue:43976", "accept": false, - "reason": "Float8 storage lookup failure and Qwen3.5 chat crash are different runtime errors." + "reason": "Qwen2.5-VL import error vs. Python 3.9/3.10 compatibility issue; unrelated." }, { - "left": "issue:41628", - "right": "issue:43976", + "left": "issue:44297", + "right": "issue:45397", "accept": false, - "reason": "AutoImageProcessor import failure and Python-version support are unrelated." + "reason": "tokenizer.save_pretrained metadata mismatch vs. gemma-4 zero3 loading failure; different functionality." }, { - "left": "issue:45561", - "right": "issue:45563", + "left": "issue:43525", + "right": "issue:43881", "accept": false, - "reason": "A pytest-xdist file race and a stale warning in generate() are different bugs with no shared code path." + "reason": "Llama4Config missing pad_token_id vs. GLM-4V-9B loading failure; different models and bugs." } ] }, @@ -10966,29 +10990,33 @@ "component_index": 1, "component_count": 17, "review_unit_index": 35, - "review_unit_count": 49, + "review_unit_count": 50, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-427", + "cluster_id": "cluster-10000-428", "nodes": [ "issue:36246", + "issue:41628", "issue:41950", "issue:42491", "issue:42757", "issue:42886", - "issue:43257", - "issue:43262", "issue:43299", "issue:43352", + "issue:43450", + "issue:43525", "issue:43526", "issue:43531", "issue:43596", + "issue:43643", "issue:43673", "issue:43901", + "issue:43976", "issue:44162", + "issue:44188", "issue:44220", "issue:44242", "issue:44246", @@ -10999,31 +11027,33 @@ "issue:44485", "issue:44496", "issue:44561", + "issue:44589", + "issue:44617", "issue:44749", "issue:44779", "issue:44857", "issue:44871", - "issue:44877", "issue:45003", - "issue:45137", "issue:45200", "issue:45250", "issue:45292", "issue:45362", - "issue:45375", "issue:45397", "issue:45412", "issue:45561", - "issue:45584", + "issue:45563", "issue:45593", - "issue:45701", - "issue:45759", - "issue:45987", - "issue:46082", "issue:46123", "issue:46132" ], "soft_pairs": [ + "issue:43450|issue:44617", + "issue:43525|issue:43643", + "issue:43901|issue:44246", + "issue:44188|issue:44336", + "issue:41628|issue:43976", + "issue:44589|issue:45362", + "issue:45561|issue:45563", "issue:42491|issue:43299", "issue:44336|issue:44373", "issue:44295|issue:45397", @@ -11036,240 +11066,226 @@ "issue:36246|issue:42757", "issue:36246|issue:44351", "issue:44336|issue:44485", - "issue:44246|issue:44871", "issue:45003|issue:45561", + "issue:44246|issue:44871", "issue:43673|issue:44561", "issue:44242|issue:44857", - "issue:44779|issue:45292", "issue:44295|issue:44749", - "issue:43352|issue:45250", "issue:36246|issue:43596", + "issue:43352|issue:45250", + "issue:44779|issue:45292", "issue:44162|issue:45200", "issue:43901|issue:45003", - "issue:45412|issue:46123", - "issue:44496|issue:45003", - "issue:44877|issue:45375", - "issue:45987|issue:46082", - "issue:43262|issue:44246", - "issue:45292|issue:45584", - "issue:43257|issue:45137", - "issue:45375|issue:45759", - "issue:45292|issue:45701" + "issue:45412|issue:46123" ], - "prepared_review_unit_hash": "556a0e5a7b9dedababd629f7c405e8d1f7a4b69c592cbc712893e76b011a5a95", + "prepared_review_unit_hash": "053fcbee34f06dc052053ce62370aee7aaad2420619ab2ac9f97aa5d46bb8d4e", "component_budget": { - "node_count": 427, - "item_count": 427, - "soft_pair_count": 1576, - "serialized_chars": 816182, - "estimated_input_tokens": 204046, - "estimated_eval_tokens": 408348 + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 }, "budget": { - "node_count": 48, - "item_count": 48, - "soft_pair_count": 31, - "serialized_chars": 36830, - "estimated_input_tokens": 9208, - "estimated_eval_tokens": 18672 + "node_count": 47, + "item_count": 47, + "soft_pair_count": 30, + "serialized_chars": 35818, + "estimated_input_tokens": 8955, + "estimated_eval_tokens": 18166 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly a set of unrelated transformers bugs that only share broad vocabulary like model names, import errors, or configuration/loading paths. The soft pairs are largely false positives from topical similarity rather than true duplicates.", - "confidence": 0.97, - "canonical_issue_reason": "No single canonical issue stands out: the cluster spans separate bugs across imports, configs, tokenizers, pipelines, and model-specific regressions, so treating any one item as the duplicate anchor would be arbitrary.", + "summary": "Mostly a loose grab-bag of unrelated issues: import/export regressions, model-specific loading bugs, pipeline behavior, performance, docs, and test infrastructure. There is no clear duplicate cluster; only a few items are superficially similar around import failures or version regressions.", + "confidence": 0.69, + "canonical_issue_reason": "41628 is the closest broad representative of the import/export regression subset (`Cannot import name 'AutoImageProcessor' from 'transformers'`), which is the nearest shared theme here, though the overall cluster is still heterogeneous.", "canonical_pr_reason": null, - "best_issue_reason": "No issue is a strong representative duplicate target; even the closest-looking pairs still describe different concrete failures or different model families/versions.", + "best_issue_reason": "41628 is the best global issue candidate because it is a generic top-level import failure and most plausibly serves as the broad representative of the small import-API regression subset.", "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:43450", + "right": "issue:44617", + "accept": false, + "reason": "Both involve video/model behavior, but one is a batched video-processor shape bug and the other is a Sam3Video OOM issue; different failure modes and fixes." + }, + { + "left": "issue:43525", + "right": "issue:43643", + "accept": false, + "reason": "A missing `pad_token_id` attribute in `Llama4Config` is not the same as missing fields from `AutoConfig.from_pretrained(..., trust_remote_code=True)`." + }, + { + "left": "issue:43901", + "right": "issue:44246", + "accept": false, + "reason": "Docs mismatch for `return_all_scores` and intermittent import slowness are unrelated." + }, + { + "left": "issue:44188", + "right": "issue:44336", + "accept": false, + "reason": "Attention-kernel divergence under `torch.compile` is unrelated to ANSI codes emitted by loading report utilities." + }, + { + "left": "issue:41628", + "right": "issue:43976", + "accept": false, + "reason": "Both mention compatibility/import problems, but one is a missing `AutoImageProcessor` export and the other is a Python-version support issue." + }, + { + "left": "issue:44589", + "right": "issue:45362", + "accept": false, + "reason": "A missing float8 storage type and a Qwen3.5 chat crash are different concrete bugs." + }, + { + "left": "issue:45561", + "right": "issue:45563", + "accept": false, + "reason": "Pytest-xdist file race and a stale `num_return_sequences` warning do not share the same code path." + }, { "left": "issue:42491", "right": "issue:43299", "accept": false, - "reason": "Both are Qwen3-MoE related, but one is a LoRA compatibility problem and the other is a model-loading regression; different concrete failures." + "reason": "Both are Qwen/MoE-related regressions, but they affect different model families and loading paths; not the same bug." }, { "left": "issue:44336", "right": "issue:44373", "accept": false, - "reason": "ANSI output in loading_report and a wrong position_ids docstring are unrelated bugs." + "reason": "ANSI escape codes in loading reports and a wrong docstring for `position_ids` are unrelated." }, { "left": "issue:44295", "right": "issue:45397", "accept": false, - "reason": "One is a position_ids buffer access error; the other is a Gemma-4 ZeRO3/from_pretrained loading issue." + "reason": "A `position_ids` buffer access error and a Gemma-4 Zero3 `from_pretrained` issue are distinct failures." }, { "left": "issue:44242", "right": "issue:45593", "accept": false, - "reason": "Both mention missing losses, but they affect different models and different code paths." + "reason": "Missing auxiliary loss in MoE routing and missing auxiliary losses for D-FINE denoising off are different model-specific training bugs." }, { "left": "issue:44246", "right": "issue:44496", "accept": false, - "reason": "Slow import and unrecognized model config are unrelated." + "reason": "Import performance and an unrecognized model/config error are unrelated." }, { "left": "issue:41950", "right": "issue:44220", "accept": false, - "reason": "Video-classification image-processor lookup and audio feature extraction are different subsystems." + "reason": "Video-classification processor lookup and `_torch_extract_fbank_features()` are different subsystems and bugs." }, { "left": "issue:43531", "right": "issue:45362", "accept": false, - "reason": "Qwen3-MoE sliding-window behavior and a Qwen3.5 chat crash are different bugs." + "reason": "Both reference Qwen, but one is a `sliding_window` bug in MoE models and the other is a chat crash on Qwen3.5-35B; not the same fix." }, { "left": "issue:43526", "right": "issue:45003", "accept": false, - "reason": "BEiT label reduction and unsafe sys.modules access are unrelated." + "reason": "A Beit image processor label-reduction bug is unrelated to unsafe `sys.modules[]` access in `modeling_utils`." }, { "left": "issue:42886", "right": "issue:46132", "accept": false, - "reason": "Offline cache loading and AttentionInterface.register behavior are different issues." + "reason": "Offline cache loading and `AttentionInterface.register` behavior changes are different problems." }, { "left": "issue:36246", "right": "issue:42757", "accept": false, - "reason": "Both are import errors, but for different symbols and different packages." + "reason": "Both are import errors, but they come from different packages and symbols (`transformers.models.qwen2_5_vl` vs `huggingface_hub`)." }, { "left": "issue:36246", "right": "issue:44351", "accept": false, - "reason": "Qwen2.5 VL image processor import failure and HybridCache import failure are unrelated." + "reason": "Missing `Qwen2_5_VLImageProcessor` and missing `HybridCache` exports are separate API regressions." }, { "left": "issue:44336", "right": "issue:44485", "accept": false, - "reason": "Loading-report ANSI codes and GLM-5 RoPE implementation are different topics." + "reason": "Loading-report ANSI output and GLM-5 RoPE implementation are unrelated." }, { - "left": "issue:44246", - "right": "issue:44871", + "left": "issue:45003", + "right": "issue:45561", "accept": false, - "reason": "Import performance and Gemma eos_token configuration mismatch are unrelated." + "reason": "Unsafe `sys.modules[]` access and a pytest-xdist file race do not describe the same bug." }, { - "left": "issue:45003", - "right": "issue:45561", + "left": "issue:44246", + "right": "issue:44871", "accept": false, - "reason": "Unsafe sys.modules access and a pytest-xdist file race are different concurrency/path bugs." + "reason": "Import slowness is unrelated to Gemma-3 `eos_token_id` config inconsistency." }, { "left": "issue:43673", "right": "issue:44561", "accept": false, - "reason": "Chunked-prefill cache handling and removal of is_torch_fx_available are separate regressions." + "reason": "Missing GenerationMixin cache in chunked prefill and removal of `is_torch_fx_available` breaking trust-remote-code models are different regressions." }, { "left": "issue:44242", "right": "issue:44857", "accept": false, - "reason": "Missing load-balancing loss and an AMP float16 crash in LwDetrImageLoss are unrelated." + "reason": "Different auxiliary-loss/loss-scaling bugs in different models; no shared concrete code path." }, { - "left": "issue:44779", - "right": "issue:45292", + "left": "issue:44295", + "right": "issue:44749", "accept": false, - "reason": "Tokenizer regression and resize_token_embeddings not updating output embeddings are different code paths." + "reason": "A `position_ids` buffer error and a slow data-filtering regression after upgrade are unrelated." }, { - "left": "issue:44295", - "right": "issue:44749", + "left": "issue:36246", + "right": "issue:43596", "accept": false, - "reason": "A position_ids buffer error and slower filtering after upgrade are unrelated." + "reason": "An import error for a vision processor and an index error in `BertModel` are unrelated." }, { "left": "issue:43352", "right": "issue:45250", "accept": false, - "reason": "One is a specific model not supporting Flash Attention 2; the other is a generic Flash Attention 2 issue." + "reason": "One is a model-specific unsupported Flash Attention 2 issue; the other is only a generic Flash Attention 2 topic." }, { - "left": "issue:36246", - "right": "issue:43596", + "left": "issue:44779", + "right": "issue:45292", "accept": false, - "reason": "ImportError for a model processor and a DeepSpeed ZeRO3 IndexError are unrelated." + "reason": "Deepseek tokenizer correctness and `resize_token_embeddings` not updating output embeddings are different bugs." }, { "left": "issue:44162", "right": "issue:45200", "accept": false, - "reason": "ESM2 breakage and Gemma-4 mm_token_type_ids handling are different model-specific bugs." + "reason": "ESM2 breakage and Gemma 4 `mm_token_type_ids` defaults concern different models and code paths." }, { "left": "issue:43901", "right": "issue:45003", "accept": false, - "reason": "A documentation mismatch and unsafe sys.modules access are not duplicates." + "reason": "A pipeline docs mismatch and unsafe `sys.modules[]` access are unrelated." }, { "left": "issue:45412", "right": "issue:46123", "accept": false, - "reason": "RT-DETR memory retention and MaskGenerationPipeline partial-batch dropping are unrelated." - }, - { - "left": "issue:44496", - "right": "issue:45003", - "accept": false, - "reason": "Unrecognized model config and sys.modules access are different failures." - }, - { - "left": "issue:44877", - "right": "issue:45375", - "accept": false, - "reason": "Granite speech strict-config loading and Qwen3.5 MoE vision config field loss are different models and bugs." - }, - { - "left": "issue:45987", - "right": "issue:46082", - "accept": false, - "reason": "Cache-length return type and LlamaConfig head_dim validation are unrelated." - }, - { - "left": "issue:43262", - "right": "issue:44246", - "accept": false, - "reason": "Audio chat-template sampling rate and import slowdown are unrelated." - }, - { - "left": "issue:45292", - "right": "issue:45584", - "accept": false, - "reason": "Embedding resize behavior and Whisper empty-transcription handling are different issues." - }, - { - "left": "issue:43257", - "right": "issue:45137", - "accept": false, - "reason": "Qwen3 MoE weight conversion under accelerate/deepspeed and a ZeRO3 deque IndexError are different bugs." - }, - { - "left": "issue:45375", - "right": "issue:45759", - "accept": false, - "reason": "Both are Qwen multimodal config problems, but one drops a vision config field and the other fails to unwrap text_config; not the same concrete bug." - }, - { - "left": "issue:45292", - "right": "issue:45701", - "accept": false, - "reason": "A specific embedding-resize bug and a broad tokenization-version change report are not duplicates." + "reason": "RT-DETR memory release and MaskGenerationPipeline dropping partial-batch results are different runtime bugs." } ] }, @@ -11285,316 +11301,295 @@ "component_index": 1, "component_count": 17, "review_unit_index": 36, - "review_unit_count": 49, + "review_unit_count": 50, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-427", + "cluster_id": "cluster-10000-428", "nodes": [ "issue:42886", "issue:43065", "issue:43232", + "issue:43257", + "issue:43262", "issue:43408", "issue:43475", "issue:43540", - "issue:43646", "issue:43653", "issue:43673", - "issue:43701", "issue:43931", "issue:44060", "issue:44077", "issue:44162", - "issue:44297", + "issue:44246", "issue:44336", - "issue:44361", "issue:44368", "issue:44466", "issue:44479", + "issue:44496", "issue:44530", - "issue:44623", "issue:44734", - "issue:44749", - "issue:44869", - "issue:44871", "issue:44877", "issue:44933", "issue:44977", "issue:44987", "issue:44995", + "issue:45003", "issue:45137", "issue:45200", "issue:45229", - "issue:45278", + "issue:45292", "issue:45335", - "issue:45356", - "issue:45381", + "issue:45375", "issue:45397", "issue:45459", "issue:45563", + "issue:45584", "issue:45588", "issue:45600", "issue:45656", + "issue:45701", + "issue:45759", + "issue:45987", + "issue:46082", "issue:46123", - "issue:46132", "issue:46170" ], "soft_pairs": [ + "issue:44877|issue:45375", + "issue:45987|issue:46082", + "issue:44496|issue:45003", + "issue:45375|issue:45759", + "issue:43262|issue:44246", + "issue:45292|issue:45584", + "issue:43257|issue:45137", + "issue:45292|issue:45701", "issue:44987|issue:45459", "issue:43232|issue:44336", - "issue:44479|issue:46170", "issue:43065|issue:43408", "issue:44995|issue:45200", + "issue:44479|issue:46170", "issue:43475|issue:43540", - "issue:44466|issue:45335", "issue:42886|issue:45229", + "issue:44466|issue:45335", "issue:45397|issue:45656", "issue:44060|issue:44077", "issue:43653|issue:44162", - "issue:44995|issue:45588", "issue:44368|issue:44977", - "issue:45137|issue:45600", + "issue:44995|issue:45588", "issue:44368|issue:45563", + "issue:45137|issue:45600", "issue:44877|issue:46170", "issue:43673|issue:44734", "issue:43931|issue:44368", - "issue:44530|issue:46170", "issue:43065|issue:46123", - "issue:44933|issue:45459", - "issue:44297|issue:44623", - "issue:44869|issue:45356", - "issue:44361|issue:44749", - "issue:43646|issue:44162", - "issue:43232|issue:43673", - "issue:44623|issue:45656", - "issue:44749|issue:45656", - "issue:44871|issue:46132", - "issue:43673|issue:45278", - "issue:43701|issue:44749", - "issue:43475|issue:45381" + "issue:44530|issue:46170", + "issue:44933|issue:45459" ], - "prepared_review_unit_hash": "516a7c96ee3b653c56fdeb847492f194620b36112230fafcf8d86ae6d32d6379", + "prepared_review_unit_hash": "1b457941161f3e28ab34de6e36bf0d02941ed7e7990abf66dfc9b0d6cd4e3e73", "component_budget": { - "node_count": 427, - "item_count": 427, - "soft_pair_count": 1576, - "serialized_chars": 816182, - "estimated_input_tokens": 204046, - "estimated_eval_tokens": 408348 + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 }, "budget": { "node_count": 47, "item_count": 47, - "soft_pair_count": 32, - "serialized_chars": 36751, - "estimated_input_tokens": 9188, - "estimated_eval_tokens": 18632 + "soft_pair_count": 29, + "serialized_chars": 35852, + "estimated_input_tokens": 8963, + "estimated_eval_tokens": 18182 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The items are mostly unrelated regression reports that only share broad themes like tokenizers, generation, or multimodal models. None of the soft-edge pairs look like the same concrete bug or change, so all candidate duplicate links should be rejected.", - "confidence": 0.86, - "canonical_issue_reason": "issue:45229 is the strongest standalone issue to use as a representative: it is concrete, high-impact, and has the most discussion activity, making it the best single anchor among this mixed set.", + "summary": "This is not a true duplicate cluster: the items span many unrelated bugs and features (tokenizers, generation/cache, SAM3, Qwen, DeepSpeed, config loading, etc.). The soft pairs mostly share broad subsystems or keywords, but not the same concrete bug or change.", + "confidence": 0.95, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "issue:45229 is the best overall issue candidate because it describes a specific reproducible runtime failure with substantial discussion and clear user impact.", + "best_issue_reason": null, "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:44987", - "right": "issue:45459", + "left": "issue:44877", + "right": "issue:45375", "accept": false, - "reason": "Both involve loading/tokenizer behavior, but one is a model-loading regression and the other is protobuf-specific error masking; different bugs and code paths." + "reason": "Both mention config strictness, but they are different models and different missing-field failures." }, { - "left": "issue:43232", - "right": "issue:44336", + "left": "issue:45987", + "right": "issue:46082", "accept": false, - "reason": "Generation kwargs after sync_gpus is unrelated to ANSI codes emitted by loading_report; no shared failing behavior." + "reason": "Cache return type and Llama head_dim validation are unrelated issues." }, { - "left": "issue:44479", - "right": "issue:46170", + "left": "issue:44496", + "right": "issue:45003", "accept": false, - "reason": "One is a video-input regression in Qwen VL models; the other is a Windows/ROCm import dependency issue. Different subsystems." + "reason": "Model auto-loading failure and unsafe sys.modules access are different problems." }, { - "left": "issue:43065", - "right": "issue:43408", + "left": "issue:45375", + "right": "issue:45759", "accept": false, - "reason": "Both are SAM3-related, but one is a dummy Conv2d in the pixel decoder and the other is a model-type warning for tracker instantiation; not the same bug." + "reason": "Both are multimodal Qwen config issues, but they affect different fields and code paths." }, { - "left": "issue:44995", - "right": "issue:45200", + "left": "issue:43262", + "right": "issue:44246", "accept": false, - "reason": "Stale indexer cache on second forward pass is unrelated to Gemma 4 token-type IDs defaulting to zeros." + "reason": "Audio sampling-rate handling and import slowness are unrelated." }, { - "left": "issue:43475", - "right": "issue:43540", + "left": "issue:45292", + "right": "issue:45584", "accept": false, - "reason": "Different models and failures: missing fpn_position_embeddings in SAM 3 video versus a video-processing ValueError in Qwen3OmniMoe." + "reason": "resize_token_embeddings bug and Whisper empty-transcription failure are different bugs." }, { - "left": "issue:44466", - "right": "issue:45335", + "left": "issue:43257", + "right": "issue:45137", "accept": false, - "reason": "Both touch tied/embedding weights, but one is serialization inconsistency and the other is resize_token_embeddings not updating decoder embeddings; not mergeable as one fix." + "reason": "DeepSpeed appears in both, but one is weight conversion and the other is a ZeRO3 deque error." }, { - "left": "issue:42886", - "right": "issue:45229", + "left": "issue:45292", + "right": "issue:45701", "accept": false, - "reason": "Offline cache loading for tokenizers is unrelated to a multi-GPU CUDA OOM in Gemma4." + "reason": "Embedding resize behavior and tokenization version drift are not the same issue." }, { - "left": "issue:45397", - "right": "issue:45656", + "left": "issue:44987", + "right": "issue:45459", "accept": false, - "reason": "Gemma-4 ZeRO3 loading failure and a double optimizer-step deepspeed bug are separate training/runtime issues." + "reason": "Model loading regression and protobuf decode-error masking are unrelated." }, { - "left": "issue:44060", - "right": "issue:44077", + "left": "issue:43232", + "right": "issue:44336", "accept": false, - "reason": "Incorrect tied-weights warning in Qwen3-Next is unrelated to patchtsmixer post_init allowance." + "reason": "Generation kwargs handling and ANSI output formatting are unrelated." }, { - "left": "issue:43653", - "right": "issue:44162", + "left": "issue:43065", + "right": "issue:43408", "accept": false, - "reason": "BigBirdTokenizer special-token registration and ESM2 breakage are different tokenizer/model regressions." + "reason": "Both are SAM3-related, but one is a dummy conv module and the other is a model-type mismatch warning." }, { "left": "issue:44995", - "right": "issue:45588", - "accept": false, - "reason": "A stale cache bug in GlmMoeDsa is unrelated to flash_attention crashing on s_aux=None." - }, - { - "left": "issue:44368", - "right": "issue:44977", - "accept": false, - "reason": "One is a tie_word_embeddings warning during LoRA fine-tuning; the other is a flash-attention generation failure. Same family, different bug." - }, - { - "left": "issue:45137", - "right": "issue:45600", - "accept": false, - "reason": "DeepSpeed ZeRO3 deque underflow and removed Sam3LiteText auto-mapping references are unrelated." - }, - { - "left": "issue:44368", - "right": "issue:45563", + "right": "issue:45200", "accept": false, - "reason": "A model-configuration warning is not the same as a stale num_return_sequences warning in paged generate()." + "reason": "A stale indexer cache crash and missing mm_token_type_ids are different bugs." }, { - "left": "issue:44877", + "left": "issue:44479", "right": "issue:46170", "accept": false, - "reason": "Strict granite_speech config loading and Windows ROCm torch.distributed import failures are unrelated." + "reason": "A video-input regression and a Windows/ROCm distributed import error are unrelated." }, { - "left": "issue:43673", - "right": "issue:44734", + "left": "issue:43475", + "right": "issue:43540", "accept": false, - "reason": "Both involve cache/generation, but chunked_prefill cache absence and KV-cache continuation tensor indexing are different concrete defects." + "reason": "SAM3 vision encoder output and Qwen3OmniMoe video processing failures are different code paths." }, { - "left": "issue:43931", - "right": "issue:44368", + "left": "issue:42886", + "right": "issue:45229", "accept": false, - "reason": "Weight-shape mismatch on Qwen3-VL loading is unrelated to a tie_word_embeddings warning." + "reason": "Tokenizer offline-cache loading and Gemma4 CUDA OOM are unrelated." }, { - "left": "issue:44530", - "right": "issue:46170", + "left": "issue:44466", + "right": "issue:45335", "accept": false, - "reason": "PagedAttentionCache linear_attention crash and Windows/ROCm import errors do not share the same failing code path." + "reason": "Both involve tied/embedding weights, but the concrete failures are different." }, { - "left": "issue:43065", - "right": "issue:46123", + "left": "issue:45397", + "right": "issue:45656", "accept": false, - "reason": "SAM3 pixel-decoder dummy layer and MaskGenerationPipeline partial-batch handling are unrelated." + "reason": "Both involve DeepSpeed, but one is from_pretrained loading and the other is double optimizer stepping." }, { - "left": "issue:44933", - "right": "issue:45459", + "left": "issue:44060", + "right": "issue:44077", "accept": false, - "reason": "A nonexistent import from image_utils is unrelated to protobuf error masking in tokenizer decoding." + "reason": "Qwen3-Next tied-weight warning and PatchTSMixer post_init policy are unrelated." }, { - "left": "issue:44297", - "right": "issue:44623", + "left": "issue:43653", + "right": "issue:44162", "accept": false, - "reason": "Both are save_pretrained-related, but one is tokenizer_class mismatch and the other is missing processor files; not the same serialization bug." + "reason": "BigBirdTokenizer special-token registration and ESM2 breakage are different issues." }, { - "left": "issue:44869", - "right": "issue:45356", + "left": "issue:44368", + "right": "issue:44977", "accept": false, - "reason": "Whisper timestamp decoding crash and Kimi-K2.5 tokenizer regression are distinct tokenizer paths with different failure modes." + "reason": "One is a config warning about tie_word_embeddings; the other is a flash-attention generation regression." }, { - "left": "issue:44361", - "right": "issue:44749", + "left": "issue:44995", + "right": "issue:45588", "accept": false, - "reason": "MLukeTokenizer AttributeError and a performance regression after upgrading Transformers are different issues." + "reason": "Stale cache crash and flash_attention s_aux=None crash are unrelated." }, { - "left": "issue:43646", - "right": "issue:44162", + "left": "issue:44368", + "right": "issue:45563", "accept": false, - "reason": "Custom model initialization breakage in Transformers 5.0.0 is unrelated to ESM2 being broken." + "reason": "A tied-weights warning and a stale num_return_sequences warning are different bugs." }, { - "left": "issue:43232", - "right": "issue:43673", + "left": "issue:45137", + "right": "issue:45600", "accept": false, - "reason": "Both mention generation/cache, but the sync_gpus kwargs update bug is not the same as missing GenerationMixin cache during chunked prefill." + "reason": "DeepSpeed ZeRO3 deque failure and Sam3LiteText auto-mapping CI breakage are unrelated." }, { - "left": "issue:44623", - "right": "issue:45656", + "left": "issue:44877", + "right": "issue:46170", "accept": false, - "reason": "Processor save_pretrained file omissions and double optimizer.step under deepspeed are unrelated." + "reason": "Strict config loading and Windows ROCm distributed import errors are unrelated." }, { - "left": "issue:44749", - "right": "issue:45656", + "left": "issue:43673", + "right": "issue:44734", "accept": false, - "reason": "A data-processing slowdown after upgrade is not the same bug as optimizer.step being called twice." + "reason": "Both mention cache/generation, but they fail in different components and modes." }, { - "left": "issue:44871", - "right": "issue:46132", + "left": "issue:43931", + "right": "issue:44368", "accept": false, - "reason": "Gemma-3 eos_token_id inconsistency and AttentionInterface.register behavior change are unrelated." + "reason": "A Qwen3-VL weight-shape mismatch and a tie_word_embeddings warning are unrelated." }, { - "left": "issue:43673", - "right": "issue:45278", + "left": "issue:43065", + "right": "issue:46123", "accept": false, - "reason": "A specific cache bug during chunked prefill is not the same as broad import errors after upgrade." + "reason": "SAM3 dummy layer handling and MaskGenerationPipeline partial-batch behavior are unrelated." }, { - "left": "issue:43701", - "right": "issue:44749", + "left": "issue:44530", + "right": "issue:46170", "accept": false, - "reason": "Checkpoint key mismatch and preprocessing slowdown are different problems." + "reason": "PagedAttentionCache linear_attention error and Windows ROCm torch.distributed import failure are unrelated." }, { - "left": "issue:43475", - "right": "issue:45381", + "left": "issue:44933", + "right": "issue:45459", "accept": false, - "reason": "SAM3 vision encoder missing embeddings and Qwen2.5-VL wrong vision_position_ids are different multimodal regressions." + "reason": "Missing image_utils import and protobuf decode-error masking are unrelated." } ] }, @@ -11610,63 +11605,75 @@ "component_index": 1, "component_count": 17, "review_unit_index": 37, - "review_unit_count": 49, + "review_unit_count": 50, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-427", + "cluster_id": "cluster-10000-428", "nodes": [ "issue:28282", "issue:30990", "issue:36010", "issue:41628", - "issue:42175", "issue:42371", "issue:42673", "issue:42831", "issue:42994", "issue:43232", - "issue:43377", "issue:43475", "issue:43482", - "issue:43504", "issue:43525", - "issue:43526", "issue:43643", + "issue:43646", "issue:43673", + "issue:43701", "issue:43704", - "issue:43824", "issue:43873", "issue:44038", + "issue:44162", "issue:44188", "issue:44222", - "issue:44242", + "issue:44297", "issue:44351", - "issue:44496", + "issue:44361", "issue:44623", "issue:44661", - "issue:44792", + "issue:44749", + "issue:44869", + "issue:44871", "issue:44933", "issue:44945", - "issue:44964", "issue:45081", "issue:45092", "issue:45103", - "issue:45137", "issue:45230", + "issue:45278", "issue:45310", + "issue:45356", "issue:45375", + "issue:45381", "issue:45397", "issue:45478", - "issue:45561", - "issue:45712", + "issue:45656", "issue:45753", "issue:46077", - "issue:46123" + "issue:46123", + "issue:46132" ], "soft_pairs": [ + "issue:44297|issue:44623", + "issue:44869|issue:45356", + "issue:44361|issue:44749", + "issue:43646|issue:44162", + "issue:43232|issue:43673", + "issue:44623|issue:45656", + "issue:44749|issue:45656", + "issue:43673|issue:45278", + "issue:44871|issue:46132", + "issue:43701|issue:44749", + "issue:43475|issue:45381", "issue:42994|issue:43482", "issue:45230|issue:45397", "issue:43475|issue:43643", @@ -11676,8 +11683,8 @@ "issue:41628|issue:44933", "issue:42371|issue:45092", "issue:43475|issue:44661", - "issue:45081|issue:45103", "issue:45375|issue:45478", + "issue:45081|issue:45103", "issue:42994|issue:43525", "issue:45310|issue:45375", "issue:43873|issue:44945", @@ -11686,242 +11693,238 @@ "issue:42673|issue:43704", "issue:42831|issue:44222", "issue:44945|issue:45753", - "issue:44222|issue:45310", "issue:36010|issue:41628", - "issue:44964|issue:45137", - "issue:42994|issue:43824", - "issue:43673|issue:44188", - "issue:43232|issue:44792", - "issue:42175|issue:45712", - "issue:44038|issue:45310", - "issue:43526|issue:45561", - "issue:43504|issue:45561", - "issue:44496|issue:45561", - "issue:43377|issue:44242" + "issue:44222|issue:45310" ], - "prepared_review_unit_hash": "173fdb5d7c1982d4487dc8d349df6a89660979cd6e1542ab024a090cd858d528", + "prepared_review_unit_hash": "cd1499c83d15e10908153d230f2a7e0b1ac66d58262725a155d9dacd2a936356", "component_budget": { - "node_count": 427, - "item_count": 427, - "soft_pair_count": 1576, - "serialized_chars": 816182, - "estimated_input_tokens": 204046, - "estimated_eval_tokens": 408348 + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 }, "budget": { - "node_count": 47, - "item_count": 47, - "soft_pair_count": 31, - "serialized_chars": 36150, - "estimated_input_tokens": 9038, - "estimated_eval_tokens": 18332 + "node_count": 48, + "item_count": 48, + "soft_pair_count": 32, + "serialized_chars": 36963, + "estimated_input_tokens": 9241, + "estimated_eval_tokens": 18738 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly heterogeneous: many unrelated import/config/generation/quantization issues were grouped by loose similarity. Only one soft edge looks like a true duplicate thread around Qwen3.5-Moe from_pretrained failures caused by missing config fields.", - "confidence": 0.52, - "canonical_issue_reason": "Issue 45310 is the best canonical issue because it states the broader user-visible failure, and it matches the only plausible duplicate thread in this set; 45375 is a narrower root-cause variant of the same Qwen3.5-Moe problem.", + "summary": "This cluster is mostly unrelated Transformers issues connected only by broad themes like imports, generation, saving/loading, or model-family regressions. There are two clear exact duplicate pairs in the packet: issue:42673/issue:43704 (Qwen3ForCausalLM VRAM leak in multiple dataloader threads) and issue:45310/issue:45478 (Qwen3.5 MoE from_pretrained error). The remaining soft edges look like conservative false positives rather than true duplicates.", + "confidence": 0.91, + "canonical_issue_reason": "issue:42673 is the strongest canonical for its exact duplicate pair because it is the earlier, fuller report of the Qwen3ForCausalLM VRAM leak; issue:43704 is an exact duplicate. Note that the packet also contains a separate exact duplicate pair, issue:45310/issue:45478, so this cluster is not a single duplicate set.", "canonical_pr_reason": null, - "best_issue_reason": "45310 is the most representative and actionable issue among the near-duplicate candidates, with the clearest general symptom description.", + "best_issue_reason": "issue:42673 is the best representative overall because it has the fullest discussion/history among the exact duplicate pairs and is the clearest stable reference point for triage.", "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:44297", + "right": "issue:44623", + "accept": false, + "reason": "Both mention save_pretrained, but one is a tokenizer_config mismatch and the other is missing processor files; different failure modes and code paths." + }, + { + "left": "issue:44869", + "right": "issue:45356", + "accept": false, + "reason": "Both are tokenizer/codec-related, but one is a Whisper timestamp decode crash and the other is a Kimi-K2.5 tokenizer regression with a different warning/codec issue." + }, + { + "left": "issue:44361", + "right": "issue:44749", + "accept": false, + "reason": "Different problems: MLukeTokenizer task AttributeError vs a general performance slowdown after upgrade." + }, + { + "left": "issue:43646", + "right": "issue:44162", + "accept": false, + "reason": "Both are Transformers v5 regressions, but one is custom model initialization and the other is ESM2-specific breakage; not the same underlying bug." + }, + { + "left": "issue:43232", + "right": "issue:43673", + "accept": false, + "reason": "Both involve generation internals, but one is sync_gpus/update_model_kwargs behavior and the other is chunked_prefill cache handling; distinct code paths." + }, + { + "left": "issue:44623", + "right": "issue:45656", + "accept": false, + "reason": "processor.save_pretrained file issues are unrelated to a deepspeed optimizer-step counting bug." + }, + { + "left": "issue:44749", + "right": "issue:45656", + "accept": false, + "reason": "A data-filtering slowdown after upgrade is unrelated to optimizer.step being called twice." + }, + { + "left": "issue:43673", + "right": "issue:45278", + "accept": false, + "reason": "One is a specific GenerationMixin cache regression; the other is a broad import-error report after upgrade." + }, + { + "left": "issue:44871", + "right": "issue:46132", + "accept": false, + "reason": "Gemma eos_token_id inconsistency is unrelated to AttentionInterface.register changing registered-function behavior." + }, + { + "left": "issue:43701", + "right": "issue:44749", + "accept": false, + "reason": "Checkpoint resume key mismatch is unrelated to dataset filtering performance." + }, + { + "left": "issue:43475", + "right": "issue:45381", + "accept": false, + "reason": "Both are vision/video model issues, but one is a missing SAM3 attribute and the other is incorrect Qwen2.5-VL video position IDs." + }, { "left": "issue:42994", "right": "issue:43482", "accept": false, - "reason": "Both involve model loading/saving failures, but they are different models and different code paths (quantized save vs GGUF load/version compatibility)." + "reason": "Quantized model saving failure is different from Qwen2.5-GGUF loading failure." }, { "left": "issue:45230", "right": "issue:45397", "accept": false, - "reason": "Generic bug report vs a specific ZeRO3/from_pretrained issue; no evidence they describe the same defect." + "reason": "issue:45230 is too generic to establish a shared bug, while 45397 is a specific gemma-4 zero3 from_pretrained problem." }, { "left": "issue:43475", "right": "issue:43643", "accept": false, - "reason": "Different subsystems and failures: missing SAM vision encoder attribute vs trust_remote_code config field loss." + "reason": "Different failures: SAM3 output attribute missing vs trust_remote_code returning missing fields from AutoConfig." }, { "left": "issue:28282", "right": "issue:30990", "accept": false, - "reason": "PyTorch missing ImportError vs Sentence Transformers loading hang are unrelated symptoms and causes." + "reason": "An ImportError about PyTorch missing is unrelated to Sentence Transformers getting stuck during load." }, { "left": "issue:44038", "right": "issue:44623", "accept": false, - "reason": "Qwen3-VL-Moe/transformers v5 error is unrelated to processor.save_pretrained missing files." + "reason": "Qwen3-VL-Moe v5 incompatibility and processor.save_pretrained missing files are unrelated issues." }, { "left": "issue:46077", "right": "issue:46123", "accept": false, - "reason": "Different pipeline behavior: Gemma4 bidirectional attention masking vs MaskGenerationPipeline final-batch handling." + "reason": "Gemma4 causal-mask construction and MaskGenerationPipeline final-batch handling are different bugs." }, { "left": "issue:41628", "right": "issue:44933", "accept": false, - "reason": "Both are import-related, but they concern different missing symbols and different implementation areas." + "reason": "Both are import errors, but for different symbols/modules and likely different fixes." }, { "left": "issue:42371", "right": "issue:45092", "accept": false, - "reason": "TensorFlow backend packaging vs remote-code checkpoint/meta-init incompatibility are different problems." + "reason": "TF32 API guidance is unrelated to remote-code checkpoint meta-initialization incompatibility." }, { "left": "issue:43475", "right": "issue:44661", "accept": false, - "reason": "Unrelated model attribute failure vs add-new-model-like/tokenizer-mapping-name logic bug." + "reason": "A SAM3 video attribute error is unrelated to add-new-model-like failing on TOKENIZER_MAPPING_NAMES." }, { - "left": "issue:45081", - "right": "issue:45103", + "left": "issue:45375", + "right": "issue:45478", "accept": false, - "reason": "Tokenizer regex crash and auto_docstring annotations crash are unrelated code paths." + "reason": "Both involve Qwen3.5 MoE, but one is a missing vision-config field and the other is a general from_pretrained error; not enough to treat as the same bug." }, { - "left": "issue:45375", - "right": "issue:45478", - "accept": true, - "reason": "Both point to the same Qwen3.5-Moe from_pretrained failure, with 45375 giving the likely root cause: a missing vision config field dropped by strict handling." + "left": "issue:45081", + "right": "issue:45103", + "accept": false, + "reason": "Mistral regex patching and auto_docstring future-annotations crashes are unrelated AttributeErrors in different subsystems." }, { "left": "issue:42994", "right": "issue:43525", "accept": false, - "reason": "Quantized model saving failure and Llama4Config missing pad_token_id are separate config/save issues." + "reason": "Quantized model saving failure is different from Llama4Config missing pad_token_id." }, { "left": "issue:45310", "right": "issue:45375", - "accept": true, - "reason": "Same underlying Qwen3.5-Moe loading bug; one describes the user-facing from_pretrained error, the other identifies the missing field causing it." + "accept": false, + "reason": "Both are Qwen3.5 MoE-related, but one is a from_pretrained error and the other is a missing vision-config field; not clearly the same underlying defect." }, { "left": "issue:43873", "right": "issue:44945", "accept": false, - "reason": "Offloading/quantization behavior is unrelated to pipeline parallelism output divergence." + "reason": "Quantization/offloading behavior is unrelated to incorrect pipeline-parallel LLM outputs." }, { "left": "issue:43232", "right": "issue:44188", "accept": false, - "reason": "Generation kwargs update bug after sync_gpus is a different problem from attention-kernel divergence under torch.compile." + "reason": "Generation kwargs after sync_gpus is unrelated to attention-kernel divergence under torch.compile." }, { "left": "issue:41628", "right": "issue:44351", "accept": false, - "reason": "Different import errors: AutoImageProcessor missing vs HybridCache missing." + "reason": "Different missing imports: AutoImageProcessor vs HybridCache." }, { "left": "issue:42673", "right": "issue:43704", - "accept": false, - "reason": "Same model family and symptom wording, but the later issue is a separate report with no evidence of being the same defect instance." + "accept": true, + "reason": "Exact same title and same reported bug: Qwen3ForCausalLM VRAM leak in multiple dataloader threads." }, { "left": "issue:42831", "right": "issue:44222", "accept": false, - "reason": "FP8 accuracy regression and FP8 moe save_pretrained failure are distinct FP8-related issues." + "reason": "Both are FP8-related, but one is an accuracy issue and the other is save_pretrained for moe models." }, { "left": "issue:44945", "right": "issue:45753", "accept": false, - "reason": "Pipeline parallelism output corruption and Qwen3_5 infinite loop on a specific image are different failures." - }, - { - "left": "issue:44222", - "right": "issue:45310", - "accept": false, - "reason": "Saving FP8 MOE checkpoints and Qwen3.5-Moe from_pretrained loading errors affect different operations." + "reason": "Pipeline-parallel wrong output is unrelated to a Qwen3_5 infinite loop on a specific image." }, { "left": "issue:36010", "right": "issue:41628", "accept": false, - "reason": "Different missing imports from different modules; not the same symbol or regression." - }, - { - "left": "issue:44964", - "right": "issue:45137", - "accept": false, - "reason": "Model loading incompatibility and DeepSpeed ZeRO3 deque error are unrelated." - }, - { - "left": "issue:42994", - "right": "issue:43824", - "accept": false, - "reason": "Quantized saving failure vs missing Qwen2.5-VL import are separate issues." - }, - { - "left": "issue:43673", - "right": "issue:44188", - "accept": false, - "reason": "GenerationMixin cache regression and attention-kernel branching divergence are different generation problems." - }, - { - "left": "issue:43232", - "right": "issue:44792", - "accept": false, - "reason": "Generation kwargs handling and janus image-generation test failure are not the same bug." - }, - { - "left": "issue:42175", - "right": "issue:45712", - "accept": false, - "reason": "TensorFlow packaging omission and leftover dummy PT classes are unrelated backend/package hygiene issues." + "reason": "Different import failures: GenerationMixin import vs AutoImageProcessor import." }, { - "left": "issue:44038", + "left": "issue:44222", "right": "issue:45310", "accept": false, - "reason": "Both mention Qwen models and transformers v5, but they involve different models and failure modes." - }, - { - "left": "issue:43526", - "right": "issue:45561", - "accept": false, - "reason": "Image processor label reduction bug and xdist race in testing utils are unrelated." - }, - { - "left": "issue:43504", - "right": "issue:45561", - "accept": false, - "reason": "Pretrained preset loading failure and xdist file-race bug are different code paths." - }, - { - "left": "issue:44496", - "right": "issue:45561", - "accept": false, - "reason": "Unrecognized model/config loading error vs testing-utils race condition are not the same issue." - }, - { - "left": "issue:43377", - "right": "issue:44242", - "accept": false, - "reason": "Padding-mask batch consistency bug in MIMI encoder vs MoE load-balancing loss omission are different algorithmic issues." + "reason": "FP8 moe save_pretrained and Qwen3.5-Moe from_pretrained errors are different bugs." } ] }, "evaluator_result": { - "accept": true, - "feedback": "The summary is grounded in the packet and stays mostly conservative. The only accepted soft edges are the Qwen3.5-Moe from_pretrained/thread of reports, which is supported by the titles and descriptions. The rest of the pair rejections are appropriately cautious and do not overclaim evidence." + "accept": false, + "feedback": "The summary is not fully grounded: the packet contains another exact duplicate pair besides 42673/43704, namely issue:45310 and issue:45478, which have the same title and clearly describe the same Qwen3.5 MoE from_pretrained error. Because the response says there is only one clear duplicate pair, it overstates the evidence. The other soft-edge rejections are generally conservative, but the missing duplicate makes the overall assessment imprecise." }, "evaluator_used": true, - "retried": false, + "retried": true, "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null @@ -11931,66 +11934,75 @@ "component_index": 1, "component_count": 17, "review_unit_index": 38, - "review_unit_count": 49, + "review_unit_count": 50, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-427", + "cluster_id": "cluster-10000-428", "nodes": [ "issue:30333", "issue:30990", "issue:33290", - "issue:34689", "issue:35141", - "issue:36331", "issue:42175", "issue:42222", - "issue:42831", "issue:42898", - "issue:43116", + "issue:42994", + "issue:43232", "issue:43278", + "issue:43377", "issue:43479", - "issue:43644", + "issue:43504", + "issue:43526", + "issue:43673", "issue:43723", "issue:43784", "issue:43824", "issue:43874", "issue:43976", + "issue:44038", "issue:44060", "issue:44079", "issue:44188", - "issue:44190", - "issue:44222", - "issue:44295", - "issue:44297", + "issue:44242", "issue:44373", "issue:44448", + "issue:44496", "issue:44568", "issue:44623", "issue:44734", "issue:44779", + "issue:44792", "issue:44869", "issue:44933", + "issue:44964", "issue:45092", "issue:45103", "issue:45125", "issue:45137", "issue:45200", - "issue:45245", "issue:45310", "issue:45356", "issue:45405", - "issue:45656", + "issue:45561", "issue:45712", - "issue:45812", - "issue:46082", - "issue:46132" + "issue:45812" ], "soft_pairs": [ - "issue:42222|issue:45200", + "issue:42994|issue:43824", + "issue:44964|issue:45137", + "issue:43232|issue:44792", + "issue:43673|issue:44188", + "issue:44038|issue:45310", + "issue:42175|issue:45712", + "issue:43526|issue:45561", + "issue:43504|issue:45561", + "issue:44496|issue:45561", + "issue:43377|issue:44242", "issue:42175|issue:44933", + "issue:42222|issue:45200", "issue:42898|issue:44568", "issue:43976|issue:44933", "issue:30990|issue:35141", @@ -12004,252 +12016,227 @@ "issue:43784|issue:43824", "issue:44734|issue:45103", "issue:43723|issue:44568", - "issue:42898|issue:44779", "issue:42898|issue:45356", + "issue:42898|issue:44779", "issue:43874|issue:45092", "issue:45405|issue:45712", - "issue:44060|issue:45125", "issue:42898|issue:44448", - "issue:44295|issue:45656", - "issue:44448|issue:46132", - "issue:44933|issue:45712", - "issue:44222|issue:44297", - "issue:44297|issue:45656", - "issue:43116|issue:44190", - "issue:43479|issue:43644", - "issue:44190|issue:46082", - "issue:30333|issue:36331", - "issue:44222|issue:44623", - "issue:42831|issue:44623", - "issue:30333|issue:34689", - "issue:45245|issue:46132" + "issue:44060|issue:45125" ], - "prepared_review_unit_hash": "d4ce0f74d1f13ad06cedaf63aeae42e8a3b11a1c802907832ba78b286e32e6fa", + "prepared_review_unit_hash": "673af8d6d3170d33f11597bec5cacef15485e4eaa6cfc5e2c4b19ee73c0b8383", "component_budget": { - "node_count": 427, - "item_count": 427, - "soft_pair_count": 1576, - "serialized_chars": 816182, - "estimated_input_tokens": 204046, - "estimated_eval_tokens": 408348 + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 }, "budget": { - "node_count": 48, - "item_count": 48, - "soft_pair_count": 34, - "serialized_chars": 37988, - "estimated_input_tokens": 9497, - "estimated_eval_tokens": 19250 + "node_count": 47, + "item_count": 47, + "soft_pair_count": 31, + "serialized_chars": 36549, + "estimated_input_tokens": 9138, + "estimated_eval_tokens": 18532 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is highly heterogeneous: tokenizer regressions, multimodal/model-loading bugs, training/DeepSpeed issues, packaging/import problems, and serving/runtime crashes are all mixed together. No soft pair looks like the same underlying bug, so I would not merge any of them. Issue 42898 is the closest thing to a hub because it is a broad v5 tokenizer-regression report, but it does not really represent the whole set.", - "confidence": 0.43, - "canonical_issue_reason": "Issue 42898 is the most central representative only in the narrow sense that it is a broad v5 tokenizer regression and overlaps thematically with several other tokenizer-related reports; the rest of the cluster is too mixed for a strong canonical fit.", + "summary": "The cluster is mostly a grab bag of unrelated issues, with a loose concentration around tokenizer/v5 regressions and multimodal/model-loading bugs; none of the soft pairs look like true duplicates. I would anchor the issue cluster on the broad tokenizer regression report, with a more concrete tokenizer bug as the strongest representative.", + "confidence": 0.74, + "canonical_issue_reason": "Issue #42898 is the broadest and most central regression report in the set: it describes a v5 tokenizer behavior change and is thematically closest to several nearby tokenizer-related reports.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 42898 is the best issue to anchor the cluster because it is the broadest, most reusable regression report among the tokenizer-related items, even though the overall cluster is not cohesive.", + "best_issue_reason": "Issue #45812 is the cleanest high-signal report: it has a concrete, model-specific tokenizer failure with a clear silent regression from v4 to v5.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:42222", - "right": "issue:45200", + "left": "issue:42994", + "right": "issue:43824", + "accept": false, + "reason": "Quantized model saving and a missing Qwen2.5-VL import symbol are unrelated failures." + }, + { + "left": "issue:44964", + "right": "issue:45137", + "accept": false, + "reason": "Phi-4 multimodal loading and a DeepSpeed ZeRO3 deque error affect different code paths." + }, + { + "left": "issue:43232", + "right": "issue:44792", + "accept": false, + "reason": "Generation kwargs syncing and the janus image-generation test failure are different bugs." + }, + { + "left": "issue:43673", + "right": "issue:44188", + "accept": false, + "reason": "Chunked-prefill cache handling is not the same as the torch.compile attention-kernel branching issue." + }, + { + "left": "issue:44038", + "right": "issue:45310", + "accept": false, + "reason": "Both involve Qwen3 VL/MoE models, but they report different versions and distinct from_pretrained failures." + }, + { + "left": "issue:42175", + "right": "issue:45712", + "accept": false, + "reason": "Missing tensorflow in the torch extra and leftover dummy PT classes are separate packaging problems." + }, + { + "left": "issue:43526", + "right": "issue:45561", + "accept": false, + "reason": "A BeitImageProcessorFast label bug is unrelated to the xdist captured_info.txt race." + }, + { + "left": "issue:43504", + "right": "issue:45561", + "accept": false, + "reason": "Legacy-field model loading and a pytest-xdist file race are not the same issue." + }, + { + "left": "issue:44496", + "right": "issue:45561", "accept": false, - "reason": "Different model families and failure modes: VitPose module breakage vs Gemma 4 mm_token_type_ids defaulting." + "reason": "An unrecognized-model config error is unrelated to the testing-utils race condition." + }, + { + "left": "issue:43377", + "right": "issue:44242", + "accept": false, + "reason": "Missing padding-mask support in MIMI encoding is different from the load-balancing-loss gating bug." }, { "left": "issue:42175", "right": "issue:44933", "accept": false, - "reason": "Both touch import/package surfaces, but one is backend extra installation and the other is a missing image_utils import." + "reason": "Backend dependency resolution and a missing image_utils import are unrelated." + }, + { + "left": "issue:42222", + "right": "issue:45200", + "accept": false, + "reason": "Broken vitpose models and Gemma 4 mm_token_type_ids defaulting are different multimodal bugs." }, { "left": "issue:42898", "right": "issue:44568", "accept": false, - "reason": "Both are tokenizer regressions in v5, but they affect different code paths: cleanup behavior vs missing BOS/EOS insertion." + "reason": "Both are tokenizer regressions in v5, but one is clean_up_tokenization_spaces behavior and the other is add_special_tokens BOS/EOS handling." }, { "left": "issue:43976", "right": "issue:44933", "accept": false, - "reason": "Unrelated problems: Python version compatibility vs a nonexistent import from image_utils." + "reason": "Python version compatibility and a missing import are unrelated." }, { "left": "issue:30990", "right": "issue:35141", "accept": false, - "reason": "Loading hang and embedding reinitialization are different bugs with no shared code path." + "reason": "Sentence-transformer loading hangs and embedding reinitialization in post_init are different problems." }, { "left": "issue:43479", "right": "issue:44869", "accept": false, - "reason": "Different subsystems: multimodal config defaulting vs Whisper timestamp decode crash." + "reason": "Default multimodal config initialization and Whisper timestamp decoding crash are unrelated." }, { "left": "issue:43479", "right": "issue:44079", "accept": false, - "reason": "Config initialization bug and ModelOutput key assignment bug are not the same issue." + "reason": "Config defaulting for multimodal submodules does not match ModelOutput key assignment logic." }, { "left": "issue:43278", "right": "issue:45137", "accept": false, - "reason": "Training/eval dtype drift is unrelated to the DeepSpeed ZeRO3 deque underflow." + "reason": "A dtype mismatch in evaluation is unrelated to the DeepSpeed deque pop error." }, { "left": "issue:44623", "right": "issue:45310", "accept": false, - "reason": "Both mention loading/saving, but one is missing save_pretrained files and the other is a Qwen3.5 Moe from_pretrained failure." + "reason": "Missing files on processor.save_pretrained and Qwen3.5 MoE from_pretrained failure are different load/save bugs." }, { - "left": "issue:42831", + "left": "issue:42898", "right": "issue:45812", "accept": false, - "reason": "FP8 accuracy degradation is unrelated to Granite tokenizer ID regression." + "reason": "Both are tokenizer regressions, but they affect different behaviors and model families, so they are not the same concrete bug." }, { "left": "issue:44188", "right": "issue:44373", "accept": false, - "reason": "Runtime attention-kernel divergence and a wrong docstring are not the same underlying bug." + "reason": "A torch.compile attention-kernel divergence is unrelated to a docstring typo." }, { "left": "issue:30333", "right": "issue:33290", "accept": false, - "reason": "MLflow reporting on failed training jobs is unrelated to Adafactor/DeepSpeed OOM." + "reason": "MLflow reporting on failed training jobs and AdaFactor OOM in DeepSpeed are separate training issues." }, { "left": "issue:43784", "right": "issue:43824", "accept": false, - "reason": "Two different import failures affecting different symbols and packages." + "reason": "A sentence-transformers import NameError and a Qwen2.5-VL import error are different import failures." }, { "left": "issue:44734", "right": "issue:45103", "accept": false, - "reason": "Serving KV-cache indexing crash and auto-docstring annotation handling are unrelated code paths." + "reason": "KV-cache continuation indexing in transformers serve is unrelated to the auto_docstring future-annotations crash." }, { "left": "issue:43723", "right": "issue:44568", "accept": false, - "reason": "Both are tokenizer-related, but one is a general AutoTokenizer loading issue and the other is a specific special-token behavior regression." + "reason": "Tokenizer loading in v5 and BOS/EOS add_special_tokens behavior are different tokenizer regressions." }, { "left": "issue:42898", - "right": "issue:44779", + "right": "issue:45356", "accept": false, - "reason": "Both are v5 tokenizer regressions, but they affect different model families and different tokenizer logic." + "reason": "Clean-up spacing behavior in v5 and Kimi-K2.5 codec handling are separate tokenizer issues." }, { "left": "issue:42898", - "right": "issue:45356", + "right": "issue:44779", "accept": false, - "reason": "Different tokenizer regressions: clean_up_tokenization_spaces vs Kimi-K2.5 codec handling/warning behavior." + "reason": "These are both tokenizer regressions, but for different models and different failure modes." }, { "left": "issue:43874", "right": "issue:45092", "accept": false, - "reason": "Missing image patch helper and remote-code meta-init incompatibility are distinct multimodal issues." + "reason": "A missing image-patch count method and an InternVL2 meta-init incompatibility are not the same bug." }, { "left": "issue:45405", "right": "issue:45712", "accept": false, - "reason": "A dependency version pin problem is unrelated to leftover dummy class exports." - }, - { - "left": "issue:44060", - "right": "issue:45125", - "accept": false, - "reason": "Different Qwen3 issues: incorrect tied-weights warning vs missing tensor-parallel plan." + "reason": "A bumped minimum PEFT version and dummy PT class cleanup are unrelated." }, { "left": "issue:42898", "right": "issue:44448", "accept": false, - "reason": "Tokenizer cleanup regression and Pegasus output drift are different failure modes." - }, - { - "left": "issue:44295", - "right": "issue:45656", - "accept": false, - "reason": "Reading a registered buffer and optimizer stepping twice are unrelated." - }, - { - "left": "issue:44448", - "right": "issue:46132", - "accept": false, - "reason": "Model output regression and AttentionInterface.register behavior change are not the same bug." - }, - { - "left": "issue:44933", - "right": "issue:45712", - "accept": false, - "reason": "Both are export/import-adjacent, but one is a missing image_utils import and the other is stale dummy PT classes." - }, - { - "left": "issue:44222", - "right": "issue:44297", - "accept": false, - "reason": "Save-pretrained behavior for FP8/MoE models is unrelated to tokenizer_class metadata mismatch." - }, - { - "left": "issue:43116", - "right": "issue:44190", - "accept": false, - "reason": "Different example scripts and different failure modes: empty classification results vs local dataset loading." - }, - { - "left": "issue:43479", - "right": "issue:43644", - "accept": false, - "reason": "Multimodal config defaults and non-persistent buffer junk are unrelated model initialization bugs." - }, - { - "left": "issue:44190", - "right": "issue:46082", - "accept": false, - "reason": "Example dataset loading failure is unrelated to LlamaConfig head_dim validation." - }, - { - "left": "issue:30333", - "right": "issue:36331", - "accept": false, - "reason": "Failed-job MLflow reporting is unrelated to CustomTrainer.compute_loss signature breakage." - }, - { - "left": "issue:44222", - "right": "issue:44623", - "accept": false, - "reason": "Both mention save_pretrained, but one is FP8/MoE model serialization and the other is missing processor artifact files." - }, - { - "left": "issue:42831", - "right": "issue:44623", - "accept": false, - "reason": "FP8 accuracy issues are unrelated to processor save_pretrained file omission." - }, - { - "left": "issue:45245", - "right": "issue:46132", - "accept": false, - "reason": "Category cardinality limits and AttentionInterface.register behavior changes are different issues." + "reason": "A tokenizer behavior regression and a model output discrepancy for Pegasus are different issues." }, { - "left": "issue:44297", - "right": "issue:45656", + "left": "issue:44060", + "right": "issue:45125", "accept": false, - "reason": "Tokenizer config mismatch is unrelated to the DeepSpeed optimizer-step double call." + "reason": "An incorrect tied-weights warning in Qwen3-Next is not the same as a missing tensor-parallel plan in Qwen3.5 MoE." } ] }, @@ -12265,68 +12252,81 @@ "component_index": 1, "component_count": 17, "review_unit_index": 39, - "review_unit_count": 49, + "review_unit_count": 50, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-427", + "cluster_id": "cluster-10000-428", "nodes": [ "issue:22355", "issue:28282", "issue:30333", "issue:30990", + "issue:34689", "issue:35141", + "issue:36331", "issue:38175", - "issue:39401", "issue:41628", "issue:42757", "issue:42831", "issue:42907", + "issue:43116", "issue:43232", "issue:43479", + "issue:43644", "issue:43723", "issue:43761", - "issue:43824", - "issue:44038", + "issue:44190", "issue:44206", "issue:44222", - "issue:44261", "issue:44263", "issue:44295", "issue:44297", "issue:44360", "issue:44373", - "issue:44442", + "issue:44448", "issue:44483", "issue:44484", "issue:44485", "issue:44514", - "issue:44568", "issue:44623", - "issue:44829", "issue:44871", - "issue:44913", + "issue:44933", "issue:44977", - "issue:44991", "issue:45071", "issue:45092", "issue:45103", - "issue:45278", + "issue:45245", "issue:45310", "issue:45357", "issue:45406", + "issue:45656", + "issue:45712", "issue:45834", - "issue:46002", "issue:46050", - "issue:46082" + "issue:46082", + "issue:46132" ], "soft_pairs": [ - "issue:22355|issue:28282", - "issue:44206|issue:45103", + "issue:44295|issue:45656", + "issue:44448|issue:46132", + "issue:44933|issue:45712", + "issue:44222|issue:44297", + "issue:44297|issue:45656", + "issue:43116|issue:44190", + "issue:43479|issue:43644", + "issue:44190|issue:46082", + "issue:44222|issue:44623", + "issue:30333|issue:36331", + "issue:42831|issue:44623", + "issue:30333|issue:34689", + "issue:45245|issue:46132", "issue:45071|issue:46082", + "issue:22355|issue:28282", "issue:41628|issue:42757", + "issue:44206|issue:45103", "issue:28282|issue:30333", "issue:30333|issue:35141", "issue:44297|issue:45310", @@ -12340,242 +12340,243 @@ "issue:44373|issue:44485", "issue:44514|issue:45103", "issue:43723|issue:44871", + "issue:45092|issue:45357", "issue:44263|issue:44360", "issue:30990|issue:38175", - "issue:45092|issue:45357", - "issue:42907|issue:46050", - "issue:43723|issue:44991", - "issue:44829|issue:46002", - "issue:43723|issue:44442", - "issue:44261|issue:45092", - "issue:42831|issue:45310", - "issue:43479|issue:44913", - "issue:41628|issue:43824", - "issue:44263|issue:44373", - "issue:42831|issue:44038", - "issue:39401|issue:44568", - "issue:42757|issue:45278" + "issue:42907|issue:46050" ], - "prepared_review_unit_hash": "f1548ea9b576a552f60db2cdd445e65c7b8d2d2a95ca9f295927e683b8da6480", + "prepared_review_unit_hash": "5a8150cde9b810f48fd881f556beaf5d795db9dcb8a9bdef078c070f149a7558", "component_budget": { - "node_count": 427, - "item_count": 427, - "soft_pair_count": 1576, - "serialized_chars": 816182, - "estimated_input_tokens": 204046, - "estimated_eval_tokens": 408348 + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 32, - "serialized_chars": 37045, - "estimated_input_tokens": 9262, - "estimated_eval_tokens": 18780 + "soft_pair_count": 34, + "serialized_chars": 37629, + "estimated_input_tokens": 9408, + "estimated_eval_tokens": 19072 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly heterogeneous: import regressions, tokenizer/model-loading bugs, config issues, and unrelated training/runtime problems. None of the soft pairs look like the same concrete bug closely enough to merge, though a few live in the same broad v5 import/tokenizer regression area.", - "confidence": 0.68, - "canonical_issue_reason": "issue:45278 is the broadest umbrella item in the set, capturing the general post-upgrade import-regression theme that also shows up in several nearby reports; the rest of the cluster is too mixed for a tighter canonical choice.", + "summary": "The items are a heterogeneous set of unrelated Transformers issues; no clear duplicate cluster emerges, and all soft candidate pairs should be rejected.", + "confidence": 0.24, + "canonical_issue_reason": "No real duplicate cluster is present; if a representative issue is needed, #44297 is the most concrete, narrowly scoped serialization bug among otherwise unrelated reports.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45278 is the best single representative of the dominant broad regression theme, but only as an umbrella issue rather than a true duplicate target for the whole cluster.", + "best_issue_reason": "#44297 is the best triage candidate because it has a specific, reproducible save_pretrained symptom and a clear code path, making it more actionable than the other heterogeneous issues.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:22355", - "right": "issue:28282", + "left": "issue:44295", + "right": "issue:45656", "accept": false, - "reason": "Different missing-module/dependency errors: ONNX import path vs PyTorch availability." + "reason": "Different bugs: buffer registration/access error versus optimizer stepping twice under deepspeed." }, { - "left": "issue:44206", - "right": "issue:45103", + "left": "issue:44448", + "right": "issue:46132", "accept": false, - "reason": "Unrelated bugs: feature-extractor arg handling vs auto-docstring annotation introspection." + "reason": "Both touch model behavior, but one is a Pegasus output change and the other is an AttentionInterface registration side effect." }, { - "left": "issue:45071", - "right": "issue:46082", + "left": "issue:44933", + "right": "issue:45712", "accept": false, - "reason": "Different config-validation problems affecting different models and parameters." + "reason": "Both are import/export related, but one is a missing image_utils import and the other is leftover dummy PT classes leaking into dir()." }, { - "left": "issue:41628", - "right": "issue:42757", + "left": "issue:44222", + "right": "issue:44297", "accept": false, - "reason": "Both are import errors, but for different symbols and different dependency surfaces." + "reason": "Both involve save_pretrained, but one is FP8 MoE model saving and the other is tokenizer metadata mismatch." }, { - "left": "issue:28282", - "right": "issue:30333", + "left": "issue:44297", + "right": "issue:45656", "accept": false, - "reason": "One is a missing PyTorch import, the other is MLFlow reporting after training failure." + "reason": "Tokenizer serialization bug is unrelated to a deepspeed optimizer double-step issue." }, { - "left": "issue:30333", - "right": "issue:35141", + "left": "issue:43116", + "right": "issue:44190", "accept": false, - "reason": "Unrelated training-job reporting bug vs embedding resize/post_init behavior." + "reason": "Example-script failures in different workflows: multi-label classification output versus local dataset loading." }, { - "left": "issue:44297", - "right": "issue:45310", + "left": "issue:43479", + "right": "issue:43644", "accept": false, - "reason": "Different tokenizer/config serialization failures; not the same concrete failure mode." + "reason": "Config default initialization bug is unrelated to non-persistent buffer corruption." }, { - "left": "issue:43723", - "right": "issue:44977", + "left": "issue:44190", + "right": "issue:46082", "accept": false, - "reason": "Tokenizer loading regression vs flash-attention generation behavior on a specific model." + "reason": "Local dataset loading issue is unrelated to LlamaConfig head_dim validation." }, { "left": "issue:44222", - "right": "issue:44484", + "right": "issue:44623", "accept": false, - "reason": "FP8 save-pretrained bug vs a question about shard sizing; no shared underlying defect." + "reason": "Different save-pretrained problems affecting different objects: FP8 MoE model versus processor artifact files." }, { - "left": "issue:43232", - "right": "issue:45834", + "left": "issue:30333", + "right": "issue:36331", "accept": false, - "reason": "Generation kwargs sync bug vs long-ocr index error in another model path." + "reason": "Both are training-related, but MLflow reporting on failure is separate from a compute_loss signature break." }, { - "left": "issue:43761", - "right": "issue:46082", + "left": "issue:42831", + "right": "issue:44623", "accept": false, - "reason": "Vision model hidden_states regression vs Llama config validation are unrelated." + "reason": "Accuracy regression in FineGrainedFP8 is unrelated to missing processor save files." }, { - "left": "issue:44263", - "right": "issue:45092", + "left": "issue:30333", + "right": "issue:34689", "accept": false, - "reason": "Different failures in different code paths: DSA indexer math vs meta-init checkpoint compatibility." + "reason": "Training-job reporting and a Llama 3.2 model-loading regression are different code paths." }, { - "left": "issue:44483", - "right": "issue:45406", + "left": "issue:45245", + "right": "issue:46132", "accept": false, - "reason": "Chat-completions request handling vs processor missing internal tokenizer attribute." + "reason": "Category-count runtime error is unrelated to AttentionInterface registration behavior." }, { - "left": "issue:44295", - "right": "issue:44623", + "left": "issue:45071", + "right": "issue:46082", "accept": false, - "reason": "Position_ids buffer access bug vs processor save_pretrained file emission bug." + "reason": "PretrainedConfig type-checking break is not the same as LlamaConfig head_dim rejection." }, { - "left": "issue:44373", - "right": "issue:44485", + "left": "issue:22355", + "right": "issue:28282", "accept": false, - "reason": "Docstring mismatch is not the same as a RoPE implementation bug." + "reason": "Both are import errors, but one is a missing transformers.onnx module and the other is missing PyTorch dependency detection." }, { - "left": "issue:44514", - "right": "issue:45103", + "left": "issue:41628", + "right": "issue:42757", "accept": false, - "reason": "Batched processor crash vs annotation-based auto-docstring crash are unrelated." + "reason": "Different missing symbols from different packages; not the same underlying failure." }, { - "left": "issue:43723", - "right": "issue:44871", + "left": "issue:44206", + "right": "issue:45103", "accept": false, - "reason": "Both involve tokenizer-related v5 issues, but different symptoms and model-specific roots." + "reason": "Feature-extractor argument crash and auto-docstring annotation parsing crash are unrelated." }, { - "left": "issue:44263", - "right": "issue:44360", + "left": "issue:28282", + "right": "issue:30333", "accept": false, - "reason": "Same subsystem name, but different defects in the GLM DSA indexer." + "reason": "ImportError for PyTorch dependency is unrelated to MLflow reporting on failed training jobs." }, { - "left": "issue:30990", - "right": "issue:38175", + "left": "issue:30333", + "right": "issue:35141", "accept": false, - "reason": "Loading hang vs zero-probability accuracy issue; no shared bug." + "reason": "Training reporting bug is unrelated to embedding reinitialization after resize_token_embeddings." }, { - "left": "issue:45092", - "right": "issue:45357", + "left": "issue:44297", + "right": "issue:45310", "accept": false, - "reason": "Checkpoint meta-init incompatibility vs incorrect visual-encoder save keys." + "reason": "Tokenizer_class mismatch in save_pretrained is not the same as Qwen3.5 MoE from_pretrained failure." }, { - "left": "issue:42907", - "right": "issue:46050", + "left": "issue:43723", + "right": "issue:44977", "accept": false, - "reason": "Both touch quantized save/load, but they are different quantization schemes and failure modes." + "reason": "Tokenizer loading regression and Qwen3.5 flash-attention generation bug are different problems." }, { - "left": "issue:43723", - "right": "issue:44991", + "left": "issue:44222", + "right": "issue:44484", "accept": false, - "reason": "Both are tokenizer-loading complaints, but they target different models and likely different causes." + "reason": "Both mention save_pretrained, but one is FP8 MoE saving and the other is shard-size default behavior." }, { - "left": "issue:44829", - "right": "issue:46002", + "left": "issue:43232", + "right": "issue:45834", "accept": false, - "reason": "Different training-path regressions: flash-attention behavior vs NaNs at init/checkout." + "reason": "Generation kwargs update after sync_gpus is unrelated to long OCR index errors." }, { - "left": "issue:43723", - "right": "issue:44442", + "left": "issue:43761", + "right": "issue:46082", "accept": false, - "reason": "Generic AutoTokenizer loading problem vs one tokenizer class loading failure; not clearly the same bug." + "reason": "CLIPVisionModel hidden_states regression is unrelated to LlamaConfig head_dim validation." }, { - "left": "issue:44261", + "left": "issue:44263", "right": "issue:45092", "accept": false, - "reason": "Layernorm epsilon config precision bug vs checkpoint/meta-init incompatibility." + "reason": "GlmMoeDsaIndexer torch.split issue is unrelated to remote-code checkpoint meta-init incompatibility." }, { - "left": "issue:42831", - "right": "issue:45310", + "left": "issue:44483", + "right": "issue:45406", "accept": false, - "reason": "FP8 accuracy issue vs Qwen3.5 Moe from_pretrained error are unrelated." + "reason": "API request rejection and Gemma4Processor crash are different service-level failures." }, { - "left": "issue:43479", - "right": "issue:44913", + "left": "issue:44295", + "right": "issue:44623", "accept": false, - "reason": "Different config persistence bugs affecting different models and fields." + "reason": "Position_ids buffer access error is unrelated to missing processor save artifacts." }, { - "left": "issue:41628", - "right": "issue:43824", + "left": "issue:44373", + "right": "issue:44485", "accept": false, - "reason": "Both are import errors, but for different exported classes and different underlying breakages." + "reason": "Wrong docstring for position_ids is not the same as the GLM-5 RoPE implementation discussion." }, { - "left": "issue:44263", - "right": "issue:44373", + "left": "issue:44514", + "right": "issue:45103", "accept": false, - "reason": "Different bug classes: runtime tensor indexing vs documentation text." + "reason": "Chat-template batching crash and auto-docstring annotation crash are unrelated." }, { - "left": "issue:42831", - "right": "issue:44038", + "left": "issue:43723", + "right": "issue:44871", "accept": false, - "reason": "Accuracy issue in FP8 path vs Qwen3-VL-Moe bug; no shared concrete code-path." + "reason": "Tokenizer loading regression is unrelated to eos_token_id configuration mismatch." }, { - "left": "issue:39401", - "right": "issue:44568", + "left": "issue:45092", + "right": "issue:45357", "accept": false, - "reason": "Different tokenizer regressions: offset mapping vs special-token insertion." + "reason": "Old InternVL2 meta-init incompatibility is unrelated to Qwen3.5 save_pretrained visual-key regression." }, { - "left": "issue:42757", - "right": "issue:45278", + "left": "issue:44263", + "right": "issue:44360", + "accept": false, + "reason": "Different GLM-5 DSA topics: a torch.split bug versus a missing ReLU discussion." + }, + { + "left": "issue:30990", + "right": "issue:38175", "accept": false, - "reason": "45278 is a broad umbrella import-regression report, but this is still not the same concrete bug as the specific huggingface_hub symbol import failure." + "reason": "Sentence Transformers loading hang and SigLIP2 zero-probability outputs are unrelated." + }, + { + "left": "issue:42907", + "right": "issue:46050", + "accept": false, + "reason": "Dequantized model save failure is unrelated to sinq quantization breakage after save/load." } ] }, @@ -12591,13 +12592,13 @@ "component_index": 1, "component_count": 17, "review_unit_index": 40, - "review_unit_count": 49, + "review_unit_count": 50, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-427", + "cluster_id": "cluster-10000-428", "nodes": [ "issue:22355", "issue:28218", @@ -12606,22 +12607,22 @@ "issue:39401", "issue:41628", "issue:42371", + "issue:42757", "issue:42831", "issue:42886", "issue:42907", "issue:42994", "issue:43066", - "issue:43116", "issue:43122", "issue:43232", - "issue:43408", - "issue:43475", + "issue:43479", "issue:43576", "issue:43646", + "issue:43723", "issue:43756", "issue:43784", - "issue:43937", - "issue:43976", + "issue:43824", + "issue:44038", "issue:44060", "issue:44117", "issue:44261", @@ -12629,31 +12630,42 @@ "issue:44295", "issue:44297", "issue:44373", - "issue:44488", - "issue:44514", + "issue:44442", "issue:44521", - "issue:44625", + "issue:44568", "issue:44671", "issue:44704", "issue:44743", "issue:44829", + "issue:44913", "issue:44945", - "issue:44964", - "issue:45200", + "issue:44991", + "issue:45092", "issue:45278", "issue:45290", + "issue:45310", "issue:45593", - "issue:45600", "issue:45753", "issue:46002", "issue:46132" ], "soft_pairs": [ + "issue:43723|issue:44991", + "issue:44829|issue:46002", + "issue:43723|issue:44442", + "issue:44261|issue:45092", + "issue:42831|issue:45310", + "issue:43479|issue:44913", + "issue:41628|issue:43824", + "issue:42831|issue:44038", + "issue:44263|issue:44373", + "issue:39401|issue:44568", + "issue:42757|issue:45278", "issue:43232|issue:45593", "issue:42886|issue:42994", "issue:42831|issue:45753", - "issue:22355|issue:28218", "issue:42886|issue:43066", + "issue:22355|issue:28218", "issue:43756|issue:44060", "issue:44671|issue:46132", "issue:43122|issue:44521", @@ -12665,222 +12677,211 @@ "issue:44295|issue:44297", "issue:44704|issue:45290", "issue:42907|issue:44945", - "issue:43232|issue:44743", "issue:41628|issue:45278", - "issue:43122|issue:44514", - "issue:43976|issue:44945", - "issue:44261|issue:44373", - "issue:43784|issue:45600", - "issue:39401|issue:44488", - "issue:43408|issue:43475", - "issue:42907|issue:43937", - "issue:39401|issue:44521", - "issue:43116|issue:45200", - "issue:44625|issue:44829", - "issue:42907|issue:44964" + "issue:43232|issue:44743" ], - "prepared_review_unit_hash": "642ea3512f475d34bd52c0c858a85e2b5eed69b34c2aad3068d7931253a639b7", + "prepared_review_unit_hash": "54b43756e2184666db45bf5017d7e95f09833010e9cdac1f4901bb6c68b5db90", "component_budget": { - "node_count": 427, - "item_count": 427, - "soft_pair_count": 1576, - "serialized_chars": 816182, - "estimated_input_tokens": 204046, - "estimated_eval_tokens": 408348 + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 29, - "serialized_chars": 35963, - "estimated_input_tokens": 8991, - "estimated_eval_tokens": 18238 + "serialized_chars": 35956, + "estimated_input_tokens": 8989, + "estimated_eval_tokens": 18234 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is a loose collection of unrelated Transformers issues spanning tokenization, multimodal processors, saving/loading, generation, and model-specific regressions. I do not see a single underlying bug/change that ties the soft pairs together, so all soft edges are rejected.", - "confidence": 0.95, - "canonical_issue_reason": null, + "summary": "This set is mostly a grab bag of unrelated transformer regressions. The only recurring theme is tokenizer/import breakage in v5, but the soft pairs still point to different concrete failures, so I would not merge any of them.", + "confidence": 0.9, + "canonical_issue_reason": "Issue 43723 is the broadest tokenizer-loading regression in v5 and is the closest hub for the tokenizer-related reports, but the overall set is too heterogeneous for a true duplicate canonical.", "canonical_pr_reason": null, - "best_issue_reason": "issue:42886 is the clearest standalone regression with a concrete repro (offline tokenizer cache loading), but it is only a loose representative because the cluster is heterogeneous rather than one duplicate family.", + "best_issue_reason": "43723 is the most representative issue for the tokenizer-loading regression subset because it targets the generic AutoTokenizer.from_pretrained failure in v5.", "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:43723", + "right": "issue:44991", + "accept": false, + "reason": "Both are tokenizer-loading complaints, but one is a generic v5 AutoTokenizer regression and the other is a model-specific failure for EMBEDDIA/est-roberta; not enough evidence of the same bug." + }, + { + "left": "issue:44829", + "right": "issue:46002", + "accept": false, + "reason": "Different problem domains: flash_attention_3 training degradation vs NaNs after checkout/init in classification heads." + }, + { + "left": "issue:43723", + "right": "issue:44442", + "accept": false, + "reason": "Shared surface area around AutoTokenizer, but one is a general v5 loading regression and the other is a specific FastSpeech2ConformerTokenizer export/loading failure." + }, + { + "left": "issue:44261", + "right": "issue:45092", + "accept": false, + "reason": "MLA layernorm precision/config issue is unrelated to InternVL2 remote-code meta-initialization incompatibility." + }, + { + "left": "issue:42831", + "right": "issue:45310", + "accept": false, + "reason": "FineGrainedFP8 accuracy regression and Qwen3.5-Moe from_pretrained error are distinct failures." + }, + { + "left": "issue:43479", + "right": "issue:44913", + "accept": false, + "reason": "Different configs and symptoms: Phi4Multimodal default-init bug vs GPTNeoX rotary_pct not persisting on reload." + }, + { + "left": "issue:41628", + "right": "issue:43824", + "accept": false, + "reason": "Both are import errors, but they concern different missing symbols and likely different registration/export paths." + }, + { + "left": "issue:42831", + "right": "issue:44038", + "accept": false, + "reason": "FineGrainedFP8 accuracy problems are not the same as Qwen3/VL-Moe load errors." + }, + { + "left": "issue:44263", + "right": "issue:44373", + "accept": false, + "reason": "The first is a code bug in torch.split handling; the second is just a wrong docstring." + }, + { + "left": "issue:39401", + "right": "issue:44568", + "accept": false, + "reason": "Qwen3 offset_mapping correctness and mdeberta add_special_tokens BOS/EOS omission are different tokenizer behaviors." + }, + { + "left": "issue:42757", + "right": "issue:45278", + "accept": false, + "reason": "Missing is_offline_mode from huggingface_hub is a specific dependency import issue; 45278 is a broad collection of import errors after upgrade." + }, { "left": "issue:43232", "right": "issue:45593", "accept": false, - "reason": "Different bugs: generation kwargs handling vs D-FINE auxiliary losses." + "reason": "Generation state handling after sync_gpus is unrelated to D-FINE auxiliary-loss behavior when denoising is off." }, { "left": "issue:42886", "right": "issue:42994", "accept": false, - "reason": "Both involve loading/saving, but one is offline tokenizer cache loading and the other is quantized model saving failure." + "reason": "Offline tokenizer cache loading and quantized model saving are separate code paths and symptoms." }, { "left": "issue:42831", "right": "issue:45753", "accept": false, - "reason": "Unrelated failures: FineGrainedFP8 accuracy vs Qwen3_5 image-triggered infinite loop." + "reason": "FP8 accuracy regression does not match a Qwen3_5 infinite-loop bug on a specific image." }, { - "left": "issue:22355", - "right": "issue:28218", + "left": "issue:42886", + "right": "issue:43066", "accept": false, - "reason": "Missing module import error vs tokenizer spacing behavior after added token; not the same bug." + "reason": "Offline cache loading failure and wrong tokenizer decoder type in v5 are different tokenizer subsystems." }, { - "left": "issue:42886", - "right": "issue:43066", + "left": "issue:22355", + "right": "issue:28218", "accept": false, - "reason": "Tokenizer offline cache regression and wrong tokenizer decoder type are different tokenizer issues." + "reason": "Missing transformers.onnx is an import/package issue; added-token spacing is a tokenizer behavior bug." }, { "left": "issue:43756", "right": "issue:44060", "accept": false, - "reason": "Different model-specific regressions: Smollm3 RoPE layer dropping vs Qwen3-Next tied-weights warning." + "reason": "Smollm3 RoPE-layer dropping and Qwen3-Next tied-weights warning are unrelated model-architecture issues." }, { "left": "issue:44671", "right": "issue:46132", "accept": false, - "reason": "Masked LM prediction bug and AttentionInterface registration behavior are unrelated." + "reason": "CamemBERT MLM prediction regression is not the same as AttentionInterface.register changing callback behavior." }, { "left": "issue:43122", "right": "issue:44521", "accept": false, - "reason": "Both touch tokenization/chat templates, but they affect different APIs and failure modes." + "reason": "Both involve tokenization, but one is a general tokenization drift across versions while the other is multimodal assistant_mask generation." }, { "left": "issue:44117", "right": "issue:44704", "accept": false, - "reason": "Tokenizer mapping None handling vs AutoProcessor kwargs forwarding are separate code paths." + "reason": "TOKENIZER_MAPPING_NAMES None handling and AutoProcessor kwargs propagation are separate loader plumbing bugs." }, { "left": "issue:43576", "right": "issue:43784", "accept": false, - "reason": "CLI/env command breakage vs sentence-transformers import NameError; not the same issue." + "reason": "Broken transformers env command and a sentence-transformers import NameError are unrelated failures." }, { "left": "issue:43646", "right": "issue:46002", "accept": false, - "reason": "Custom model init breakage and NaNs in classification heads are different regressions." + "reason": "Custom model initialization breakage is not the same as NaNs appearing in classification heads after checkout/init." }, { "left": "issue:30990", "right": "issue:33357", "accept": false, - "reason": "Sentence Transformers loading hang vs MacOS bus error on CLIP; unrelated symptoms." + "reason": "Sentence Transformers loading stalls and a MacOS bus error on CLIP are different runtime failures." }, { "left": "issue:42371", "right": "issue:44263", "accept": false, - "reason": "TF32 configuration guidance issue vs torch.split behavior in GlmMoeDsaIndexer." + "reason": "TF32 API guidance and a torch.split indexing bug do not describe the same issue." }, { "left": "issue:44295", "right": "issue:44297", "accept": false, - "reason": "Position_ids buffer access error vs tokenizer_class mismatch in saved config; different bugs." + "reason": "Registering position_ids as a buffer causing read errors is unrelated to tokenizer_class metadata mismatch on save." }, { "left": "issue:44704", "right": "issue:45290", "accept": false, - "reason": "AutoProcessor cached_file kwargs forwarding and chat-template tool-call crash are separate paths." + "reason": "Passing kwargs to cached_file and crashing on tool-call assistant messages are different apply_chat_template/processor bugs." }, { "left": "issue:42907", "right": "issue:44945", "accept": false, - "reason": "Quantized model saving failure vs incorrect pipeline-parallel LLM output; unrelated." - }, - { - "left": "issue:43232", - "right": "issue:44743", - "accept": false, - "reason": "Both involve generation/state, but one is sync_gpus kwargs update and the other is recurrent state reset in a specific model module." + "reason": "Saving dequantized models and incorrect output under pipeline parallelism are unrelated model-save vs inference-parallelism issues." }, { "left": "issue:41628", "right": "issue:45278", "accept": false, - "reason": "Single missing import vs broad import breakage after upgrade; too broad and not one concrete bug." - }, - { - "left": "issue:43122", - "right": "issue:44514", - "accept": false, - "reason": "Tokenization/version difference and Qwen2_5_VL batched chat-template crash are distinct issues." - }, - { - "left": "issue:43976", - "right": "issue:44945", - "accept": false, - "reason": "Python version support regression vs incorrect output under pipeline parallelism." - }, - { - "left": "issue:44261", - "right": "issue:44373", - "accept": false, - "reason": "Runtime precision bug vs docstring documentation issue." - }, - { - "left": "issue:43784", - "right": "issue:45600", - "accept": false, - "reason": "Import-time NameError in sentence-transformers integration vs stale auto_mappings references breaking CI." - }, - { - "left": "issue:39401", - "right": "issue:44488", - "accept": false, - "reason": "Different models and failures: wrong offset mappings vs model loading failure." - }, - { - "left": "issue:43408", - "right": "issue:43475", - "accept": false, - "reason": "Same SAM3 family, but a model-type warning and a missing attribute are different concrete bugs." - }, - { - "left": "issue:42907", - "right": "issue:43937", - "accept": false, - "reason": "Quantized saving failure vs invalid GenerationConfig; unrelated." - }, - { - "left": "issue:39401", - "right": "issue:44521", - "accept": false, - "reason": "Wrong offset mapping and zero assistant masks are different tokenizer/multimodal template bugs." - }, - { - "left": "issue:43116", - "right": "issue:45200", - "accept": false, - "reason": "Example script multi-label output bug vs Gemma 4 token-type defaults; unrelated." - }, - { - "left": "issue:44625", - "right": "issue:44829", - "accept": false, - "reason": "Config propagation bug vs flash_attention_3 training degeneration; different subsystems." + "reason": "AutoImageProcessor import failure is one symbol missing; 45278 reports many import errors after upgrade, so they are not the same concrete bug." }, { - "left": "issue:42907", - "right": "issue:44964", + "left": "issue:43232", + "right": "issue:44743", "accept": false, - "reason": "Saving dequantized models and loading Phi-4 multimodal are unrelated issues." + "reason": "Generation kwargs update after sync_gpus and Qwen3.5 recurrent-state reset are different generation-path bugs." } ] }, @@ -12896,64 +12897,74 @@ "component_index": 1, "component_count": 17, "review_unit_index": 41, - "review_unit_count": 49, + "review_unit_count": 50, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-427", + "cluster_id": "cluster-10000-428", "nodes": [ "issue:22355", "issue:28218", - "issue:29127", "issue:29159", "issue:30990", "issue:36296", - "issue:38617", + "issue:39401", "issue:42175", - "issue:42222", - "issue:42371", "issue:42673", "issue:42757", "issue:42907", - "issue:43066", "issue:43116", "issue:43122", "issue:43278", - "issue:43441", + "issue:43408", "issue:43452", + "issue:43475", "issue:43493", - "issue:43611", "issue:43784", "issue:43824", - "issue:43874", "issue:43937", "issue:43976", - "issue:43992", "issue:44038", - "issue:44075", "issue:44222", + "issue:44261", "issue:44295", "issue:44297", "issue:44351", "issue:44360", + "issue:44373", "issue:44485", - "issue:44704", + "issue:44488", + "issue:44514", + "issue:44521", + "issue:44625", "issue:44779", - "issue:44936", + "issue:44829", "issue:44938", "issue:44945", + "issue:44964", "issue:44977", "issue:45072", + "issue:45200", "issue:45216", - "issue:45278", "issue:45446", "issue:45478", - "issue:45753", - "issue:46002" + "issue:45600", + "issue:45753" ], "soft_pairs": [ + "issue:43122|issue:44514", + "issue:43976|issue:44945", + "issue:39401|issue:44488", + "issue:43784|issue:45600", + "issue:44261|issue:44373", + "issue:43408|issue:43475", + "issue:42907|issue:43937", + "issue:39401|issue:44521", + "issue:43116|issue:45200", + "issue:44625|issue:44829", + "issue:42907|issue:44964", "issue:44360|issue:44485", "issue:44038|issue:44297", "issue:28218|issue:30990", @@ -12971,227 +12982,219 @@ "issue:43122|issue:44977", "issue:42175|issue:44938", "issue:43493|issue:45753", - "issue:42673|issue:45072", - "issue:43066|issue:44779", - "issue:43992|issue:44704", - "issue:38617|issue:45278", - "issue:22355|issue:29127", - "issue:43611|issue:46002", - "issue:43937|issue:44485", - "issue:43976|issue:45446", - "issue:43116|issue:43493", - "issue:42222|issue:43441", - "issue:42371|issue:43874", - "issue:44075|issue:44936" + "issue:42673|issue:45072" ], - "prepared_review_unit_hash": "c46592cb210fb78e325d20f8b96e3ffc7dd7325fcebbdff5e5a6bde5985495d5", + "prepared_review_unit_hash": "992c2b8cd9aa81120e9c54b8b8ec04b9391b15c5d66fa0c264532c80ba2bd6ea", "component_budget": { - "node_count": 427, - "item_count": 427, - "soft_pair_count": 1576, - "serialized_chars": 816182, - "estimated_input_tokens": 204046, - "estimated_eval_tokens": 408348 + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 }, "budget": { - "node_count": 48, - "item_count": 48, + "node_count": 47, + "item_count": 47, "soft_pair_count": 29, - "serialized_chars": 35760, - "estimated_input_tokens": 8940, - "estimated_eval_tokens": 18136 + "serialized_chars": 35357, + "estimated_input_tokens": 8840, + "estimated_eval_tokens": 17936 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The cluster is mostly a loose collection of unrelated Transformer issues; the soft-similarity pairs mostly share broad subsystem words (tokenizer/import/model) rather than the same concrete bug. No PR candidates are present.", - "confidence": 0.34, - "canonical_issue_reason": "issue:43122 is the best anchor for the tokenizer-regression subgroup because it describes a broad cross-version tokenization change and sits closest to the other tokenizer-related candidates.", + "summary": "Mostly unrelated issues across tokenization, model loading, and backend regressions. The Qwen3.5 save/load reports are thematically close, but #45216 and #45478 are not strong enough evidence of the same underlying bug to merge as duplicates.", + "confidence": 0.63, + "canonical_issue_reason": "Issue #45216 is the broadest and clearest representative of the Qwen3.5 5.4.0 save_pretrained serialization regression, but it should not absorb the separate from_pretrained error in #45478 without stronger body-level evidence of a shared root cause.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43122 is the most representative issue overall: it is specific enough to define a real bug, but broad enough to capture the shared tokenizer-regression theme better than the more model-specific reports.", + "best_issue_reason": "#45216 is the best anchor for the Qwen3.5 save/load regression theme because it describes the core checkpoint corruption symptom most generally.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:44360", - "right": "issue:44485", + "left": "issue:43122", + "right": "issue:44514", "accept": false, - "reason": "Different subsystems and failure modes: DSA/ReLU discussion vs GLM-5 RoPE implementation." + "reason": "Both involve tokenizer/chat-template behavior, but one is a tokenizer regression and the other is a batched apply_chat_template crash; different symptoms and code paths." }, { - "left": "issue:44038", - "right": "issue:44297", + "left": "issue:43976", + "right": "issue:44945", "accept": false, - "reason": "One is a Qwen3-VL-Moe loading bug; the other is tokenizer save_pretrained metadata mismatch." + "reason": "A Python version compatibility issue and an incorrect output under pipeline parallelism are unrelated bugs." }, { - "left": "issue:28218", - "right": "issue:30990", + "left": "issue:39401", + "right": "issue:44488", "accept": false, - "reason": "Tokenizer extra-space bug is unrelated to Sentence Transformers loading hang." + "reason": "Wrong offset_mapping for Qwen3 tokenizer and failure to load a specific model are different problems; only the model family overlaps." }, { - "left": "issue:44038", - "right": "issue:44295", + "left": "issue:43784", + "right": "issue:45600", "accept": false, - "reason": "Qwen3-VL-Moe loading issue does not match a position_ids buffer read error." + "reason": "Both are import/load failures, but one is a missing nn symbol in sentence-transformers integration and the other is stale auto_mappings references." }, { - "left": "issue:43122", - "right": "issue:44779", + "left": "issue:44261", + "right": "issue:44373", "accept": false, - "reason": "Both concern tokenizer regressions, but they describe different concrete failures and model-specific behaviors." + "reason": "A numerical precision/config issue in MLA and a wrong docstring for position_ids are unrelated." }, { - "left": "issue:42907", - "right": "issue:43452", + "left": "issue:43408", + "right": "issue:43475", "accept": false, - "reason": "Dequantized save failure is unrelated to gguf_file loading behavior in AutoTokenizer/AutoModel." + "reason": "Both are SAM 3 related, but one is a model-type warning and the other is a missing attribute on a vision encoder output." }, { - "left": "issue:36296", - "right": "issue:43116", + "left": "issue:42907", + "right": "issue:43937", "accept": false, - "reason": "Tensor parallel training bug is unrelated to example-script multi-label classification output." + "reason": "Saving dequantized models and invalid GenerationConfig are distinct code paths and symptoms." }, { - "left": "issue:44038", - "right": "issue:44222", + "left": "issue:39401", + "right": "issue:44521", "accept": false, - "reason": "Different Qwen3-VL-Moe loading path vs FP8 save_pretrained behavior." + "reason": "Both touch tokenizer/multimodal processing, but wrong offset mapping and all-zero assistant masks are separate bugs." }, { - "left": "issue:43824", - "right": "issue:44351", + "left": "issue:43116", + "right": "issue:45200", "accept": false, - "reason": "Separate import errors for different symbols and modules." + "reason": "A classification example returning empty results and a Gemma 4 token-type-id requirement are not the same underlying issue." }, { - "left": "issue:45216", - "right": "issue:45478", + "left": "issue:44625", + "right": "issue:44829", "accept": false, - "reason": "Both are Qwen3.5-related, but one is save_pretrained checkpoint corruption and the other is a from_pretrained error." + "reason": "num_labels propagation and flash_attention_3 degenerate training are unrelated model configuration/performance bugs." }, { - "left": "issue:42757", - "right": "issue:44351", + "left": "issue:42907", + "right": "issue:44964", "accept": false, - "reason": "Missing huggingface_hub symbol vs missing transformers symbol; unrelated import breakages." + "reason": "Saving dequantized models and loading a multimodal Phi-4 model are different problems." }, { - "left": "issue:43784", - "right": "issue:45446", + "left": "issue:44360", + "right": "issue:44485", "accept": false, - "reason": "Sentence-transformers import failure is unrelated to the flex_attention PyTorch version check." + "reason": "DSA indexer activation logic and GLM-5 RoPE implementation are unrelated." }, { - "left": "issue:43278", - "right": "issue:44945", + "left": "issue:44038", + "right": "issue:44297", "accept": false, - "reason": "Dtype drift in evaluate is not the same bug as incorrect output under pipeline parallelism." + "reason": "A Qwen3-VL-Moe loading issue and tokenizer_class mismatch on save_pretrained are different regressions." }, { - "left": "issue:22355", - "right": "issue:29127", + "left": "issue:28218", + "right": "issue:30990", "accept": false, - "reason": "Old missing-module issue is unrelated to LayoutLMv3 error-message clarity." + "reason": "Tokenizer adding an extra space and Sentence Transformers getting stuck loading do not describe the same bug." }, { - "left": "issue:43122", - "right": "issue:44977", + "left": "issue:44038", + "right": "issue:44295", "accept": false, - "reason": "Both are tokenizer regressions in v5, but they affect different models and show different symptoms." + "reason": "Qwen3-VL-Moe loading and reading position_ids after buffer registration are unrelated." }, { - "left": "issue:42175", - "right": "issue:44938", + "left": "issue:43122", + "right": "issue:44779", "accept": false, - "reason": "Packaging backend selection issue is unrelated to Python 3.14 import/load failure." + "reason": "Both are tokenizer regressions, but one is version-difference behavior and the other is Deepseek incorrect results; different models and failure modes." }, { - "left": "issue:43493", - "right": "issue:45753", + "left": "issue:42907", + "right": "issue:43452", "accept": false, - "reason": "SigLIP2 JAX/HF discrepancy is unrelated to Qwen3_5 infinite-loop behavior on one image." + "reason": "Saving dequantized models and gguf_file breaking AutoTokenizer/AutoModel loading are distinct issues." }, { - "left": "issue:42673", - "right": "issue:45072", + "left": "issue:36296", + "right": "issue:43116", "accept": false, - "reason": "VRAM leak in Qwen3ForCausalLM is unrelated to dtype mismatches in inference." + "reason": "Tensor parallel training and a multi-label classification example returning empty results are unrelated." }, { - "left": "issue:43066", - "right": "issue:44779", + "left": "issue:44038", + "right": "issue:44222", "accept": false, - "reason": "Tokenizer decoder-type bug and Deepseek tokenization regression are related only at a very high level, not the same concrete defect." + "reason": "Qwen3-VL-Moe loading and FP8 save_pretrained for moe are both MoE-related but not the same code-path failure." }, { - "left": "issue:43992", - "right": "issue:44704", + "left": "issue:43824", + "right": "issue:44351", "accept": false, - "reason": "Missing embed_tokens weight on from_pretrained is unrelated to AutoProcessor kwarg forwarding." + "reason": "Both are import errors, but they involve different missing exports and different fixes." }, { - "left": "issue:38617", - "right": "issue:45278", + "left": "issue:45216", + "right": "issue:45478", "accept": false, - "reason": "Specific missing import is not the same underlying issue as a broad set of import errors after upgrade." + "reason": "The titles point to a related Qwen3.5 save/load theme, but one is a save_pretrained checkpoint corruption report and the other is a from_pretrained load error for Qwen3.5 MoE; that's not enough to prove the same bug." }, { - "left": "issue:22355", - "right": "issue:29127", + "left": "issue:42757", + "right": "issue:44351", "accept": false, - "reason": "No shared code path: ONNX import failure vs LayoutLMv3 error-message wording." + "reason": "Missing is_offline_mode from huggingface_hub and missing HybridCache from transformers are separate import/export issues." }, { - "left": "issue:43611", - "right": "issue:46002", + "left": "issue:43784", + "right": "issue:45446", "accept": false, - "reason": "base_model_prefix loading breakage is unrelated to NaNs in classification heads." + "reason": "Importing sentence-transformers with latest transformers and a PyTorch version check for AuxRequest are unrelated." }, { - "left": "issue:43937", - "right": "issue:44485", + "left": "issue:43278", + "right": "issue:44945", "accept": false, - "reason": "GenerationConfig validation for GLM-5 is not the same as a RoPE implementation discussion." + "reason": "Embedding dtype changing between training and evaluate is not the same as incorrect output under pipeline parallelism." }, { - "left": "issue:43976", - "right": "issue:45446", + "left": "issue:22355", + "right": "issue:29159", "accept": false, - "reason": "Python version support breakage is unrelated to an incorrect PyTorch-version guard in flex_attention." + "reason": "No module named transformers.onnx and slow/fast tokenizer inconsistency are different historical bugs." }, { - "left": "issue:43116", - "right": "issue:43493", + "left": "issue:43122", + "right": "issue:44977", "accept": false, - "reason": "Example-script multi-label output bug is unrelated to SigLIP2 implementation discrepancies." + "reason": "Tokenizer version-difference behavior and Qwen3.5 flash-attention generation problems are unrelated." }, { - "left": "issue:42222", - "right": "issue:43441", + "left": "issue:42175", + "right": "issue:44938", "accept": false, - "reason": "VitPose breakage and Ministral-3 FlashAttention failure affect different models and code paths." + "reason": "Missing TensorFlow backend in an install extra and Python 3.14 loading failure are different packaging/runtime issues." }, { - "left": "issue:42371", - "right": "issue:43874", + "left": "issue:43493", + "right": "issue:45753", "accept": false, - "reason": "TF32 API guidance is unrelated to missing image-patch counting in GLM46VImageProcessorFast." + "reason": "SigLIP2 HF/JAX discrepancy and Qwen3_5 infinite loop on a specific image are separate model bugs." }, { - "left": "issue:44075", - "right": "issue:44936", + "left": "issue:42673", + "right": "issue:45072", "accept": false, - "reason": "Unused SGD args and trainer.evaluate() failure after train() are different bugs." + "reason": "VRAM leak in Qwen3ForCausalLM multithreaded loading and dtype mismatches in bfloat16 inference are unrelated." } ] }, - "evaluator_result": null, - "evaluator_used": false, - "retried": false, + "evaluator_result": { + "accept": false, + "feedback": "The overall summary is fine, but the accepted edge between issue:45216 and issue:45478 is too speculative for the packet. The titles only show a Qwen3.5 / 5.4.0 save_pretrained checkpoint corruption report and a separate from_pretrained load error for Qwen3.5 MoE; that is a plausible theme match, but not enough evidence to call them the same bugfix. A more conservative verdict would keep this pair as non-duplicate unless the bodies show a shared serialization root cause." + }, + "evaluator_used": true, + "retried": true, "accepted_nontrivial_soft_edge": false, "error_kind": null, "error_message": null @@ -13201,63 +13204,74 @@ "component_index": 1, "component_count": 17, "review_unit_index": 42, - "review_unit_count": 49, + "review_unit_count": 50, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-427", + "cluster_id": "cluster-10000-428", "nodes": [ + "issue:22355", + "issue:29127", "issue:33453", "issue:38617", "issue:39401", - "issue:41628", "issue:42222", - "issue:42490", - "issue:42617", + "issue:42371", "issue:42673", "issue:43064", + "issue:43066", "issue:43116", - "issue:43408", - "issue:43421", + "issue:43441", "issue:43450", - "issue:43582", + "issue:43493", + "issue:43611", "issue:43673", - "issue:43742", "issue:43825", "issue:43827", + "issue:43874", "issue:43906", "issue:43937", + "issue:43976", "issue:43986", + "issue:43992", "issue:44038", + "issue:44075", "issue:44117", "issue:44291", "issue:44380", - "issue:44410", "issue:44466", - "issue:44484", + "issue:44485", + "issue:44704", "issue:44779", - "issue:44829", + "issue:44936", "issue:44977", "issue:45071", "issue:45081", "issue:45083", - "issue:45216", - "issue:45310", - "issue:45341", + "issue:45278", "issue:45357", "issue:45405", "issue:45446", "issue:45478", - "issue:45561", "issue:45563", "issue:45588", "issue:46002", - "issue:46077", - "issue:46178" + "issue:46077" ], "soft_pairs": [ + "issue:43992|issue:44704", + "issue:43066|issue:44779", + "issue:38617|issue:45278", + "issue:22355|issue:29127", + "issue:43611|issue:46002", + "issue:43937|issue:44485", + "issue:43976|issue:45446", + "issue:42222|issue:43441", + "issue:43116|issue:43493", + "issue:42371|issue:43874", + "issue:44075|issue:44936", "issue:43064|issue:43673", "issue:45357|issue:45478", "issue:44038|issue:45478", @@ -13275,197 +13289,210 @@ "issue:43450|issue:43986", "issue:44466|issue:46002", "issue:45405|issue:45446", - "issue:42222|issue:46178", - "issue:44117|issue:45081", - "issue:45341|issue:45561", - "issue:38617|issue:41628", - "issue:44410|issue:45310", - "issue:44829|issue:45216", - "issue:39401|issue:43421", - "issue:42617|issue:45341", - "issue:42673|issue:43582", - "issue:44484|issue:45310", - "issue:42490|issue:42673", - "issue:43408|issue:43742" + "issue:44117|issue:45081" ], - "prepared_review_unit_hash": "f1fbe9b36c7339e2bc304e6691bacdc89cfa1748b238d576269738a9b92b9767", + "prepared_review_unit_hash": "9262a5943d5f2df4cf9b9ca845cc446d74839acbbc5c6c595f35b0e2af74a07b", "component_budget": { - "node_count": 427, - "item_count": 427, - "soft_pair_count": 1576, - "serialized_chars": 816182, - "estimated_input_tokens": 204046, - "estimated_eval_tokens": 408348 + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 }, "budget": { "node_count": 47, "item_count": 47, "soft_pair_count": 29, - "serialized_chars": 35838, - "estimated_input_tokens": 8960, - "estimated_eval_tokens": 18176 + "serialized_chars": 35667, + "estimated_input_tokens": 8917, + "estimated_eval_tokens": 18090 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly heterogeneous transformer-regression issues with no true duplicate soft pairs. The only clear duplicate-like theme is the repeated Qwen3.5 Moe from_pretrained bug, where 45310/45478 look like the same report family.", - "confidence": 0.72, - "canonical_issue_reason": "Issue 45310 is the earlier report for the repeated Qwen3.5 Moe from_pretrained failure; it\u2019s the better canonical anchor for that duplicate pair.", + "summary": "These issues are largely unrelated regressions across different models, APIs, and failure modes. None of the soft pairs look like the same underlying bug or change, so I would not merge any of them as duplicates.", + "confidence": 0.93, + "canonical_issue_reason": "issue:45278 is the broadest representative item in the set: it describes multiple import errors after upgrading to 5.5.0, which makes it the most central umbrella issue among these otherwise distinct reports.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 45478 is the cleanest overall representative because it has the same concrete bug as 45310 but a slightly stronger report signal and clearer version-specific wording.", + "best_issue_reason": "issue:45278 is the best single issue to anchor the cluster because it captures a broad upgrade regression and is more representative than the model-specific or narrowly scoped reports.", "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:43992", + "right": "issue:44704", + "accept": false, + "reason": "Different code paths: UMT5Encoder weight loading vs AutoProcessor forwarding kwargs to cached_file." + }, + { + "left": "issue:43066", + "right": "issue:44779", + "accept": false, + "reason": "Both tokenizer-related, but one is wrong decoder type and the other is incorrect Deepseek tokenization results; not the same bug." + }, + { + "left": "issue:38617", + "right": "issue:45278", + "accept": false, + "reason": "One is a specific missing import, the other is a broad set of import errors after upgrade; related theme but not the same underlying defect." + }, + { + "left": "issue:22355", + "right": "issue:29127", + "accept": false, + "reason": "Completely different problems: missing transformers.onnx module vs a confusing layoutlmv3 error message." + }, + { + "left": "issue:43611", + "right": "issue:46002", + "accept": false, + "reason": "Loading with base_model_prefix broke, but the other report is NaNs in classification heads after checkout/init; unrelated failure modes." + }, + { + "left": "issue:43937", + "right": "issue:44485", + "accept": false, + "reason": "Both mention GLM-5, but one is invalid GenerationConfig and the other is RoPE implementation discussion; different bugs." + }, + { + "left": "issue:43976", + "right": "issue:45446", + "accept": false, + "reason": "Python version compatibility regression vs an incorrect PyTorch version check in flex_attention; unrelated." + }, + { + "left": "issue:42222", + "right": "issue:43441", + "accept": false, + "reason": "VitPose model breakage and Ministral FlashAttention failure are different model-specific issues." + }, + { + "left": "issue:43116", + "right": "issue:43493", + "accept": false, + "reason": "Example-script multi-label classification behavior is unrelated to SigLIP2 implementation discrepancy." + }, + { + "left": "issue:42371", + "right": "issue:43874", + "accept": false, + "reason": "TF32 settings and a missing image-processor method are unrelated." + }, + { + "left": "issue:44075", + "right": "issue:44936", + "accept": false, + "reason": "Optimizer argument handling and trainer.evaluate() after train are different trainer/optimizer issues." + }, { "left": "issue:43064", "right": "issue:43673", "accept": false, - "reason": "Different bugs: optimizer-state placement under FSDP2/PEFT vs GenerationMixin cache handling during chunked prefill." + "reason": "Distributed optimizer-state corruption and missing GenerationMixin cache are unrelated code paths." }, { "left": "issue:45357", "right": "issue:45478", "accept": false, - "reason": "Same model family, but different failures: incorrect saved visual encoder keys vs from_pretrained load error." + "reason": "Both concern Qwen3.5, but save_pretrained key serialization and from_pretrained loading errors are different bugs." }, { "left": "issue:44038", "right": "issue:45478", "accept": false, - "reason": "Both mention Qwen3/VL/Moe, but the concrete breakage differs and the model/load paths are not the same." + "reason": "Same general model family, but Qwen3-VL-Moe vs Qwen3.5 MoE from_pretrained are not the same failure." }, { "left": "issue:43116", "right": "issue:43825", "accept": false, - "reason": "Example-script multi-label output bug vs a misleading pipeline error message; different code paths and symptoms." + "reason": "Classification example output bug vs pipeline translation-task messaging are unrelated." }, { "left": "issue:45083", "right": "issue:45446", "accept": false, - "reason": "Unrelated: qwen3_omni_moe feature-extraction helper behavior vs a flex_attention import version check." + "reason": "A feature-extraction helper issue and a PyTorch version check bug are unrelated." }, { "left": "issue:43937", "right": "issue:45563", "accept": false, - "reason": "GenerationConfig validation failure is unrelated to the stale warning for num_return_sequences." + "reason": "GenerationConfig invalidity and a stale warning for num_return_sequences are different issues." }, { "left": "issue:44380", "right": "issue:46077", "accept": false, - "reason": "Both involve attention behavior, but GPT2 scaling ignored and Gemma4 bidirectional masking are distinct model/backend bugs." + "reason": "Attention-scaling ignored in GPT2 backend vs Gemma4 bidirectional mask construction are different model/backend bugs." }, { "left": "issue:33453", "right": "issue:39401", "accept": false, - "reason": "Tokenizer regressions, but one is loading regression and the other is wrong offset_mapping; not the same underlying defect." + "reason": "Both tokenizer-related regressions, but generic loading regression vs Qwen3 offset_mapping bug are not the same defect." }, { "left": "issue:44117", "right": "issue:45588", "accept": false, - "reason": "Tokenizer mapping fallback issue is unrelated to flash_attention s_aux=None crash." + "reason": "Tokenizer mapping None handling and flash_attention crashing on s_aux=None are unrelated." }, { "left": "issue:43827", "right": "issue:45446", "accept": false, - "reason": "Docs still referencing pipeline() is unrelated to the AuxRequest PyTorch version check." + "reason": "Documentation references to pipeline() vs a runtime version check bug have no shared underlying code path." }, { "left": "issue:39401", "right": "issue:44779", "accept": false, - "reason": "Different tokenizer bugs on different models; both are regressions, but not one shared root cause." + "reason": "Different model-specific tokenizer regressions: Qwen3 offset mapping vs Deepseek incorrect results." }, { "left": "issue:42673", "right": "issue:43906", "accept": false, - "reason": "VRAM leak in Qwen3ForCausalLM threads is unrelated to an isolated reproduction of another issue." + "reason": "VRAM leak in Qwen3ForCausalLM threads is not the same as a reproduction of issue 38071." }, { "left": "issue:43937", "right": "issue:44977", "accept": false, - "reason": "GenerationConfig invalid vs flash-attention generation failure are separate problems." + "reason": "Different GLM/Qwen generation failures with different symptoms and likely different causes." }, { "left": "issue:44291", "right": "issue:45071", "accept": false, - "reason": "Both are v5 compatibility issues, but init_empty_weights argument handling and PretrainedConfig type checking are different code-path failures." + "reason": "init_empty_weights argument incompatibility is unrelated to PretrainedConfig type-checking breakage." }, { "left": "issue:43450", "right": "issue:43986", "accept": false, - "reason": "Video batch-shape bug and AutoProcessor without torchvision crash are different issues in the video stack." + "reason": "Batched video processor shape bug vs missing torchvision crash when loading via AutoProcessor are distinct issues." }, { "left": "issue:44466", "right": "issue:46002", "accept": false, - "reason": "Serialization of tied lm_head weights is unrelated to NaNs in classification heads during checkout/init." + "reason": "Serialization of tied lm_head weights and NaNs in classification heads are unrelated." }, { "left": "issue:45405", "right": "issue:45446", "accept": false, - "reason": "PEFT version pinning and a PyTorch version check for flex_attention are unrelated maintenance/runtime issues." - }, - { - "left": "issue:42222", - "right": "issue:46178", - "accept": false, - "reason": "VitPose model breakage and tensorflow_text backend mapping errors are unrelated." + "reason": "A PEFT minimum-version bump issue is unrelated to the flex_attention import version check." }, { "left": "issue:44117", "right": "issue:45081", "accept": false, - "reason": "Tokenizer mapping None and Mistral regex patch crashing on backend_tokenizer are different failures." - }, - { - "left": "issue:45341", - "right": "issue:45561", - "accept": false, - "reason": "A small testing_utils bug and a pytest-xdist race in captured_info.txt are not the same underlying problem." - }, - { - "left": "issue:38617", - "right": "issue:41628", - "accept": false, - "reason": "Different import errors for different symbols; they do not share the same broken code path." - }, - { - "left": "issue:44410", - "right": "issue:45310", - "accept": false, - "reason": "Missing qwen3next projections and Qwen3.5 Moe from_pretrained loading error are distinct model-specific issues." - }, - { - "left": "issue:44829", - "right": "issue:45216", - "accept": false, - "reason": "FlashAttention training degeneration and incorrect save_pretrained checkpointing are separate bugs." - }, - { - "left": "issue:39401", - "right": "issue:43421", - "accept": false, - "reason": "Wrong offset_mapping and runtime post-processor updates after special-token changes are different tokenizer backend defects." - }, - { - "left": "issue:42617", - "right": "issue:45341", - "accept": false, - "reason": "3d_parallel.py execution failure is unrelated to a testing_utils bug." + "reason": "Tokenizer mapping fallback handling and a Mistral regex patch crash are different tokenizer bugs." } ] }, @@ -13481,42 +13508,42 @@ "component_index": 1, "component_count": 17, "review_unit_index": 43, - "review_unit_count": 49, + "review_unit_count": 50, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-427", + "cluster_id": "cluster-10000-428", "nodes": [ "issue:38617", + "issue:39401", "issue:41628", "issue:42175", "issue:42222", "issue:42371", + "issue:42490", + "issue:42617", + "issue:42673", "issue:42831", "issue:43012", "issue:43066", "issue:43097", "issue:43122", - "issue:43335", "issue:43408", + "issue:43421", "issue:43441", - "issue:43452", "issue:43577", + "issue:43582", "issue:43673", + "issue:43742", "issue:43874", - "issue:43931", "issue:43937", - "issue:43986", - "issue:44038", "issue:44075", - "issue:44117", "issue:44220", "issue:44261", "issue:44263", "issue:44360", - "issue:44368", "issue:44410", "issue:44483", "issue:44484", @@ -13525,28 +13552,39 @@ "issue:44496", "issue:44829", "issue:44933", - "issue:44987", - "issue:45003", "issue:45071", + "issue:45216", "issue:45278", + "issue:45310", "issue:45341", - "issue:45372", "issue:45464", "issue:45478", "issue:45542", + "issue:45561", "issue:45656", - "issue:46123", - "issue:46172" + "issue:46172", + "issue:46178" ], "soft_pairs": [ + "issue:45341|issue:45561", + "issue:42222|issue:46178", + "issue:38617|issue:41628", + "issue:44410|issue:45310", + "issue:39401|issue:43421", + "issue:44829|issue:45216", + "issue:42617|issue:45341", + "issue:44484|issue:45310", + "issue:42673|issue:43582", + "issue:42490|issue:42673", + "issue:43408|issue:43742", "issue:42371|issue:44261", "issue:44933|issue:45278", "issue:44483|issue:45464", - "issue:42222|issue:45542", - "issue:43937|issue:44496", "issue:38617|issue:44220", - "issue:42175|issue:42222", + "issue:43937|issue:44496", + "issue:42222|issue:45542", "issue:43066|issue:43122", + "issue:42175|issue:42222", "issue:42371|issue:42831", "issue:44075|issue:45656", "issue:43673|issue:45341", @@ -13561,258 +13599,240 @@ "issue:44263|issue:46172", "issue:44360|issue:46172", "issue:43012|issue:43408", - "issue:43441|issue:44829", - "issue:43931|issue:44410", - "issue:43931|issue:44038", - "issue:43335|issue:45071", - "issue:43673|issue:46123", - "issue:43097|issue:43452", - "issue:45003|issue:45341", - "issue:43986|issue:45341", - "issue:44117|issue:44987", - "issue:43408|issue:44368", - "issue:43986|issue:44484", - "issue:43874|issue:44263", - "issue:43986|issue:45372" + "issue:43441|issue:44829" ], - "prepared_review_unit_hash": "3d55347442ba5d03e117ca1fcccc221bf191e5818452a50809e4dcebb04e5e10", + "prepared_review_unit_hash": "573ac12730a34a75c005a551453139f4b133089737cba965ec9930cc7ce31a15", "component_budget": { - "node_count": 427, - "item_count": 427, - "soft_pair_count": 1576, - "serialized_chars": 816182, - "estimated_input_tokens": 204046, - "estimated_eval_tokens": 408348 + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 35, - "serialized_chars": 37926, - "estimated_input_tokens": 9482, - "estimated_eval_tokens": 19220 + "soft_pair_count": 34, + "serialized_chars": 37402, + "estimated_input_tokens": 9351, + "estimated_eval_tokens": 18958 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly a loose collection of unrelated Transformers regression issues, with only broad thematic overlap around import/loading and model-configuration breakages. I did not find any soft pair that is clearly the same underlying bug or change.", - "confidence": 0.83, - "canonical_issue_reason": "issue:45278 is the broadest umbrella item in the set: it describes multiple import failures after upgrading from 4.57.0 to 5.5.0, making it the closest representative of the cluster theme.", + "summary": "This cluster is heterogeneous: most items are distinct import, tokenizer, model-loading, backend, or training bugs rather than true duplicates. None of the soft pairs look like the same concrete issue, so all are rejected.", + "confidence": 0.8, + "canonical_issue_reason": "issue:45278 is the broadest umbrella in the set ('Many import errors after update from 4.57.0 to 5.5.0') and is the closest representative of the import-breakage subset, though the cluster overall is not a single duplicate group.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45278 is the best single representative because it is the most general user-facing regression report; the rest are narrower, model-specific, or unrelated bugs.", + "best_issue_reason": "issue:45278 is the most representative high-level regression report and the best anchor if one issue must be chosen from this mixed set.", "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:45341", + "right": "issue:45561", + "accept": false, + "reason": "Both mention testing_utils, but one is a vague bug report and the other is a specific xdist race on captured_info.txt; different failures." + }, + { + "left": "issue:42222", + "right": "issue:46178", + "accept": false, + "reason": "Different subsystems: vitpose model breakage vs backend-mapping error for tensorflow_text." + }, + { + "left": "issue:38617", + "right": "issue:41628", + "accept": false, + "reason": "Both are import errors, but for different missing symbols and code paths." + }, + { + "left": "issue:44410", + "right": "issue:45310", + "accept": false, + "reason": "Different Qwen issues: missing layer projections vs a Qwen3.5 MoE from_pretrained error." + }, + { + "left": "issue:39401", + "right": "issue:43421", + "accept": false, + "reason": "Both are tokenizer-related, but wrong offset_mapping and runtime post-processor updates are different problems." + }, + { + "left": "issue:44829", + "right": "issue:45216", + "accept": false, + "reason": "FlashAttention-3 degenerate training and Qwen3.5 save_pretrained checkpoint corruption are unrelated regressions." + }, + { + "left": "issue:42617", + "right": "issue:45341", + "accept": false, + "reason": "A script failure and a testing_utils bug are unrelated." + }, + { + "left": "issue:44484", + "right": "issue:45310", + "accept": false, + "reason": "Shard-size behavior in save_pretrained is unrelated to the Qwen3.5 from_pretrained regression." + }, + { + "left": "issue:42673", + "right": "issue:43582", + "accept": false, + "reason": "VRAM leakage in multi-threaded loading and an Apple Silicon TypeError are distinct issues." + }, + { + "left": "issue:42490", + "right": "issue:42673", + "accept": false, + "reason": "Trailing-slash path warning vs VRAM leak under multiple dataloader threads are different bugs." + }, + { + "left": "issue:43408", + "right": "issue:43742", + "accept": false, + "reason": "Model-type mismatch warning for sam3_tracker is unrelated to the MobileLLM key error." + }, { "left": "issue:42371", "right": "issue:44261", "accept": false, - "reason": "Both mention precision-related behavior, but one is about TF32 settings and the other about a missing rms_norm_eps in MLA; different code paths and different failures." + "reason": "TF32 settings warning and MLA rms_norm_eps precision error are different problems." }, { "left": "issue:44933", "right": "issue:45278", "accept": false, - "reason": "Both involve imports, but one is a specific nonexistent image_utils import while the other is a broad post-upgrade import regression; not the same bug." + "reason": "Both are import-related, but missing image_utils import and broad 5.5.0 import errors are not the same concrete failure." }, { "left": "issue:44483", "right": "issue:45464", "accept": false, - "reason": "Both mention chat/completions, but one is a general API request rejection and the other is a Qwen3.5 streaming inference failure; too different to merge." + "reason": "Different API failures: generic chat/completions rejection vs Qwen3.5 streaming inference failure." }, { - "left": "issue:42222", - "right": "issue:45542", + "left": "issue:38617", + "right": "issue:44220", "accept": false, - "reason": "These are unrelated: vitpose model breakage vs TensorFlow backend packaging confusion." + "reason": "ImportError for layer_type_validation is unrelated to _torch_extract_fbank_features()." }, { "left": "issue:43937", "right": "issue:44496", "accept": false, - "reason": "Both are loading-time errors, but one is an invalid GenerationConfig and the other is an unrecognized model/config issue; different root causes." + "reason": "Invalid GenerationConfig for GLM-5 is different from unrecognized model config.json." }, { - "left": "issue:38617", - "right": "issue:44220", + "left": "issue:42222", + "right": "issue:45542", "accept": false, - "reason": "An import error for layer_type_validation is unrelated to a torch audio feature-extraction bug." + "reason": "vitpose model breakage and missing TensorFlow backend are unrelated." }, { - "left": "issue:42175", - "right": "issue:42222", + "left": "issue:43066", + "right": "issue:43122", "accept": false, - "reason": "TensorFlow installation/backend packaging is unrelated to vitpose model breakage." + "reason": "Tokenizer decoder type mismatch and changed tokenization output are distinct regressions." }, { - "left": "issue:43066", - "right": "issue:43122", + "left": "issue:42175", + "right": "issue:42222", "accept": false, - "reason": "Both concern tokenizer behavior changes, but one is a decoder-type mismatch while the other is changed tokenization output; not enough evidence of the same defect." + "reason": "TensorFlow extra/backend packaging issue vs vitpose model breakage." }, { "left": "issue:42371", "right": "issue:42831", "accept": false, - "reason": "Both touch numerical settings, but TF32 API guidance and FineGrainedFP8 accuracy are different features and failure modes." + "reason": "TF32 API warning is unrelated to FineGrainedFP8 accuracy." }, { "left": "issue:44075", "right": "issue:45656", "accept": false, - "reason": "Different optimizer problems: ignored SGD args vs optimizer.step being called twice under deepspeed." + "reason": "Optimizer SGD args ignored and optimizer step called twice under deepspeed are different optimizer bugs." }, { "left": "issue:43673", "right": "issue:45341", "accept": false, - "reason": "Cache handling during chunked prefill is unrelated to a testing_utils bug." + "reason": "Generation cache missing in v5 RC is unrelated to the testing_utils bug." }, { "left": "issue:44410", "right": "issue:45478", "accept": false, - "reason": "Different model families and different loading errors: qwen3next missing projections vs Qwen3.5 MoE from_pretrained failure." + "reason": "Missing projections and Qwen3.5 from_pretrained error are different code-path failures." }, { "left": "issue:43012", "right": "issue:43577", "accept": false, - "reason": "Both involve dtype/precision, but one is a compile-time warning and the other is model submodules staying float32 after load; not the same bug." + "reason": "bfloat16 precision warning and BLIP2 dtype propagation bug are different dtype issues." }, { "left": "issue:44484", "right": "issue:45478", "accept": false, - "reason": "save_pretrained shard size behavior is unrelated to a Qwen3.5 loading error." + "reason": "Shard-size question and Qwen3.5 from_pretrained regression are unrelated." }, { "left": "issue:43097", "right": "issue:44493", "accept": false, - "reason": "One is about a removed config field, the other about unexpected position-id keys; different config regressions." + "reason": "Removed tie_embeddings_and_encoder_decoder API and unexpected key with position id are separate regressions." }, { "left": "issue:41628", "right": "issue:45071", "accept": false, - "reason": "An AutoImageProcessor import failure and a PretrainedConfig type-checking break are separate API regressions." + "reason": "AutoImageProcessor ImportError and PreTrainedConfig type-checking breakage are unrelated." }, { "left": "issue:43874", "right": "issue:44261", "accept": false, - "reason": "Missing image-patch helper methods are unrelated to MLA normalization precision handling." + "reason": "Missing image patch-count helper and MLA eps precision bug are different issues." }, { "left": "issue:44263", "right": "issue:44485", "accept": false, - "reason": "A torch.split return-value issue in a DSA indexer is not the same as the GLM-5 RoPE implementation discussion." + "reason": "torch.split return handling in an indexer and GLM-5 RoPE implementation are unrelated." }, { "left": "issue:45071", "right": "issue:45542", "accept": false, - "reason": "Type-checking regressions and TensorFlow backend packaging are unrelated." + "reason": "Type-checking breakage is unrelated to the missing TensorFlow backend error." }, { "left": "issue:44263", "right": "issue:46172", "accept": false, - "reason": "Different models and failure modes: GlmMoeDsaIndexer split handling vs deepseek-v4 tensor transpose behavior." + "reason": "DSA indexer split handling and deepseek-v4 transpose/dtype issue are unrelated." }, { "left": "issue:44360", "right": "issue:46172", "accept": false, - "reason": "A missing ReLU in the DSA indexer is unrelated to the deepseek-v4 tensor operation bug." + "reason": "DSA indexer lacking ReLU and deepseek-v4 comb transpose bug are different problems." }, { "left": "issue:43012", "right": "issue:43408", "accept": false, - "reason": "A bfloat16 precision warning is unrelated to a sam3_video vs sam3_tracker model-type mismatch." + "reason": "bfloat16 warning and sam3 model-type mismatch are different model-loading problems." }, { "left": "issue:43441", "right": "issue:44829", "accept": false, - "reason": "Both mention FlashAttention, but one is a Ministral-3 load-time failure and the other is degenerate training with flash_attention_3; not the same fix." - }, - { - "left": "issue:43931", - "right": "issue:44410", - "accept": false, - "reason": "Different Qwen variants and different symptoms: weight-shape mismatch vs missing attention projections." - }, - { - "left": "issue:43931", - "right": "issue:44038", - "accept": false, - "reason": "Both are Qwen3-VL-related loading issues, but they concern different models and different error classes; not clearly one bug." - }, - { - "left": "issue:43335", - "right": "issue:45071", - "accept": false, - "reason": "SwitchTransformers sparse-layer construction is unrelated to config type checking." - }, - { - "left": "issue:43673", - "right": "issue:46123", - "accept": false, - "reason": "Generation cache behavior during chunked prefill is unrelated to MaskGenerationPipeline dropping results on partial batches." - }, - { - "left": "issue:43097", - "right": "issue:43452", - "accept": false, - "reason": "A removed embedding-tie config option is unrelated to gguf_file breaking tokenizer/model loading." - }, - { - "left": "issue:45003", - "right": "issue:45341", - "accept": false, - "reason": "Unsafe sys.modules access and a testing_utils bug are different parts of the codebase with no shared failure mode." - }, - { - "left": "issue:43986", - "right": "issue:45341", - "accept": false, - "reason": "A missing torchvision dependency for AutoProcessor video models is unrelated to testing_utils." - }, - { - "left": "issue:44117", - "right": "issue:44987", - "accept": false, - "reason": "A tokenizer mapping None-handling bug and a specific repo loading failure are too loosely related to treat as duplicates." - }, - { - "left": "issue:43408", - "right": "issue:44368", - "accept": false, - "reason": "Model-type mismatch warning and tie_word_embeddings warning are separate configuration warnings." - }, - { - "left": "issue:43986", - "right": "issue:44484", - "accept": false, - "reason": "AutoProcessor/torchvision dependency handling is unrelated to save_pretrained shard-size defaults." - }, - { - "left": "issue:43874", - "right": "issue:44263", - "accept": false, - "reason": "Missing image patch counting helper and torch.split return values are unrelated implementation bugs." - }, - { - "left": "issue:43986", - "right": "issue:45372", - "accept": false, - "reason": "Different missing dependency/import problems: torchvision absence vs missing ReasoningEffort from mistral_common." + "reason": "FlashAttention failure in Ministral-3 and flash_attention_3 training degeneration are unrelated." } ] }, @@ -13828,67 +13848,80 @@ "component_index": 1, "component_count": 17, "review_unit_index": 44, - "review_unit_count": 49, + "review_unit_count": 50, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-427", + "cluster_id": "cluster-10000-428", "nodes": [ "issue:38617", "issue:42757", - "issue:42907", - "issue:43012", "issue:43064", "issue:43066", "issue:43097", "issue:43208", - "issue:43352", + "issue:43335", + "issue:43408", + "issue:43452", "issue:43502", "issue:43531", "issue:43550", - "issue:43792", + "issue:43673", "issue:43824", "issue:43825", "issue:43873", + "issue:43874", "issue:43906", - "issue:43976", + "issue:43931", "issue:43986", "issue:43992", "issue:44038", "issue:44062", "issue:44117", - "issue:44162", "issue:44261", + "issue:44263", "issue:44315", "issue:44351", "issue:44368", - "issue:44380", "issue:44410", "issue:44484", "issue:44485", "issue:44655", "issue:44704", "issue:44938", + "issue:44987", + "issue:45003", + "issue:45071", "issue:45103", "issue:45341", + "issue:45372", "issue:45375", "issue:45440", - "issue:45479", "issue:45588", - "issue:45753", "issue:46050", "issue:46123", - "issue:46132", "issue:46172", "issue:46178" ], "soft_pairs": [ + "issue:43931|issue:44410", + "issue:43931|issue:44038", + "issue:43335|issue:45071", + "issue:43673|issue:46123", + "issue:43097|issue:43452", + "issue:45003|issue:45341", + "issue:43986|issue:45341", + "issue:44117|issue:44987", + "issue:43408|issue:44368", + "issue:43986|issue:44484", + "issue:43874|issue:44263", + "issue:43986|issue:45372", "issue:45440|issue:46172", "issue:43531|issue:45341", - "issue:44261|issue:44485", "issue:38617|issue:42757", + "issue:44261|issue:44485", "issue:43064|issue:44315", "issue:43097|issue:46050", "issue:43097|issue:45588", @@ -13905,239 +13938,234 @@ "issue:44938|issue:46178", "issue:43208|issue:43550", "issue:43992|issue:44484", - "issue:43825|issue:44655", - "issue:43012|issue:43550", - "issue:43012|issue:44368", - "issue:42907|issue:45753", - "issue:42907|issue:44655", - "issue:43352|issue:44380", - "issue:44162|issue:46172", - "issue:43352|issue:43976", - "issue:44380|issue:45479", - "issue:43792|issue:46123", - "issue:44380|issue:46132", - "issue:43064|issue:44380" + "issue:43825|issue:44655" ], - "prepared_review_unit_hash": "0d87e5b665a42ad11dff13527e5274a2643bfcb507dbb3fdf101bc901aad343c", + "prepared_review_unit_hash": "dccbd5bd37c7d5b432acbef53af0a81bca7ef2d960739c78fbd39d4bd2d5fc0c", "component_budget": { - "node_count": 427, - "item_count": 427, - "soft_pair_count": 1576, - "serialized_chars": 816182, - "estimated_input_tokens": 204046, - "estimated_eval_tokens": 408348 + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 }, "budget": { - "node_count": 47, - "item_count": 47, - "soft_pair_count": 32, - "serialized_chars": 36511, - "estimated_input_tokens": 9128, - "estimated_eval_tokens": 18512 + "node_count": 48, + "item_count": 48, + "soft_pair_count": 33, + "serialized_chars": 37511, + "estimated_input_tokens": 9378, + "estimated_eval_tokens": 19012 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is heterogeneous: the soft matches are mostly superficial Transformers/v5 or model-family wording overlaps, not true duplicates. I would not merge any of the soft edge pairs.", - "confidence": 0.93, - "canonical_issue_reason": "No safe canonical issue: the items span unrelated import errors, model-specific bugs, pipeline behavior, and version-migration regressions.", + "summary": "This cluster is a grab bag of mostly unrelated Transformers regression reports, many centered on v5 migration breakages. None of the soft pairs look like the same underlying bug or change, so all are rejected as duplicates.", + "confidence": 0.91, + "canonical_issue_reason": "Issue 43097 is the broadest, clearest v5 regression report in the set and reads like a representative umbrella complaint about a breaking API removal; the other issues are separate bugs in different code paths.", "canonical_pr_reason": null, - "best_issue_reason": "No single issue is a strong cluster representative because the similarity is broad and lexical rather than the same underlying bug.", + "best_issue_reason": "Issue 43097 is the best single representative of the cluster\u2019s general v5-breaking-change theme, but it should not be treated as a duplicate target for the other reports.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:45440", - "right": "issue:46172", + "left": "issue:43931", + "right": "issue:44410", "accept": false, - "reason": "Different models and failure modes: DeepSeekV3 divergence vs a deepseek-v4 transpose/dtype bug." + "reason": "Both are Qwen-related loading/model-structure bugs, but they describe different failures: weight-shape mismatch vs missing projections in a different architecture." }, { - "left": "issue:43531", - "right": "issue:45341", + "left": "issue:43931", + "right": "issue:44038", "accept": false, - "reason": "Qwen3-MoE sliding-window behavior is unrelated to a small testing_utils bug." + "reason": "Same broad model family, but the reported failures are distinct Qwen-VL issues with different symptoms and likely different code paths." }, { - "left": "issue:44261", - "right": "issue:44485", + "left": "issue:43335", + "right": "issue:45071", "accept": false, - "reason": "Both are model/config precision issues, but one is an MLA layernorm epsilon mismatch and the other is a GLM-5 RoPE implementation complaint." + "reason": "SwitchTransformers sparse-layer construction and PreTrainedConfig type checking are unrelated bugs." }, { - "left": "issue:38617", - "right": "issue:42757", + "left": "issue:43673", + "right": "issue:46123", "accept": false, - "reason": "Separate import errors from different modules and symbols; not the same missing export." + "reason": "Chunked prefill cache handling and MaskGenerationPipeline final-batch handling are different generation/pipeline defects." }, { - "left": "issue:43064", - "right": "issue:44315", + "left": "issue:43097", + "right": "issue:43452", "accept": false, - "reason": "Different training-stack problems: FSDP2/PEFT optimizer state mismatch vs Liger Kernel not applied with model_init." + "reason": "Both are v5 breakage reports, but one is about a removed config helper and the other about tokenizer/model loading for gguf_file; not the same bug." }, { - "left": "issue:43097", - "right": "issue:46050", + "left": "issue:45003", + "right": "issue:45341", "accept": false, - "reason": "One is a removed model config flag; the other is quantized model save/load breakage." + "reason": "Unsafe sys.modules access and a testing_utils bug are unrelated." }, { - "left": "issue:43097", - "right": "issue:45588", + "left": "issue:43986", + "right": "issue:45341", "accept": false, - "reason": "Unrelated: missing tie_embeddings flag vs flash_attention crash on sink-less models." + "reason": "Video-model loading without torchvision and a testing utility issue do not share the same failure path." }, { - "left": "issue:44062", - "right": "issue:44117", + "left": "issue:44117", + "right": "issue:44987", "accept": false, - "reason": "Tokenizer construction bug vs mapping lookup returning None; not the same code path." + "reason": "TOKENIZER_MAPPING_NAMES returning None and a model-loading failure for physical-intelligence/fast are not the same underlying defect." }, { - "left": "issue:44410", - "right": "issue:45375", + "left": "issue:43408", + "right": "issue:44368", "accept": false, - "reason": "Both mention Qwen variants, but one is missing projections while the other is a config field being dropped." + "reason": "These are different warning/regression reports involving different model families and different config fields." }, { - "left": "issue:38617", - "right": "issue:44351", + "left": "issue:43986", + "right": "issue:44484", "accept": false, - "reason": "Different import failures for different names and modules." + "reason": "AutoProcessor/torchvision loading crash is unrelated to save_pretrained shard-size behavior." }, { - "left": "issue:44038", - "right": "issue:44410", + "left": "issue:43874", + "right": "issue:44263", "accept": false, - "reason": "Qwen3-VL-Moe bug and qwen3next projection issue are distinct model implementations." + "reason": "Different GLM bugs in different components: image processor missing a method versus torch.split return handling." }, { - "left": "issue:43502", - "right": "issue:43992", + "left": "issue:43986", + "right": "issue:45372", "accept": false, - "reason": "Offline-mode network requests and UMT5 embed_tokens loading are different loader bugs." + "reason": "Missing torchvision and missing ReasoningEffort are separate import/dependency problems." }, { - "left": "issue:44704", - "right": "issue:45103", + "left": "issue:45440", + "right": "issue:46172", "accept": false, - "reason": "Both involve pre/post-processing kwargs, but one is AutoProcessor caching arguments and the other is auto_docstring annotation handling." + "reason": "Both are DeepSeek-family issues, but one is an implementation divergence and the other is a specific tensor transpose expression; not clearly the same bug." }, { - "left": "issue:43097", - "right": "issue:43873", + "left": "issue:43531", + "right": "issue:45341", "accept": false, - "reason": "Removed config API and quantization offloading behavior are unrelated." + "reason": "Qwen3-MoE sliding_window behavior and a testing_utils bug are unrelated." }, { - "left": "issue:43986", - "right": "issue:44117", + "left": "issue:38617", + "right": "issue:42757", "accept": false, - "reason": "Video-model loading without torchvision is unrelated to tokenizer mapping assumptions." + "reason": "Both are import errors, but they involve different missing symbols from different modules and different breakages." }, { - "left": "issue:43824", - "right": "issue:43986", + "left": "issue:44261", + "right": "issue:44485", "accept": false, - "reason": "Different surface area: missing Qwen2.5-VL class export vs AutoProcessor/video-model crash." + "reason": "MLA layer-norm epsilon precision and GLM RoPE implementation are different numerical/modeling issues." }, { - "left": "issue:43066", - "right": "issue:43906", + "left": "issue:43064", + "right": "issue:44315", "accept": false, - "reason": "Wrong tokenizer decoder type is not the same as the reproduced issue linked to #38071." + "reason": "Trainer/FSDP2 optimizer-state corruption and Liger Kernel not being applied are separate training-path bugs." }, { - "left": "issue:44938", - "right": "issue:46178", + "left": "issue:43097", + "right": "issue:46050", "accept": false, - "reason": "Python 3.14 load failure vs TensorFlow Text backend mapping error are unrelated environment issues." + "reason": "A removed tie_embeddings_and_encoder_decoder API and a quantized-model save/load regression are different failure modes." }, { - "left": "issue:43208", - "right": "issue:43550", + "left": "issue:43097", + "right": "issue:45588", "accept": false, - "reason": "xLSTM training bugs and Bamba torch.compile/SDPA failure are different model/runtime problems." + "reason": "A v5 API removal and a flash-attention sinkless-model crash do not describe the same code-path problem." }, { - "left": "issue:43992", - "right": "issue:44484", + "left": "issue:44062", + "right": "issue:44117", "accept": false, - "reason": "Loading missing weights and save_pretrained shard-size defaults are not the same bug." + "reason": "AddedToken keyword duplication and tokenizer mapping None are tokenizer-related but distinct defects." }, { - "left": "issue:43825", - "right": "issue:44655", + "left": "issue:44410", + "right": "issue:45375", "accept": false, - "reason": "Pipeline error message regression vs inability to save Pipeline objects." + "reason": "Different Qwen config/model bugs: missing projections versus a missing vision config field." }, { - "left": "issue:43012", - "right": "issue:43550", + "left": "issue:38617", + "right": "issue:44351", "accept": false, - "reason": "Precision warning during compile is not the same as a compile-time SDPA failure." + "reason": "Different missing imports from different namespaces; these are not the same underlying regression." }, { - "left": "issue:43012", - "right": "issue:44368", + "left": "issue:44038", + "right": "issue:44410", "accept": false, - "reason": "Both mention warnings, but one is PyTorch precision output and the other is a tie_word_embeddings config warning." + "reason": "Both involve Qwen models, but one is a VL-MoE bug and the other is a missing-projections issue in a different model variant." }, { - "left": "issue:42907", - "right": "issue:45753", + "left": "issue:43502", + "right": "issue:43992", "accept": false, - "reason": "Dequantized save failure and Qwen3_5 infinite-loop generation are separate bugs." + "reason": "Local_files_only network leakage and UMT5Encoder weight loading are separate from_pretrained issues with different causes." }, { - "left": "issue:42907", - "right": "issue:44655", + "left": "issue:44704", + "right": "issue:45103", "accept": false, - "reason": "Saving dequantized models and saving Pipeline objects are different persistence paths." + "reason": "AutoProcessor kwargs propagation and auto_docstring annotation handling are unrelated." }, { - "left": "issue:43352", - "right": "issue:44380", + "left": "issue:43097", + "right": "issue:43873", "accept": false, - "reason": "FlashAttention support error for a specific model vs GPT2 attention scaling being ignored under fast attention backends." + "reason": "A removed v5 API and quantization offloading behavior are different classes of bugs." }, { - "left": "issue:44162", - "right": "issue:46172", + "left": "issue:43986", + "right": "issue:44117", + "accept": false, + "reason": "Video-model loading without torchvision and tokenizer mapping assumptions are separate loading paths." + }, + { + "left": "issue:43824", + "right": "issue:43986", "accept": false, - "reason": "ESM2 breakage in scientific workflows is unrelated to deepseek-v4 tensor conversion/transpose behavior." + "reason": "Different processor/model import failures for different model families." }, { - "left": "issue:43352", - "right": "issue:43976", + "left": "issue:43066", + "right": "issue:43906", "accept": false, - "reason": "Model-specific Flash Attention support and Python-version compatibility are unrelated." + "reason": "Wrong tokenizer decoder type and an isolated reproduction issue are not the same bug." }, { - "left": "issue:44380", - "right": "issue:45479", + "left": "issue:44938", + "right": "issue:46178", "accept": false, - "reason": "Attention scaling ignored under SDPA/FlashAttention is unrelated to a degenerate zero-loss classification bug." + "reason": "Python 3.14 import/load failure and an unsupported tensorflow_text backend error are unrelated environment problems." }, { - "left": "issue:43792", - "right": "issue:46123", + "left": "issue:43208", + "right": "issue:43550", "accept": false, - "reason": "Whisper model runtime failure and mask-generation partial-batch dropping are different pipeline bugs." + "reason": "xLSTM training blockers and Bamba torch.compile/SDPA failure are different model/backend bugs." }, { - "left": "issue:44380", - "right": "issue:46132", + "left": "issue:43992", + "right": "issue:44484", "accept": false, - "reason": "GPT2 backend scaling bug is unrelated to AttentionInterface.register changing function behavior." + "reason": "Missing embed_tokens.weight and default max_shard_size are unrelated save/load behaviors." }, { - "left": "issue:43064", - "right": "issue:44380", + "left": "issue:43825", + "right": "issue:44655", "accept": false, - "reason": "Distributed training optimizer-state mismatch is not the same as backend attention scaling being ignored." + "reason": "Pipeline translation error messaging and pipeline save_pretrained support are different pipeline features." } ] }, @@ -14153,22 +14181,23 @@ "component_index": 1, "component_count": 17, "review_unit_index": 45, - "review_unit_count": 49, + "review_unit_count": 50, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-427", + "cluster_id": "cluster-10000-428", "nodes": [ - "issue:36010", "issue:36683", "issue:38617", "issue:42490", "issue:42757", "issue:42898", + "issue:42907", "issue:42994", "issue:43012", + "issue:43064", "issue:43066", "issue:43208", "issue:43299", @@ -14176,41 +14205,51 @@ "issue:43475", "issue:43502", "issue:43526", + "issue:43550", "issue:43582", - "issue:43611", "issue:43618", "issue:43756", "issue:43761", + "issue:43792", "issue:43824", "issue:43867", "issue:43901", - "issue:43937", "issue:43976", "issue:44038", - "issue:44077", + "issue:44162", "issue:44222", "issue:44230", + "issue:44368", "issue:44380", "issue:44483", "issue:44485", "issue:44514", - "issue:44625", "issue:44655", "issue:44683", - "issue:44938", - "issue:45216", "issue:45276", "issue:45290", "issue:45381", "issue:45478", - "issue:45593", + "issue:45479", "issue:45698", "issue:45753", "issue:46077", - "issue:46082", + "issue:46123", + "issue:46132", "issue:46172" ], "soft_pairs": [ + "issue:43012|issue:43550", + "issue:43012|issue:44368", + "issue:42907|issue:45753", + "issue:42907|issue:44655", + "issue:43352|issue:44380", + "issue:44162|issue:46172", + "issue:43352|issue:43976", + "issue:43792|issue:46123", + "issue:44380|issue:45479", + "issue:44380|issue:46132", + "issue:43064|issue:44380", "issue:44483|issue:44514", "issue:44222|issue:45478", "issue:36683|issue:43582", @@ -14221,8 +14260,8 @@ "issue:42757|issue:43824", "issue:45698|issue:46172", "issue:43618|issue:44380", - "issue:44038|issue:44230", "issue:38617|issue:43502", + "issue:44038|issue:44230", "issue:44483|issue:45290", "issue:44483|issue:45381", "issue:43756|issue:44485", @@ -14231,239 +14270,235 @@ "issue:45276|issue:46077", "issue:43901|issue:44655", "issue:42994|issue:44655", - "issue:43299|issue:45753", "issue:42898|issue:43066", - "issue:36010|issue:42757", - "issue:44077|issue:44380", - "issue:36683|issue:44077", - "issue:43611|issue:44655", - "issue:43352|issue:43937", - "issue:44625|issue:46077", - "issue:43867|issue:45216", - "issue:43976|issue:44938", - "issue:46077|issue:46082", - "issue:43352|issue:45593" + "issue:43299|issue:45753" ], - "prepared_review_unit_hash": "03fdb2946b675eec4b4b8987dbf93a96593d54675b0580aabc0d94f318ecfe5a", + "prepared_review_unit_hash": "e67c0f3c353fb1e5b67f8b538fe2638ad28ae3f7bfffc7ab2a237dfa5afb3179", "component_budget": { - "node_count": 427, - "item_count": 427, - "soft_pair_count": 1576, - "serialized_chars": 816182, - "estimated_input_tokens": 204046, - "estimated_eval_tokens": 408348 + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 32, - "serialized_chars": 36824, - "estimated_input_tokens": 9206, - "estimated_eval_tokens": 18668 + "soft_pair_count": 33, + "serialized_chars": 37291, + "estimated_input_tokens": 9323, + "estimated_eval_tokens": 18902 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly a grab-bag of unrelated Hugging Face Transformers issues that only share broad themes like imports, save/load, or model-family regressions. I found no soft pair that clearly meets the same-underlying-bug threshold, so all soft edges are rejected.", - "confidence": 0.78, - "canonical_issue_reason": "issue:45698 is the broadest open bug in the set and the most plausible umbrella for related save/load regressions around save_pretrained/from_pretrained and custom module handling.", + "summary": "These items are not a true duplicate cluster; they span unrelated regressions across model loading, saving, compilation, tokenization, and API behavior. No soft edge looks safe to merge, so the best outcome is to keep them separate and use one broad issue only as a loose cluster anchor.", + "confidence": 0.37, + "canonical_issue_reason": "issue:44483 is the broadest, highest-impact regression in the set and the most plausible cluster anchor by title and severity; the rest are mostly model-specific or orthogonal failures.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45698 is the most central and actionable issue in the cluster: open, user-facing, and tied to a core serialization/loading path rather than a model-specific edge case.", + "best_issue_reason": "issue:44483 is the best single representative issue if one must choose one, but it should not be treated as a duplicate magnet for the other items.", "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:43012", + "right": "issue:43550", + "accept": false, + "reason": "Different failures: a PyTorch precision warning during compile versus a Bamba SDPA torch.compile error." + }, + { + "left": "issue:43012", + "right": "issue:44368", + "accept": false, + "reason": "Unrelated problems: bfloat16 compile warning versus a config warning about tie_word_embeddings in Qwen3.5 fine-tuning." + }, + { + "left": "issue:42907", + "right": "issue:45753", + "accept": false, + "reason": "Both involve model saving/loading, but one is dequantized Ministral/Devstral save failure and the other is an image-specific infinite-loop inference bug." + }, + { + "left": "issue:42907", + "right": "issue:44655", + "accept": false, + "reason": "Saving a dequantized model is not the same as saving Pipeline objects; different code paths and failure modes." + }, + { + "left": "issue:43352", + "right": "issue:44380", + "accept": false, + "reason": "Flash Attention support error for NemotronH versus GPT2 attention scaling being ignored under SDPA/FlashAttention are distinct bugs." + }, + { + "left": "issue:44162", + "right": "issue:46172", + "accept": false, + "reason": "ESM2 breakage is a model-loading/runtime issue, while deepseek-v4 hits a specific tensor transpose bug." + }, + { + "left": "issue:43352", + "right": "issue:43976", + "accept": false, + "reason": "Nemotron Flash Attention support and Python version compatibility are unrelated." + }, + { + "left": "issue:43792", + "right": "issue:46123", + "accept": false, + "reason": "Whisper inference failure is unrelated to MaskGenerationPipeline dropping the final partial batch." + }, + { + "left": "issue:44380", + "right": "issue:45479", + "accept": false, + "reason": "Attention scaling being ignored is different from a degenerate zero-loss bug in sequence classification." + }, + { + "left": "issue:44380", + "right": "issue:46132", + "accept": false, + "reason": "GPT2 backend-specific attention scaling is unrelated to AttentionInterface.register changing callback behavior." + }, + { + "left": "issue:43064", + "right": "issue:44380", + "accept": false, + "reason": "Trainer/FSDP2 optimizer-state corruption is a distributed training bug, not an attention-backend config regression." + }, { "left": "issue:44483", "right": "issue:44514", "accept": false, - "reason": "Different failure surfaces: chat-completions request validation vs batched processor templating with tool calls; not the same concrete bug." + "reason": "Chat/completions request rejection is unrelated to Qwen2_5_VLProcessor crashing on batched chat templates." }, { "left": "issue:44222", "right": "issue:45478", "accept": false, - "reason": "Both involve MoE/save/load wording, but one is FP8 save_pretrained and the other is a from_pretrained load error; too different to merge." + "reason": "Both mention FP8/quantization and MoE, but one is save_pretrained failure and the other is from_pretrained loading error." }, { "left": "issue:36683", "right": "issue:43582", "accept": false, - "reason": "Completely unrelated: Gemma3Config vocab_size attribute error vs Apple Silicon TypeError in caching_allocator_warmup." + "reason": "Gemma3Config missing vocab_size and Apple Silicon caching_allocator_warmup TypeError are unrelated." }, { "left": "issue:43012", "right": "issue:44683", "accept": false, - "reason": "Both touch PyTorch compilation behavior, but one is a warning and the other is a torch>=2.9 flex_attention failure; different bugs." + "reason": "A warning during bfloat16 compile is not the same as flex_attention failing on torch >= 2.9." }, { "left": "issue:43208", "right": "issue:43526", "accept": false, - "reason": "xLSTM training bugs and BeitImageProcessorFast label reduction are unrelated code paths." + "reason": "xLSTM training bugs and BeitImageProcessorFast reduce_labels truncation are different subsystems and bugs." }, { "left": "issue:42490", "right": "issue:43012", "accept": false, - "reason": "A trailing-slash import warning and a bfloat16 compilation warning are not the same underlying issue." + "reason": "Path-trailing-slash import warning differs completely from a compile-time PyTorch precision warning." }, { "left": "issue:43618", "right": "issue:43761", "accept": false, - "reason": "Both are CLIP v5 regressions, but one is missing attentions and the other missing hidden_states; related area, not clearly the same bug." + "reason": "CLIP attentions being unassigned and CLIPVisionModel returning hidden_states=None are separate v5 regressions." }, { "left": "issue:42757", "right": "issue:43824", "accept": false, - "reason": "Distinct import errors for different missing symbols and different packages/modules." + "reason": "Missing is_offline_mode and missing Qwen2_5_VLForConditionalGeneration are different import errors from different symbols." }, { "left": "issue:45698", "right": "issue:46172", "accept": false, - "reason": "General save/load custom-module mismatch vs a DeepSeek-V4 dtype/transpose bug; unrelated." + "reason": "A custom-module reload bug after save_pretrained is unrelated to a low-level tensor transpose bug in deepseek-v4." }, { "left": "issue:43618", "right": "issue:44380", "accept": false, - "reason": "CLIP output fields missing vs GPT2 attention scaling ignored under SDPA/FlashAttention are separate code paths." + "reason": "CLIP output attentions regression and GPT2 attention scaling behavior are distinct model/backend issues." }, { - "left": "issue:44038", - "right": "issue:44230", + "left": "issue:38617", + "right": "issue:43502", "accept": false, - "reason": "Same model family name appears, but one is a loading bug and the other is an FP8 inference support request." + "reason": "Missing layer_type_validation import and local_files_only still making API requests are unrelated import/network bugs." }, { - "left": "issue:38617", - "right": "issue:43502", + "left": "issue:44038", + "right": "issue:44230", "accept": false, - "reason": "Importing a removed utility and local_files_only still making API requests are unrelated." + "reason": "Both involve Qwen3-VL-Moe, but one is a generic transformers loading bug and the other is FP8 inference support." }, { "left": "issue:44483", "right": "issue:45290", "accept": false, - "reason": "API request rejection and chat template crash with tool-call messages are different problems." + "reason": "Chat/completions request handling is unrelated to apply_chat_template crashing on assistant tool-call messages." }, { "left": "issue:44483", "right": "issue:45381", "accept": false, - "reason": "Chat/completions request handling and Qwen2.5-VL video position IDs are unrelated." + "reason": "API request acceptance and Qwen2.5-VL video vision_position_ids are different failure modes in different code paths." }, { "left": "issue:43756", "right": "issue:44485", "accept": false, - "reason": "Both mention RoPE, but they concern different models and different implementation defects." + "reason": "Smollm3 RoPE layer dropping and GLM-5 RoPE implementation discussion are not the same concrete bug." }, { "left": "issue:43299", "right": "issue:43867", "accept": false, - "reason": "Qwen3VL MoE loading regression and sorted state_dict load error do not appear to be the same bug." + "reason": "Qwen3VL-Moe loading breakage and state_dict sorting load errors are different model-loading problems." }, { "left": "issue:36683", "right": "issue:43475", "accept": false, - "reason": "Gemma3Config vocab_size and SAM3 video output missing fpn_position_embeddings are unrelated attribute errors." + "reason": "Gemma3Config vocab_size AttributeError and SAM 3 Video missing fpn_position_embeddings are unrelated missing-attribute bugs." }, { "left": "issue:45276", "right": "issue:46077", "accept": false, - "reason": "Both are Gemma4 issues, but embedding resize propagation and bidirectional attention mask construction are separate defects." + "reason": "Gemma4 token embedding resize behavior and bidirectional attention mask construction are different bugs in different parts of the model." }, { "left": "issue:43901", "right": "issue:44655", "accept": false, - "reason": "Documentation mismatch for return_all_scores and pipeline save_pretrained failure are not the same issue." + "reason": "Deprecated docs wording for return_all_scores is unrelated to pipeline save_pretrained failure." }, { "left": "issue:42994", "right": "issue:44655", "accept": false, - "reason": "Both involve saving, but quantized model save failure and Pipeline serialization are different paths." - }, - { - "left": "issue:43299", - "right": "issue:45753", - "accept": false, - "reason": "Model loading failure and an infinite-loop inference bug on a specific image are unrelated." + "reason": "Quantized model saving failure and pipeline serialization failure are different save_pretrained bugs." }, { "left": "issue:42898", "right": "issue:43066", "accept": false, - "reason": "Tokenizer behavior change and wrong decoder type are both tokenizer-related, but they are distinct regressions." - }, - { - "left": "issue:36010", - "right": "issue:42757", - "accept": false, - "reason": "Different missing imports from different modules; no common underlying change is evident." - }, - { - "left": "issue:44077", - "right": "issue:44380", - "accept": false, - "reason": "PatchTSMixer post_init validation and GPT2 attention scaling under SDPA/FlashAttention are unrelated." - }, - { - "left": "issue:36683", - "right": "issue:44077", - "accept": false, - "reason": "Gemma3Config attribute failure and PatchTSMixer post_init behavior are unrelated." - }, - { - "left": "issue:43611", - "right": "issue:44655", - "accept": false, - "reason": "General base_model_prefix loading regression and pipeline save_pretrained failure are separate serialization/loading bugs." - }, - { - "left": "issue:43352", - "right": "issue:43937", - "accept": false, - "reason": "Flash Attention support error for Nemotron and GLM-5 GenerationConfig validation are unrelated." - }, - { - "left": "issue:44625", - "right": "issue:46077", - "accept": false, - "reason": "Qwen3.5 num_labels propagation and Gemma4 causal-mask behavior are different model/config bugs." - }, - { - "left": "issue:43867", - "right": "issue:45216", - "accept": false, - "reason": "State_dict ordering load error and Qwen3.5 checkpoint save corruption are distinct save/load problems." - }, - { - "left": "issue:43976", - "right": "issue:44938", - "accept": false, - "reason": "Both are Python compatibility complaints, but the versions and failure modes differ enough that they are not the same bug." - }, - { - "left": "issue:46077", - "right": "issue:46082", - "accept": false, - "reason": "Gemma4 attention mask construction and LlamaConfig head_dim validation are unrelated." + "reason": "Tokenizer clean_up_tokenization_spaces behavior changes and wrong decoder type are unrelated v5 issues." }, { - "left": "issue:43352", - "right": "issue:45593", + "left": "issue:43299", + "right": "issue:45753", "accept": false, - "reason": "Flash Attention support on Nemotron and missing auxiliary losses in D-FINE are completely different issues." + "reason": "Qwen3.5-VL/Moe loading failure and a specific-image infinite loop are different bugs with different triggers." } ] }, @@ -14479,50 +14514,49 @@ "component_index": 1, "component_count": 17, "review_unit_index": 46, - "review_unit_count": 49, + "review_unit_count": 50, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-427", + "cluster_id": "cluster-10000-428", "nodes": [ "issue:28282", "issue:33453", + "issue:36010", "issue:36296", "issue:36683", "issue:41669", "issue:42490", + "issue:42757", "issue:42890", "issue:43010", "issue:43064", "issue:43352", - "issue:43502", "issue:43525", "issue:43575", - "issue:43576", + "issue:43611", "issue:43618", - "issue:43723", "issue:43867", "issue:43931", + "issue:43937", "issue:43976", "issue:43992", - "issue:44062", "issue:44075", "issue:44077", - "issue:44112", "issue:44242", "issue:44246", "issue:44355", "issue:44360", + "issue:44380", "issue:44484", "issue:44492", - "issue:44556", + "issue:44625", + "issue:44655", "issue:44855", "issue:44857", "issue:44938", - "issue:44964", - "issue:44977", "issue:45071", "issue:45216", "issue:45278", @@ -14532,11 +14566,22 @@ "issue:45440", "issue:45478", "issue:45561", - "issue:45600", + "issue:45593", "issue:45850", - "issue:46132" + "issue:46077", + "issue:46082" ], "soft_pairs": [ + "issue:36010|issue:42757", + "issue:44077|issue:44380", + "issue:36683|issue:44077", + "issue:43352|issue:43937", + "issue:43611|issue:44655", + "issue:43867|issue:45216", + "issue:44625|issue:46077", + "issue:43976|issue:44938", + "issue:46077|issue:46082", + "issue:43352|issue:45593", "issue:36683|issue:43525", "issue:43618|issue:45216", "issue:36683|issue:44242", @@ -14560,217 +14605,240 @@ "issue:28282|issue:33453", "issue:43867|issue:44484", "issue:43618|issue:44857", - "issue:44360|issue:45440", - "issue:43618|issue:44977", - "issue:42890|issue:44964", - "issue:43502|issue:44062", - "issue:42890|issue:44112", - "issue:42890|issue:45600", - "issue:43576|issue:46132", - "issue:43723|issue:44556" + "issue:44360|issue:45440" ], - "prepared_review_unit_hash": "0da1c5b4ef5da8f892711965d003d7b96853879d717f0b0cf355f377f3254ab4", + "prepared_review_unit_hash": "c90995dda5f7afcae37de3aa51afdba66abafbd636ca643fc106dcf81c6114df", "component_budget": { - "node_count": 427, - "item_count": 427, - "soft_pair_count": 1576, - "serialized_chars": 816182, - "estimated_input_tokens": 204046, - "estimated_eval_tokens": 408348 + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 31, - "serialized_chars": 36497, - "estimated_input_tokens": 9125, - "estimated_eval_tokens": 18506 + "soft_pair_count": 34, + "serialized_chars": 37386, + "estimated_input_tokens": 9347, + "estimated_eval_tokens": 18950 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is highly heterogeneous: the soft-similarity pairs mostly share broad v5/loading/import wording but do not describe the same underlying bug or change, so none should be merged as duplicates.", - "confidence": 0.34, - "canonical_issue_reason": "issue:45278 is the broadest and most representative report in the set: a general import-failure regression after upgrading from 4.57.0 to 5.5.0.", + "summary": "This cluster is very heterogeneous: it mixes unrelated import errors, model/config regressions, save/load bugs, training issues, and test/performance complaints. None of the soft edges look like true duplicates of the same underlying bug, so I would not merge any of them.", + "confidence": 0.93, + "canonical_issue_reason": "If forced to pick one issue, issue 36010 is the cleanest concrete bug report: a specific import failure with a clear missing symbol traceback. It is only a weak canonical choice because the surrounding items are mostly unrelated.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45278 is the best single issue candidate because it is the most general, user-facing regression report and fits the cluster\u2019s broad import-regression theme better than the model-specific cases.", + "best_issue_reason": "Issue 36010 is the best representative issue for triage clarity, but not because it matches the rest of the cluster; it is simply the most specific and actionable report among a mixed set.", "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:36010", + "right": "issue:42757", + "accept": false, + "reason": "Both are import-name errors, but they involve different missing symbols from different packages and likely different fixes." + }, + { + "left": "issue:44077", + "right": "issue:44380", + "accept": false, + "reason": "One is about optimizer arguments not being used; the other is about attention scaling being ignored in SDPA/FlashAttention. Different code paths." + }, + { + "left": "issue:36683", + "right": "issue:44077", + "accept": false, + "reason": "Gemma3 config missing vocab_size is unrelated to SGD optimizer args not being used." + }, + { + "left": "issue:43352", + "right": "issue:43937", + "accept": false, + "reason": "FlashAttention unsupported for NemotronHForCausalLM is unrelated to GLM-5 generation config validation." + }, + { + "left": "issue:43611", + "right": "issue:44655", + "accept": false, + "reason": "base_model_prefix loading regression and Pipeline save_pretrained failure are different save/load behaviors with different failure modes." + }, + { + "left": "issue:43867", + "right": "issue:45216", + "accept": false, + "reason": "State-dict ordering load error is not the same bug as Qwen3.5 checkpoint saving regression." + }, + { + "left": "issue:44625", + "right": "issue:46077", + "accept": false, + "reason": "num_labels propagation in Qwen3.5 is unrelated to Gemma4 attention mask construction." + }, + { + "left": "issue:43976", + "right": "issue:44938", + "accept": false, + "reason": "Both mention Python versions, but one is a 5.1.0 compatibility issue and the other is a Python 3.14 import failure with a different cause." + }, + { + "left": "issue:46077", + "right": "issue:46082", + "accept": false, + "reason": "Gemma4 bidirectional attention masking and LlamaConfig head_dim validation are separate configuration bugs." + }, + { + "left": "issue:43352", + "right": "issue:45593", + "accept": false, + "reason": "Unsupported FlashAttention for Nemotron and D-FINE auxiliary-loss behavior are unrelated model issues." + }, { "left": "issue:36683", "right": "issue:43525", "accept": false, - "reason": "Both are config-attribute errors, but for different models and different missing fields (`vocab_size` vs `pad_token_id`)." + "reason": "Missing vocab_size in Gemma3Config and missing pad_token_id in Llama4Config are similar symptom types but different configs and fields." }, { "left": "issue:43618", "right": "issue:45216", "accept": false, - "reason": "One is a CLIP output-attentions regression; the other is a Qwen3.5 save_pretrained checkpoint regression." + "reason": "CLIPOutput attentions assignment and Qwen3.5 save_pretrained regression are distinct subsystems." }, { "left": "issue:36683", "right": "issue:44242", "accept": false, - "reason": "Unrelated bugs: missing config field for Gemma3 vs missing load-balancing loss when router logits are off." + "reason": "A missing Gemma3 config attribute is not the same as load balancing loss omitted when output_router_logits=False." }, { "left": "issue:43352", "right": "issue:43931", "accept": false, - "reason": "Different concrete failures: FlashAttention 2 not supported vs a Qwen3-VL weight-shape mismatch." + "reason": "Nemotron FlashAttention support and Qwen3-VL weight-shape mismatch are different model-loading problems." }, { "left": "issue:43064", "right": "issue:44492", "accept": false, - "reason": "FSDP/PEFT optimizer-state corruption is unrelated to a cache-strategy typo." + "reason": "FSDP2 optimizer-state corruption is unrelated to a typo in cache strategy docs/code." }, { "left": "issue:43010", "right": "issue:45278", "accept": false, - "reason": "Cache `update()` no_grad behavior is a different code path from generic import errors after upgrade." + "reason": "Cache update no_grad behavior and broad import errors after upgrade are not the same issue." }, { "left": "issue:43992", "right": "issue:45310", "accept": false, - "reason": "UMT5 embedding-loading loss and Qwen3.5 MoE `from_pretrained` errors are different model-loading bugs." + "reason": "UMT5Encoder missing embed_tokens.weight is unrelated to Qwen3.5 MoE from_pretrained errors." }, { "left": "issue:44355", "right": "issue:44855", "accept": false, - "reason": "Compiled-file import failures are not the same as the Python 3.13 `torch.jit.script` parsing/IndentationError issue." + "reason": "Both are import-related, but one concerns compiled Python files and the other is a Python 3.13 IndentationError from a torch.jit.script decorator/comment parsing issue." }, { "left": "issue:43867", "right": "issue:45375", "accept": false, - "reason": "State-dict sorting load errors and missing config fields in Qwen3.5 vision config are different serialization issues." + "reason": "State-dict sorting load error and missing deepstack_visual_indexes are different Qwen3.5 failure modes." }, { "left": "issue:41669", "right": "issue:44246", "accept": false, - "reason": "Import-star performance regressions and the router-loss bug affect different behavior and code paths." + "reason": "Both touch import time, but one is about removing import * usage and the other is about slow imports; not the same bug." }, { "left": "issue:36683", "right": "issue:43064", "accept": false, - "reason": "A missing config attribute is unrelated to the FSDP2/PEFT optimizer-state bug." + "reason": "Gemma3 config missing vocab_size and optimizer-state corruption under FSDP2 are unrelated." }, { "left": "issue:44938", "right": "issue:45071", "accept": false, - "reason": "Python 3.14 import failure and PretrainedConfig type-checking breakage are different regression types." + "reason": "Python 3.14 load failure and PretrainedConfig type-checking breakage are different regression reports." }, { "left": "issue:42890", "right": "issue:45561", "accept": false, - "reason": "Missing `set_seed()` in tests is unrelated to the xdist race on `captured_info.txt`." + "reason": "Missing set_seed in SamHQ tests and xdist race on captured_info.txt are separate test-flakiness issues." }, { "left": "issue:43010", "right": "issue:45850", "accept": false, - "reason": "A specific cache-update bug is not the same as the vague 'too many models broken' report." + "reason": "A specific cache/no_grad bug is not the same as a broad 'too many models broken' complaint." }, { "left": "issue:43867", "right": "issue:45357", "accept": false, - "reason": "Load failure from sorted state_dicts is different from incorrect visual-encoder keys saved by Qwen3.5." + "reason": "Sorted state-dict loading error and incorrect visual encoder keys in save_pretrained are different save/load regressions." }, { "left": "issue:42490", "right": "issue:44075", "accept": false, - "reason": "A trailing-slash local path warning is unrelated to ignored SGD optimizer args." + "reason": "Local-path slash import warning is unrelated to SGD args not being used." }, { "left": "issue:43010", "right": "issue:43976", "accept": false, - "reason": "Cache update/no_grad behavior is unrelated to the Python 3.9 vs 3.10 compatibility problem." + "reason": "Cache update decoration and Python-version compatibility are unrelated." }, { "left": "issue:43618", "right": "issue:44077", "accept": false, - "reason": "CLIP attentions missing and patchtsmixer `post_init` handling are different model-specific issues." + "reason": "CLIPOutput attentions missing and optimizer argument handling are different code paths." }, { "left": "issue:43992", "right": "issue:45478", "accept": false, - "reason": "Both mention Qwen3.5/Qwen3.5-MoE loading, but the concrete failures differ and are not clearly the same bug." + "reason": "Missing embed_tokens.weight in UMT5Encoder is unrelated to Qwen3.5 MoE from_pretrained errors." }, { "left": "issue:36296", "right": "issue:43575", "accept": false, - "reason": "Tensor-parallel training logic and TP-induced OOM on a large model are different problems." + "reason": "Tensor-parallel training bug and Qwen2-57B-A14B-Instruct OOM are related only at a very broad level; they are different failures." }, { "left": "issue:28282", "right": "issue:33453", "accept": false, - "reason": "AutoModel PyTorch import errors and tokenizer-loading regressions are unrelated." + "reason": "AutoModel missing PyTorch and tokenizer-loading regression are unrelated import/loading issues." }, { "left": "issue:43867", "right": "issue:44484", "accept": false, - "reason": "A state_dict load-order bug is unrelated to the `max_shard_size` question." + "reason": "Model loading error from sorted state_dict and a question about max_shard_size are not duplicates." }, { "left": "issue:43618", "right": "issue:44857", "accept": false, - "reason": "Missing CLIP attentions and an AMP/CUDA crash in LwDetrImageLoss are different failures." + "reason": "Missing CLIPOutput attentions and an AMP/CUDA crash in LwDetrImageLoss are unrelated bugs." }, { "left": "issue:44360", "right": "issue:45440", "accept": false, - "reason": "The DSA ReLU discussion and DeepSeekV3MoE divergence are separate model/math issues." - }, - { - "left": "issue:43502", - "right": "issue:44062", - "accept": false, - "reason": "Local-files-only network leakage and `AddedToken()` argument collisions are unrelated." - }, - { - "left": "issue:42890", - "right": "issue:44112", - "accept": false, - "reason": "Seed-setting test flakiness is not the same as the GraniteSpeech CI device-override failure." - }, - { - "left": "issue:42890", - "right": "issue:45600", - "accept": false, - "reason": "Missing `set_seed()` in tests and stale `auto_mappings.py` references are different CI problems." - }, - { - "left": "issue:43576", - "right": "issue:46132", - "accept": false, - "reason": "A broken `transformers env` command and changed `AttentionInterface.register` behavior are different APIs and code paths." - }, - { - "left": "issue:43723", - "right": "issue:44556", - "accept": false, - "reason": "Tokenizer loading in v5 and reloading a v4.57 checkpoint after upgrading are related only broadly by versioning, not the same bug." + "reason": "DSA indexer ReLU discussion and DeepseekV3MoE implementation divergence do not describe the same underlying problem." } ] }, @@ -14786,13 +14854,13 @@ "component_index": 1, "component_count": 17, "review_unit_index": 47, - "review_unit_count": 49, + "review_unit_count": 50, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-427", + "cluster_id": "cluster-10000-428", "nodes": [ "issue:30333", "issue:30990", @@ -14811,13 +14879,14 @@ "issue:43232", "issue:43299", "issue:43502", - "issue:43526", "issue:43576", - "issue:43673", + "issue:43618", + "issue:43723", "issue:43824", "issue:43827", "issue:43874", - "issue:43901", + "issue:44062", + "issue:44112", "issue:44222", "issue:44230", "issue:44315", @@ -14826,30 +14895,37 @@ "issue:44360", "issue:44387", "issue:44485", - "issue:44492", "issue:44556", "issue:44655", "issue:44749", - "issue:44908", "issue:44961", + "issue:44964", + "issue:44977", "issue:45071", "issue:45125", "issue:45341", "issue:45375", "issue:45468", "issue:45478", - "issue:46143", - "issue:46144", + "issue:45600", + "issue:46132", "issue:46178" ], "soft_pairs": [ + "issue:43618|issue:44977", + "issue:42890|issue:44964", + "issue:43502|issue:44062", + "issue:42890|issue:44112", + "issue:42890|issue:45600", + "issue:43576|issue:46132", + "issue:43723|issue:44556", "issue:44387|issue:44556", "issue:42890|issue:45341", "issue:44222|issue:45125", "issue:43010|issue:43065", "issue:42548|issue:44351", - "issue:43502|issue:44485", "issue:44355|issue:44749", + "issue:43502|issue:44485", "issue:45125|issue:45478", "issue:30333|issue:33453", "issue:43010|issue:43232", @@ -14874,270 +14950,270 @@ "issue:44961|issue:45468", "issue:41628|issue:42548", "issue:38617|issue:43576", - "issue:44355|issue:46178", - "issue:46143|issue:46144", - "issue:44908|issue:44961", - "issue:44485|issue:44961", - "issue:43673|issue:44492", - "issue:36296|issue:43901", - "issue:36296|issue:43526" + "issue:44355|issue:46178" ], - "prepared_review_unit_hash": "a640026e536d5601b96c7257ece350959d882fea6e9b95e760911273dea230f3", + "prepared_review_unit_hash": "6996115dfa2504d97f548376f92dc036288768318c2f4004e795050628e5cd43", "component_budget": { - "node_count": 427, - "item_count": 427, - "soft_pair_count": 1576, - "serialized_chars": 816182, - "estimated_input_tokens": 204046, - "estimated_eval_tokens": 408348 + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 }, "budget": { - "node_count": 47, - "item_count": 47, - "soft_pair_count": 38, - "serialized_chars": 37920, - "estimated_input_tokens": 9480, - "estimated_eval_tokens": 19216 + "node_count": 48, + "item_count": 48, + "soft_pair_count": 39, + "serialized_chars": 38792, + "estimated_input_tokens": 9698, + "estimated_eval_tokens": 19652 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is a loose mix of Transformer regression reports, import/load failures, cache/generation bugs, quantization issues, and unrelated docs/test items. The soft pairs mostly share only broad subsystem keywords, not the same concrete underlying defect, so none should be merged as duplicates.", - "confidence": 0.96, - "canonical_issue_reason": "Issue 44556 is the clearest broad regression report in the set: a v4.57\u2192v5.2/5.3 checkpoint reload break with obvious user impact and a theme that matches several other version-migration/loading complaints in the cluster.", + "summary": "This cluster is a loose mix of Transformers v5 regression reports, import/load failures, and a few unrelated docs/test/model-specific bugs. I do not see any strong duplicate pairs; all soft edges look like separate underlying issues.", + "confidence": 0.42, + "canonical_issue_reason": "Issue 44556 is the broadest, most representative loading-regression report in the set: a v4.57-to-v5.x checkpoint reload breakage. It fits the common compatibility-break theme better than the narrower import, model-specific, test, or docs issues.", "canonical_pr_reason": null, - "best_issue_reason": "44556 is the most representative user-facing issue here because it cleanly captures the cluster\u2019s strongest recurring theme: upgrade-to-v5 breakage in model loading/reloading.", + "best_issue_reason": "44556 is the best single issue to anchor this cluster because it captures the general v5 upgrade/loading incompatibility pattern without being tied to one niche model or auxiliary symptom.", "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:43618", + "right": "issue:44977", + "accept": false, + "reason": "Different model-level bugs: CLIP attentions assignment vs Qwen3.5 flash-attention generation." + }, + { + "left": "issue:42890", + "right": "issue:44964", + "accept": false, + "reason": "Both involve loading, but one is a SAM test flake and the other is a Phi-4 multimodal model-load failure; different code paths." + }, + { + "left": "issue:43502", + "right": "issue:44062", + "accept": false, + "reason": "Unrelated failures: local_files_only still making requests vs AddedToken keyword duplication." + }, + { + "left": "issue:42890", + "right": "issue:44112", + "accept": false, + "reason": "A seeding-related SAM CI flake is not the same as a GraniteSpeech stale device override failure." + }, + { + "left": "issue:42890", + "right": "issue:45600", + "accept": false, + "reason": "Different areas: SAM integration test flakiness vs broken auto_mappings references." + }, + { + "left": "issue:43576", + "right": "issue:46132", + "accept": false, + "reason": "Different subsystems: v5 env CLI behavior vs AttentionInterface.register changing callbacks." + }, + { + "left": "issue:43723", + "right": "issue:44556", + "accept": false, + "reason": "Both are loading regressions, but one is tokenizer loading and the other is checkpoint reload compatibility; not the same bug." + }, { "left": "issue:44387", "right": "issue:44556", "accept": false, - "reason": "Both mention v5 regressions, but one is an int4 CUDA memory/OOM problem and the other is checkpoint reload failure; different code paths and symptoms." + "reason": "Int4 memory/OOM regression is a different issue from checkpoint reload failure." }, { "left": "issue:42890", "right": "issue:45341", "accept": false, - "reason": "Both touch testing, but one is an integration-test seeding flake and the other is a generic testing_utils bug; not the same defect." + "reason": "Both touch tests, but this is SAM test seeding flakiness vs a generic testing_utils bug." }, { "left": "issue:44222", "right": "issue:45125", "accept": false, - "reason": "One is FP8 save_pretrained behavior, the other is missing _tp_plan for tensor parallelism; related model family, different bug." + "reason": "Serialization bug in FP8 save_pretrained is not the same as missing tensor-parallel plan metadata." }, { "left": "issue:43010", "right": "issue:43065", "accept": false, - "reason": "Cache no_grad decoration and a dummy Conv2d in Sam3PixelDecoder are unrelated implementation issues." + "reason": "No common underlying bug: no_grad decoration for cache updates vs dummy Conv2d in Sam3PixelDecoder." }, { "left": "issue:42548", "right": "issue:44351", "accept": false, - "reason": "Both are import errors, but they involve different missing symbols and different breakages." + "reason": "Different missing imports: PreTrainedModel vs HybridCache." }, { - "left": "issue:43502", - "right": "issue:44485", + "left": "issue:44355", + "right": "issue:44749", "accept": false, - "reason": "local_files_only network requests and GLM-5 RoPE implementation are unrelated problems." + "reason": "Compiled-file errors and a Chinese report about preprocessing slowdown are unrelated." }, { - "left": "issue:44355", - "right": "issue:44749", + "left": "issue:43502", + "right": "issue:44485", "accept": false, - "reason": "Compiled Python file errors and slow filtering after upgrade are different issues with no shared code path." + "reason": "Network requests despite local_files_only is unrelated to GLM-5 RoPE implementation." }, { "left": "issue:45125", "right": "issue:45478", "accept": false, - "reason": "Both involve Qwen3.5 MoE, but one is tensor-parallel metadata and the other is a from_pretrained load error; too different to merge." + "reason": "Tensor-parallel plan metadata and Qwen3.5 Moe loading failure are related only by model family, not the same concrete bug." }, { "left": "issue:30333", "right": "issue:33453", "accept": false, - "reason": "MLFlow reporting on training failure and tokenizer loading regression are unrelated." + "reason": "MLFlow failure reporting and tokenizer loading regression are unrelated." }, { "left": "issue:43010", "right": "issue:43232", "accept": false, - "reason": "Both are generation/cache-related, but one is no_grad decoration and the other is sync_gpus kwargs handling; not the same bug." + "reason": "Both generation-adjacent, but one is a no_grad decoration request and the other is a sync_gpus generation state bug." }, { "left": "issue:45125", "right": "issue:45375", "accept": false, - "reason": "Missing _tp_plan and a config field dropped by @strict are different Qwen3.5 MoE issues." + "reason": "Missing _tp_plan and missing deepstack_visual_indexes are separate model config issues." }, { "left": "issue:30990", "right": "issue:31795", "accept": false, - "reason": "Sentence Transformers loading hangs and docs confusion about model.forward arguments are not the same underlying problem." + "reason": "A loading hang in Sentence Transformers is not the same as a forward-docs clarification issue." }, { "left": "issue:43502", "right": "issue:44222", "accept": false, - "reason": "Network requests despite local_files_only and FP8 save_pretrained moe are unrelated." + "reason": "local_files_only network leakage and FP8 save_pretrained moe are different bugs." }, { "left": "issue:43299", "right": "issue:44556", "accept": false, - "reason": "Both are loading regressions, but they target different models and different failure causes; not one mergeable fix." + "reason": "Both are version-related loading regressions, but they affect different model families and failure modes." }, { "left": "issue:42831", "right": "issue:43874", "accept": false, - "reason": "FP8 accuracy loss and a missing image-patch-count method are unrelated." + "reason": "FP8 accuracy regression is unrelated to the GLM46V image-patch counting AttributeError." }, { "left": "issue:43502", "right": "issue:44961", "accept": false, - "reason": "A local_files_only network leak and a placeholder 'racoon' issue are clearly unrelated." + "reason": "local_files_only behavior and the racoon issue are unrelated." }, { "left": "issue:44315", "right": "issue:44360", "accept": false, - "reason": "Liger kernel application under model_init and a DSA indexer ReLU discussion are unrelated." + "reason": "Liger Kernel application under model_init is unrelated to the DSA indexer ReLU discussion." }, { "left": "issue:43827", "right": "issue:44961", "accept": false, - "reason": "Docs still referencing pipeline() and the 'racoon' issue do not describe the same bug." + "reason": "Docs still referencing pipeline() after removals is unrelated to racoon." }, { "left": "issue:30333", "right": "issue:36296", "accept": false, - "reason": "MLFlow failure reporting and tensor-parallel training bugs are different subsystems and problems." + "reason": "MLFlow reporting on failed training jobs is unrelated to tensor-parallel training." }, { "left": "issue:44355", "right": "issue:45071", "accept": false, - "reason": "Compiled Python file execution errors and PretrainedConfig type checking are unrelated." + "reason": "Compiled Python file errors are not the same as PretrainedConfig type-checking breakage." }, { "left": "issue:44355", "right": "issue:44655", "accept": false, - "reason": "Compiled Python file errors and save_pretrained failure for Pipeline objects are different issues." + "reason": "Compiled-file errors and Pipeline save_pretrained failure are different problems." }, { "left": "issue:33453", "right": "issue:34689", "accept": false, - "reason": "Tokenizer loading regression and Llama 3.2 vision model loading break are both load-related, but they are distinct regressions." + "reason": "Tokenizer loading regression and Llama 3.2 90B Vision model loading breakage are different concrete failures." }, { "left": "issue:42757", "right": "issue:43576", "accept": false, - "reason": "Missing huggingface_hub export and a broken transformers env command are separate API/CLI issues." + "reason": "Missing is_offline_mode import and broken transformers env command are unrelated." }, { "left": "issue:44222", "right": "issue:44230", "accept": false, - "reason": "FP8 save_pretrained moe and FP8 inference support are different paths and would not be one PR." + "reason": "FP8 save_pretrained moe vs FP8 inference support for Qwen3.5 VL are distinct issues." }, { "left": "issue:43576", "right": "issue:45071", "accept": false, - "reason": "Broken env command and PretrainedConfig type checking are unrelated." + "reason": "Broken env CLI and PretrainedConfig type checking are unrelated." }, { "left": "issue:33453", "right": "issue:36296", "accept": false, - "reason": "Tokenizer loading regression and tensor-parallel training bug do not share the same underlying defect." + "reason": "Tokenizer loading regression and tensor-parallel training bug are different issues." }, { "left": "issue:43576", "right": "issue:43824", "accept": false, - "reason": "Broken env CLI command and missing Qwen2_5_VLForConditionalGeneration import are unrelated." + "reason": "CLI env command issues are unrelated to the Qwen2.5 VL import error." }, { "left": "issue:31795", "right": "issue:38617", "accept": false, - "reason": "Documentation confusion and a missing layer_type_validation import are different kinds of issues." + "reason": "Documentation confusion and an ImportError for layer_type_validation are unrelated." }, { "left": "issue:44961", "right": "issue:45468", "accept": false, - "reason": "The 'racoon' issue is not meaningfully related to Gemma-4 audio positional encoding." + "reason": "racoon is not related to Gemma-4 audio positional encoding." }, { "left": "issue:41628", "right": "issue:42548", "accept": false, - "reason": "AutoImageProcessor and PreTrainedModel import errors are different missing exports." + "reason": "AutoImageProcessor import failure and PreTrainedModel import failure are different symbols and fixes." }, { "left": "issue:38617", "right": "issue:43576", "accept": false, - "reason": "Missing layer_type_validation import and broken env CLI are unrelated v5 breakages." + "reason": "ImportError for layer_type_validation is unrelated to the transformers env command regression." }, { "left": "issue:44355", "right": "issue:46178", "accept": false, - "reason": "Compiled-file execution errors and an unknown backend in BACKENDS_MAPPING are different failures." - }, - { - "left": "issue:46143", - "right": "issue:46144", - "accept": false, - "reason": "Both concern RoFormer, but one is kwargs passthrough and the other is attention-interface usage; not the same code-path bug." - }, - { - "left": "issue:44908", - "right": "issue:44961", - "accept": false, - "reason": "Scheduler kwargs handling and the unrelated 'racoon' issue are not duplicates." - }, - { - "left": "issue:44485", - "right": "issue:44961", - "accept": false, - "reason": "GLM-5 RoPE implementation and the 'racoon' issue are unrelated." - }, - { - "left": "issue:43673", - "right": "issue:44492", - "accept": false, - "reason": "Missing cache during chunked_prefill and a typo in cache strategies are different problems." - }, - { - "left": "issue:36296", - "right": "issue:43901", - "accept": false, - "reason": "Tensor-parallel training bugs and TextClassificationPipeline docs mismatch are unrelated." - }, - { - "left": "issue:36296", - "right": "issue:43526", - "accept": false, - "reason": "Tensor-parallel training and BeitImageProcessorFast reduce_labels behavior are unrelated." + "reason": "Compiled-file errors are unrelated to the tensorflow_text backend mapping ValueError." } ] }, @@ -15153,26 +15229,27 @@ "component_index": 1, "component_count": 17, "review_unit_index": 48, - "review_unit_count": 49, + "review_unit_count": 50, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-427", + "cluster_id": "cluster-10000-428", "nodes": [ - "issue:31795", - "issue:33453", + "issue:36296", "issue:41669", "issue:42371", "issue:42890", "issue:43352", "issue:43408", "issue:43525", + "issue:43526", "issue:43572", "issue:43576", "issue:43673", "issue:43827", + "issue:43901", "issue:43976", "issue:44038", "issue:44077", @@ -15183,34 +15260,39 @@ "issue:44297", "issue:44315", "issue:44351", - "issue:44360", "issue:44393", + "issue:44485", "issue:44492", "issue:44556", "issue:44623", - "issue:44655", "issue:44704", - "issue:44829", "issue:44861", "issue:44908", "issue:44945", + "issue:44961", "issue:45003", "issue:45071", "issue:45092", "issue:45125", "issue:45230", "issue:45278", - "issue:45310", "issue:45341", "issue:45362", - "issue:45468", "issue:45478", "issue:45542", "issue:45753", "issue:45850", - "issue:46095" + "issue:46095", + "issue:46143", + "issue:46144" ], "soft_pairs": [ + "issue:46143|issue:46144", + "issue:44908|issue:44961", + "issue:44485|issue:44961", + "issue:43673|issue:44492", + "issue:36296|issue:43901", + "issue:36296|issue:43526", "issue:45362|issue:45850", "issue:44623|issue:45230", "issue:44273|issue:45542", @@ -15219,9 +15301,9 @@ "issue:44393|issue:45071", "issue:44230|issue:46095", "issue:44393|issue:45753", - "issue:43576|issue:44556", "issue:41669|issue:43673", "issue:43525|issue:45850", + "issue:43576|issue:44556", "issue:44556|issue:45092", "issue:43572|issue:45850", "issue:44492|issue:45850", @@ -15234,8 +15316,8 @@ "issue:43976|issue:44393", "issue:43352|issue:44393", "issue:44297|issue:45230", - "issue:44295|issue:45230", "issue:43576|issue:44908", + "issue:44295|issue:45230", "issue:44908|issue:45071", "issue:41669|issue:44351", "issue:41669|issue:45341", @@ -15244,286 +15326,283 @@ "issue:44393|issue:45478", "issue:44077|issue:44861", "issue:43827|issue:44908", - "issue:41669|issue:44908", - "issue:31795|issue:33453", - "issue:44315|issue:44829", - "issue:43576|issue:44273", - "issue:44360|issue:45468", - "issue:44655|issue:44861", - "issue:45230|issue:45310" + "issue:41669|issue:44908" ], - "prepared_review_unit_hash": "fe6670131ae20b2a21551f7ec13f0cc20c0148ea1d5ffc4cf1410023f4538735", + "prepared_review_unit_hash": "aa1aee864624b8b3559eba34a87729c8a81a1ff9a19a098123ba0a032d89d730", "component_budget": { - "node_count": 427, - "item_count": 427, - "soft_pair_count": 1576, - "serialized_chars": 816182, - "estimated_input_tokens": 204046, - "estimated_eval_tokens": 408348 + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 40, - "serialized_chars": 39071, - "estimated_input_tokens": 9768, - "estimated_eval_tokens": 19792 + "serialized_chars": 38998, + "estimated_input_tokens": 9750, + "estimated_eval_tokens": 19756 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly a mix of unrelated v5 regressions, with two real duplicate groups: an open import/lazy-loading issue set and a checkpoint reload/meta-init compatibility set. The checkpoint pair is a credible duplicate; the import-slowdown vs lazy-loading pair is not strong enough to merge.", - "confidence": 0.77, - "canonical_issue_reason": "issue:44273 is the best open, actionable umbrella for the lazy-loading subgroup; it is the clearest live duplicate target in the cluster.", + "summary": "The cluster is noisy and mostly contains unrelated issues; the only clearly related subtheme is import/lazy-loading performance regressions, but the soft edges are too weak to merge. I did not find any PRs in the input.", + "confidence": 0.58, + "canonical_issue_reason": "issue:44246 is the broadest, most representative report of the import/performance subtheme and has the clearest user-facing framing; it reads like the umbrella issue for the cluster.", "canonical_pr_reason": null, - "best_issue_reason": "issue:44273 is the strongest global issue target because it is open and tracks a concrete import/lazy-loading regression rather than a vague version-wide complaint.", + "best_issue_reason": "issue:44246 is the best single representative issue because it captures the main recurring symptom in this set: slow or problematic imports around lazy loading.", "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:46143", + "right": "issue:46144", + "accept": false, + "reason": "Same RoFormer area, but one is kwargs propagation and the other is attention-interface usage; related but not the same bug." + }, + { + "left": "issue:44908", + "right": "issue:44961", + "accept": false, + "reason": "Scheduler kwargs vs an unrelated placeholder issue title; no shared underlying defect." + }, + { + "left": "issue:44485", + "right": "issue:44961", + "accept": false, + "reason": "RoPE implementation discussion is unrelated to the placeholder issue." + }, + { + "left": "issue:43673", + "right": "issue:44492", + "accept": false, + "reason": "Chunked prefill cache regression and a cache-strategy typo are different cache problems." + }, + { + "left": "issue:36296", + "right": "issue:43901", + "accept": false, + "reason": "Tensor parallel training and TextClassificationPipeline docs are unrelated." + }, + { + "left": "issue:36296", + "right": "issue:43526", + "accept": false, + "reason": "Tensor parallel training and BeitImageProcessorFast label reduction are unrelated defects." + }, { "left": "issue:45362", "right": "issue:45850", "accept": false, - "reason": "Both mention Qwen3.5, but one is a specific chat crash and the other is a broad claim that v5 breaks many models; not the same bug." + "reason": "Both mention model breakage, but one is a specific Qwen3.5 chat crash and the other is a broad complaint." }, { "left": "issue:44623", "right": "issue:45230", "accept": false, - "reason": "A specific processor.save_pretrained file-loss bug vs. an unspecific bug report; no clear shared code path." + "reason": "Processor save_pretrained missing files is not the same as a generic bug report." }, { "left": "issue:44273", "right": "issue:45542", "accept": false, - "reason": "Lazy-loading/import behavior and TensorFlow backend installation issues are different failures." + "reason": "Lazy loading issues and a TensorFlow backend import error are different failure modes." }, { "left": "issue:44038", "right": "issue:45125", "accept": false, - "reason": "Both involve Qwen3 MoE, but they describe different concrete problems: model support/compatibility vs. missing tensor-parallel plan." + "reason": "Qwen3-VL-Moe support bug and missing _tp_plan for tensor parallelism are distinct model issues." }, { "left": "issue:44273", "right": "issue:45003", "accept": false, - "reason": "Import/lazy-loading regression is not the same as unsafe sys.modules access in modeling_utils." + "reason": "Both touch import mechanics, but sys.modules access and lazy loading are not clearly the same bug." }, { "left": "issue:44393", "right": "issue:45071", "accept": false, - "reason": "Qwen3-VL output/bbox error vs. PretrainedConfig type-checking regression; different layers and symptoms." + "reason": "Qwen3-VL 2D bbox errors and PretrainedConfig type checking breakage are unrelated." }, { "left": "issue:44230", "right": "issue:46095", "accept": false, - "reason": "FP8 inference support request vs. a weight-conversion script question; not the same underlying issue." + "reason": "FP8 inference support and a weight conversion-script question are not the same change or bug." }, { "left": "issue:44393", "right": "issue:45753", "accept": false, - "reason": "Both are Qwen3-VL reports, but one is bbox hallucination/error and the other is an infinite-loop case on a specific image." - }, - { - "left": "issue:43576", - "right": "issue:44556", - "accept": false, - "reason": "Broken transformers env command vs. checkpoint reload failure after upgrade; different commands and failure modes." + "reason": "Both involve Qwen3-VL models, but one is bbox output and the other is an infinite-loop image case." }, { "left": "issue:41669", "right": "issue:43673", "accept": false, - "reason": "Import-time slowdown from import * usage vs. chunked_prefill cache missing; too different in cause and symptom." + "reason": "Import-star performance regression and chunked-prefill cache regression are different code-path problems." }, { "left": "issue:43525", "right": "issue:45850", "accept": false, - "reason": "A specific Llama4Config attribute error is not the same as a broad v5 breakage complaint." + "reason": "A missing pad_token_id attribute and a broad v5 breakage complaint are not the same issue." + }, + { + "left": "issue:43576", + "right": "issue:44556", + "accept": false, + "reason": "Broken env command and checkpoint reload incompatibility are different upgrade regressions." }, { "left": "issue:44556", "right": "issue:45092", - "accept": true, - "reason": "Both report v5 upgrade incompatibility when loading older checkpoints; same underlying checkpoint-loading/meta-init regression." + "accept": false, + "reason": "Both are upgrade-related, but one is generic reload failure and the other is InternVL2 meta-init incompatibility." }, { "left": "issue:43572", "right": "issue:45850", "accept": false, - "reason": "Missing pad_token_idx in StableLmConfig is a specific model config bug, not the same as the broad v5 breakage report." + "reason": "Missing pad_token_idx in StableLmConfig is unrelated to the broad breakage report." }, { "left": "issue:44492", "right": "issue:45850", "accept": false, - "reason": "A typo in cache strategies is unrelated to a generic v5 breakage complaint." + "reason": "Cache-strategy typo is a narrow docs/code typo, not the same as a broad compatibility complaint." }, { "left": "issue:42371", "right": "issue:44492", "accept": false, - "reason": "TF32 API guidance and a cache-strategy typo are unrelated." + "reason": "TF32 settings guidance and cache-strategy typo are unrelated." }, { "left": "issue:45278", "right": "issue:45850", "accept": false, - "reason": "Both mention v5 regressions, but one is many import errors after upgrade and the other is a broad assertion that v5 breaks too many models; too generic to be duplicates." + "reason": "Many import errors after upgrade and a vague 'breaks too many models' complaint are too broad to call the same bug." }, { "left": "issue:44246", "right": "issue:44273", "accept": false, - "reason": "The titles only loosely overlap on import behavior; 44246 is an intermittent import slowdown report, while 44273 is a lazy-loading failure, so shared root cause is not established." + "reason": "Both are import/lazy-loading themed, but one is slow import performance and the other is a broader lazy-loading failure; not clearly the same underlying defect." }, { "left": "issue:42890", "right": "issue:44908", "accept": false, - "reason": "Test seeding flakiness vs. scheduler kwargs being ignored; unrelated bugs." + "reason": "Missing set_seed in SAM HQ tests and scheduler kwargs are unrelated." }, { "left": "issue:44704", "right": "issue:44908", "accept": false, - "reason": "AutoProcessor kwargs forwarding bug vs. inverse_sqrt scheduler kwargs bug; different APIs and code paths." + "reason": "AutoProcessor kwargs passing and scheduler kwargs are different APIs and bugs." }, { "left": "issue:43408", "right": "issue:44315", "accept": false, - "reason": "Model-type mismatch warning vs. Liger Kernel not applied with model_init; not the same issue." + "reason": "SAM3 model-type warning and Liger Kernel application via model_init are unrelated." }, { "left": "issue:43976", "right": "issue:44393", "accept": false, - "reason": "Python-version compatibility bug vs. Qwen3-VL output error; unrelated." + "reason": "Python-version compatibility and Qwen3-VL bbox errors are unrelated." }, { "left": "issue:43352", "right": "issue:44393", "accept": false, - "reason": "Flash Attention 2 support gap vs. Qwen3-VL bbox error; different problems." + "reason": "Flash Attention support for NemotronH and Qwen3-VL bbox errors are unrelated." }, { "left": "issue:44297", "right": "issue:45230", "accept": false, - "reason": "Tokenizer save_pretrained mismatch vs. an unspecific bug report; no shared concrete failure." + "reason": "Tokenizer class mismatch on save_pretrained is not the same as a generic bug report." + }, + { + "left": "issue:43576", + "right": "issue:44908", + "accept": false, + "reason": "Broken env command and scheduler kwargs are unrelated." }, { "left": "issue:44295", "right": "issue:45230", "accept": false, - "reason": "position_ids buffer access error vs. generic bug report; not enough evidence of the same bug." + "reason": "position_ids buffer access error is not the same as a generic bug report." }, { - "left": "issue:43576", - "right": "issue:44908", + "left": "issue:44908", + "right": "issue:45071", "accept": false, - "reason": "Broken env command vs. scheduler kwargs handling; unrelated." + "reason": "Scheduler kwargs handling and PretrainedConfig type checking are unrelated." }, { "left": "issue:41669", "right": "issue:44351", "accept": false, - "reason": "Import-time performance regression vs. missing HybridCache import; different import-path issues." + "reason": "Import-star performance and a HybridCache import error are not the same concrete bug." }, { "left": "issue:41669", "right": "issue:45341", "accept": false, - "reason": "Import-performance regression vs. a small testing utility bug; no clear overlap." + "reason": "Import performance regression and a testing_utils typo are unrelated." }, { "left": "issue:44492", "right": "issue:44945", "accept": false, - "reason": "Cache-strategy typo is unrelated to pipeline-parallelism output corruption." + "reason": "Cache strategy typo and incorrect pipeline-parallel output are different bugs." }, { - "left": "issue:43976", - "right": "issue:45850", - "accept": false, - "reason": "Python support regression vs. broad v5 breakage complaint; not the same bug." - }, - { - "left": "issue:44393", - "right": "issue:45478", - "accept": false, - "reason": "Different Qwen3.5 issues: bbox output error vs. from_pretrained failure." - }, - { - "left": "issue:44077", - "right": "issue:44861", - "accept": false, - "reason": "Optional post_init validation issue vs. AttributeError on tied-weight keys list; different code paths." - }, - { - "left": "issue:43827", - "right": "issue:44908", - "accept": false, - "reason": "Docs still referencing removed pipeline() vs. scheduler kwargs bug; unrelated." - }, - { - "left": "issue:41669", - "right": "issue:44908", - "accept": false, - "reason": "Import-time performance regression vs. scheduler kwargs handling; unrelated." - }, - { - "left": "issue:31795", - "right": "issue:33453", - "accept": false, - "reason": "Documentation confusion in model.forward vs. tokenizer loading regression; unrelated." - }, - { - "left": "issue:44315", - "right": "issue:44829", + "left": "issue:43976", + "right": "issue:45850", "accept": false, - "reason": "Liger Kernel application with model_init vs. flash_attention_3 causing degenerate training; different failures." + "reason": "Python-version incompatibility and broad model breakage are too different to merge." }, { - "left": "issue:43576", - "right": "issue:44273", + "left": "issue:44393", + "right": "issue:45478", "accept": false, - "reason": "Broken env command vs. lazy-loading regression; different user-facing failures." + "reason": "Different Qwen3.5 model failures with different symptoms and likely different code paths." }, { - "left": "issue:44360", - "right": "issue:45468", + "left": "issue:44077", + "right": "issue:44861", "accept": false, - "reason": "DSA indexer discussion vs. Gemma-4 audio positional encoding bug; unrelated." + "reason": "Optional post_init policy and tied-weight-key crash are unrelated." }, { - "left": "issue:44655", - "right": "issue:44861", + "left": "issue:43827", + "right": "issue:44908", "accept": false, - "reason": "Pipeline save_pretrained failure vs. tied-weight key AttributeError; not the same bug." + "reason": "Docs still mentioning pipeline() and scheduler kwargs are unrelated." }, { - "left": "issue:45230", - "right": "issue:45310", + "left": "issue:41669", + "right": "issue:44908", "accept": false, - "reason": "Generic bug report vs. specific Qwen3.5 MoE from_pretrained failure; no duplicate evidence." + "reason": "Import-star slowdown and inverse_sqrt scheduler kwargs are unrelated." } ] }, - "evaluator_result": { - "accept": false, - "feedback": "The cluster-level summary is mostly grounded, but at least one soft-edge acceptance is too strong for the evidence provided. In particular, issue:44246 vs issue:44273 is only supported by broad import-related similarity in the titles; 44246 is an intermittent import-slowdown report, while 44273 is a lazy-loading failure, so the shared root cause is not established. The accepted checkpoint pair (44556/45092) is more plausible, but the overall soft-edge handling is not consistently conservative enough for acceptance." - }, - "evaluator_used": true, - "retried": true, - "accepted_nontrivial_soft_edge": true, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, "error_kind": null, "error_message": null }, @@ -15532,13 +15611,13 @@ "component_index": 1, "component_count": 17, "review_unit_index": 49, - "review_unit_count": 49, + "review_unit_count": 50, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-427", + "cluster_id": "cluster-10000-428", "nodes": [ "issue:30333", "issue:31515", @@ -15547,38 +15626,38 @@ "issue:34689", "issue:36296", "issue:37428", - "issue:37934", "issue:38617", "issue:41669", "issue:42371", "issue:42913", "issue:42936", "issue:43066", - "issue:43240", "issue:43298", "issue:43502", "issue:43519", "issue:43525", "issue:43572", + "issue:43576", "issue:43606", "issue:43618", - "issue:43668", "issue:43704", "issue:43717", - "issue:43813", - "issue:44031", "issue:44230", "issue:44261", "issue:44263", "issue:44273", "issue:44315", + "issue:44360", "issue:44485", "issue:44492", "issue:44623", + "issue:44655", + "issue:44829", "issue:44861", "issue:44908", "issue:44998", "issue:45092", + "issue:45230", "issue:45290", "issue:45310", "issue:45341", @@ -15590,16 +15669,22 @@ "issue:46172" ], "soft_pairs": [ - "issue:46018|issue:46172", + "issue:31795|issue:33453", + "issue:44315|issue:44829", + "issue:43576|issue:44273", + "issue:44360|issue:45468", + "issue:44655|issue:44861", + "issue:45230|issue:45310", "issue:44273|issue:44908", + "issue:46018|issue:46172", "issue:43519|issue:44485", "issue:43519|issue:44908", "issue:41669|issue:43704", "issue:31795|issue:34689", "issue:31795|issue:37428", "issue:44315|issue:45092", - "issue:44485|issue:45468", "issue:43704|issue:44485", + "issue:44485|issue:45468", "issue:43298|issue:46095", "issue:42371|issue:43704", "issue:43502|issue:43519", @@ -15626,328 +15711,425 @@ "issue:31515|issue:33453", "issue:43298|issue:44230", "issue:43717|issue:46144", - "issue:42936|issue:46018", - "issue:43668|issue:44031", - "issue:42936|issue:43240", - "issue:43240|issue:46144", - "issue:42936|issue:44031", - "issue:42936|issue:46144", - "issue:42936|issue:43813", - "issue:43240|issue:43668", - "issue:43668|issue:43813", - "issue:37934|issue:43668", - "issue:42936|issue:43668" + "issue:42936|issue:46018" ], - "prepared_review_unit_hash": "67b496fa8fe554d284d6fb823784b76413453c1d3dc8169e9cf73511bb7196e2", + "prepared_review_unit_hash": "f72ddff805dccf4641ee58a32d7a817595d4739fc79e571f5f1a3438292d9ddd", "component_budget": { - "node_count": 427, - "item_count": 427, - "soft_pair_count": 1576, - "serialized_chars": 816182, - "estimated_input_tokens": 204046, - "estimated_eval_tokens": 408348 + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 47, - "serialized_chars": 40971, - "estimated_input_tokens": 10243, - "estimated_eval_tokens": 20742 + "soft_pair_count": 43, + "serialized_chars": 39838, + "estimated_input_tokens": 9960, + "estimated_eval_tokens": 20176 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "No soft pair looks like a safe duplicate. The candidates mostly share broad themes like Transformers v5 regressions, model loading, or tokenizer/processor behavior, but they point to different code paths, different model families, or different bug classes.", - "confidence": 0.84, - "canonical_issue_reason": "issue:45092 is the clearest concrete regression in the set: it describes a specific Transformers v5 meta-initialization incompatibility for old InternVL2 checkpoints with a reproducible loading path.", + "summary": "This cluster is dominated by Transformer v5 regression reports, especially model/tokenizer loading and configuration/API breakages. The broad umbrella issue about v5 breaking many models is the best representative; none of the soft pairs look like the same concrete bug/change.", + "confidence": 0.77, + "canonical_issue_reason": "issue:45850 is the broadest and most cluster-representative report: it explicitly frames a v5-wide breakage affecting many models, which matches the surrounding loading/config/import regression issues.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45092 is the most actionable representative issue here because it is narrowly scoped, technically specific, and tied to a concrete failing model-loading path.", + "best_issue_reason": "issue:45850 best captures the overall theme of the cluster as a general v5 regression umbrella rather than a narrow model-specific bug.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:46018", - "right": "issue:46172", + "left": "issue:31795", + "right": "issue:33453", + "accept": false, + "reason": "Both touch model/tokenizer docs vs tokenizer loading regressions, but one is a documentation confusion issue and the other is a runtime loading regression; not the same underlying bug." + }, + { + "left": "issue:44315", + "right": "issue:44829", + "accept": false, + "reason": "Liger kernel application via model_init is unrelated to flash_attention_3 causing degenerate training; different code paths and failure modes." + }, + { + "left": "issue:43576", + "right": "issue:44273", + "accept": false, + "reason": "Broken `transformers env` CLI and lazy loading malfunction are separate features with no shared concrete bug." + }, + { + "left": "issue:44360", + "right": "issue:45468", + "accept": false, + "reason": "DSA indexer ReLU concern and Gemma-4 audio positional encoding are different model internals; no shared fix target." + }, + { + "left": "issue:44655", + "right": "issue:44861", + "accept": false, + "reason": "Pipeline `save_pretrained` failure and `_get_tied_weight_keys` crashing on list input are separate save/load helper bugs." + }, + { + "left": "issue:45230", + "right": "issue:45310", "accept": false, - "reason": "Different bugs: shared-expert gating logic vs a tensor conversion/transposition issue in deepseek-v4." + "reason": "Generic bug report and Qwen3.5 MoE `from_pretrained` error are too vague vs specific to be the same issue." }, { "left": "issue:44273", "right": "issue:44908", "accept": false, - "reason": "Lazy-loading behavior and inverse_sqrt scheduler kwargs are unrelated subsystems." + "reason": "Lazy loading failure and inverse_sqrt scheduler kwargs being ignored are unrelated subsystems." + }, + { + "left": "issue:46018", + "right": "issue:46172", + "accept": false, + "reason": "DeepSeek-V4 gating logic and a tensor transpose bug in `comb.to(dtype)` are different implementation defects." }, { "left": "issue:43519", "right": "issue:44485", "accept": false, - "reason": "Different model-specific failures: Qwen3VL timestamp calculation vs GLM-5 RoPE implementation." + "reason": "Qwen3VL timestamp calculation and GLM-5 RoPE implementation are separate model-specific math bugs." }, { "left": "issue:43519", "right": "issue:44908", "accept": false, - "reason": "Processor timestamp math and scheduler keyword handling are not the same bug." + "reason": "Processor timestamp math and scheduler kwargs handling are unrelated." }, { "left": "issue:41669", "right": "issue:43704", "accept": false, - "reason": "Import-star performance issue vs VRAM leak in multi-threaded use; unrelated symptoms and code paths." + "reason": "Import-star performance issue and Qwen3ForCausalLM VRAM leak are different problems despite both being runtime regressions." }, { "left": "issue:31795", "right": "issue:34689", "accept": false, - "reason": "Documentation confusion about forward args is not the same as a tokenizer/model loading regression." + "reason": "Documentation confusion around inputs/past_key_values is not the same as a Llama 3.2 model loading regression." }, { "left": "issue:31795", "right": "issue:37428", "accept": false, - "reason": "Doc clarity issue vs ImportError from a missing symbol; different problem classes." + "reason": "Forward-argument docs and a flash-attention import error are unrelated." }, { "left": "issue:44315", "right": "issue:45092", "accept": false, - "reason": "Applying Liger Kernel in model_init and meta-init incompatibility for old checkpoints are separate loading concerns." + "reason": "Applying Liger Kernel in model_init and meta-initialization incompatibility for old InternVL2 checkpoints are different integration bugs." }, { - "left": "issue:44485", - "right": "issue:45468", + "left": "issue:43704", + "right": "issue:44485", "accept": false, - "reason": "GLM-5 RoPE implementation and Gemma-4 audio positional encoding are different model implementations." + "reason": "VRAM leakage across dataloader threads is unrelated to RoPE implementation correctness." }, { - "left": "issue:43704", - "right": "issue:44485", + "left": "issue:44485", + "right": "issue:45468", "accept": false, - "reason": "VRAM leak under threaded loading is unrelated to a RoPE implementation bug." + "reason": "GLM-5 RoPE math and Gemma-4 audio positional encoding are different model-specific position-encoding issues." }, { "left": "issue:43298", "right": "issue:46095", "accept": false, - "reason": "Unsupported audio continuous batching and a request for a weight conversion script are not the same change or bug." + "reason": "Continuous batching not supporting audio models and a DeepSeek-V4 weight conversion question are not the same bug." }, { "left": "issue:42371", "right": "issue:43704", "accept": false, - "reason": "TF32 API migration guidance is unrelated to a VRAM leak in Qwen3ForCausalLM." + "reason": "TF32 API guidance and a VRAM leak do not share the same concrete code path." }, { "left": "issue:43502", "right": "issue:43519", "accept": false, - "reason": "Local-files-only network requests and Qwen3VL timestamp calculation are different bugs." + "reason": "Accidental API requests despite `local_files_only=True` and incorrect Qwen3VL timestamp computation are unrelated." }, { "left": "issue:43519", "right": "issue:44623", "accept": false, - "reason": "Timestamp calculation in a processor and missing files in save_pretrained do not share the same failure mode." + "reason": "Timestamp calculation bug and missing files from `processor.save_pretrained(...)` are separate processor issues." }, { "left": "issue:43519", "right": "issue:43704", "accept": false, - "reason": "Processor timestamp math and threaded VRAM leakage are unrelated." + "reason": "Qwen3VL timestamp math and Qwen3ForCausalLM thread-related VRAM leak are different bugs." }, { "left": "issue:43606", "right": "issue:45468", "accept": false, - "reason": "CPU offload device mismatch for bark-small is not the same as Gemma-4 audio positional encoding." + "reason": "CPU offload device mismatch for bark-small and Gemma-4 audio positional encoding are both audio-related but not the same failure." }, { "left": "issue:44263", "right": "issue:45468", "accept": false, - "reason": "torch.split return-value handling and audio positional encoding are separate code paths." + "reason": "`torch.split` return handling in GlmMoeDsaIndexer is unrelated to Gemma-4 audio positional encoding." }, { "left": "issue:44261", "right": "issue:45468", "accept": false, - "reason": "MLA q_a_layernorm epsilon precision and Gemma-4 audio positional encoding are different model bugs." + "reason": "MLA layernorm epsilon precision issue and Gemma-4 audio positional encoding are distinct model math bugs." }, { "left": "issue:42913", "right": "issue:43066", "accept": false, - "reason": "Both are tokenizer-related v5 regressions, but one is a broad behavior difference and the other is a specific decoder-type bug; not the same concrete issue." + "reason": "Tokenizer behavior change from v4 to v5 and wrong tokenizer decoder type in v5 are both tokenizer-related but not clearly the same concrete defect." }, { "left": "issue:44908", "right": "issue:44998", "accept": false, - "reason": "A scheduler kwargs bug does not match the placeholder/non-bug issue." + "reason": "Scheduler kwargs ignored and a non-technical issue ('Unemployment') are obviously unrelated." }, { "left": "issue:44492", "right": "issue:44998", "accept": false, - "reason": "Cache-strategy typo and an unrelated placeholder issue are not duplicates." + "reason": "Cache strategies typo and a non-technical issue are unrelated." }, { "left": "issue:44861", "right": "issue:45290", "accept": false, - "reason": "Tied-weights key handling and chat-template tool-call crashes are different failures." + "reason": "Tied-weight key handling and chat template crashes with tool calls are different code paths." }, { "left": "issue:43618", "right": "issue:44998", "accept": false, - "reason": "CLIPOutput attentions regression and an unrelated placeholder issue are not the same bug." + "reason": "CLIPOutput attentions regression and a non-technical issue are unrelated." }, { "left": "issue:44998", "right": "issue:45850", "accept": false, - "reason": "A placeholder title is not the same as the broad complaint that v5 breaks many models." + "reason": "A non-technical issue cannot be merged with a broad v5 regression tracker." }, { "left": "issue:44998", "right": "issue:45341", "accept": false, - "reason": "Unemployment/placeholder content does not match a testing_utils bug." + "reason": "A non-technical issue and a small testing_utils bug are unrelated." }, { "left": "issue:43525", "right": "issue:44861", "accept": false, - "reason": "Missing pad_token_id in Llama4Config and tied-weight key crashes are separate config/model issues." + "reason": "Missing `pad_token_id` on `Llama4Config` and tied-weight key crash are different attribute/utility problems." }, { "left": "issue:43572", "right": "issue:44861", "accept": false, - "reason": "Missing pad_token_idx in StableLmConfig is unrelated to tied-weight key handling." + "reason": "Missing `pad_token_idx` in `StableLmConfig` is a config-API regression, not the tied-weight list handling crash." }, { "left": "issue:31515", "right": "issue:38617", "accept": false, - "reason": "Slow checkpoint loading and an ImportError from configuration_utils are different issues." + "reason": "Slow checkpoint loading and missing `layer_type_validation` import are both loading-related in a broad sense, but not the same bug or fix." }, { "left": "issue:44230", "right": "issue:44623", "accept": false, - "reason": "FP8 support request and missing processor save files are unrelated." + "reason": "FP8 support request for Qwen3-VL/MoE and missing files on processor save are unrelated." }, { "left": "issue:30333", "right": "issue:31515", "accept": false, - "reason": "MLFlow job reporting after failure is unrelated to from_pretrained checkpoint loading speed." + "reason": "MLFlow reporting on failed training jobs and slow checkpoint loading are completely different concerns." }, { "left": "issue:31515", "right": "issue:36296", "accept": false, - "reason": "Checkpoint loading performance and tensor-parallel training behavior are different bugs." + "reason": "Slow `from_pretrained` checkpoint loading and tensor parallel training bugs are different subsystems and failure modes." }, { "left": "issue:31515", "right": "issue:31795", "accept": false, - "reason": "Slow checkpoint loading is not the same as confusing forward-argument documentation." + "reason": "Checkpoint load slowness and forward doc confusion are not the same issue." }, { "left": "issue:44230", "right": "issue:45310", "accept": false, - "reason": "FP8 inference support request and a Qwen3.5 MoE from_pretrained error are different concerns." + "reason": "FP8 support request and Qwen3.5 MoE `from_pretrained` error are not the same bug." }, { "left": "issue:31515", "right": "issue:33453", "accept": false, - "reason": "Loading speed and tokenizer loading regression are not the same failure." + "reason": "Slow checkpoint loading and tokenizer loading regression are both load-time issues, but they are not the same concrete defect." }, { "left": "issue:43298", "right": "issue:44230", "accept": false, - "reason": "Audio continuous batching support and FP8 inference support are separate feature requests." + "reason": "Continuous batching audio-model support and FP8 inference support are different feature gaps." }, { "left": "issue:43717", "right": "issue:46144", "accept": false, - "reason": "Mamba init_weights usage and RoFormer attention interface are different implementation issues." + "reason": "Mamba init-weights usage and RoFormer attention interface usage are unrelated architecture implementation issues." }, { "left": "issue:42936", "right": "issue:46018", "accept": false, - "reason": "Mask2Former ignore_value handling and DeepSeek-V4 shared-expert gating are unrelated models and bugs." - }, + "reason": "Mask2Former ignore_value misuse and DeepSeek-V4 gating logic are distinct model bugs." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 17, + "review_unit_index": 50, + "review_unit_count": 50, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-428", + "nodes": [ + "issue:37934", + "issue:42936", + "issue:43240", + "issue:43668", + "issue:43813", + "issue:44031", + "issue:46144" + ], + "soft_pairs": [ + "issue:43668|issue:44031", + "issue:42936|issue:43240", + "issue:43240|issue:46144", + "issue:42936|issue:44031", + "issue:42936|issue:46144", + "issue:42936|issue:43813", + "issue:43240|issue:43668", + "issue:43668|issue:43813", + "issue:37934|issue:43668", + "issue:42936|issue:43668" + ], + "prepared_review_unit_hash": "ac5bc3e5a5b27972796756213c96bec1d52767ac4a255a2677055e4dd84054da", + "component_budget": { + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 + }, + "budget": { + "node_count": 7, + "item_count": 7, + "soft_pair_count": 10, + "serialized_chars": 6938, + "estimated_input_tokens": 1735, + "estimated_eval_tokens": 3726 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "These issues are only loosely similar at the wording level; they cover unrelated bugs across norms, loss plumbing, configuration typing, typo fixes, tokenizer warnings, and model attention implementation. I would not merge any of them as duplicates.", + "confidence": 0.97, + "canonical_issue_reason": "No clear canonical issue: the cluster is a set of distinct bug reports affecting different models and code paths, with no shared targets or concrete overlap.", + "canonical_pr_reason": null, + "best_issue_reason": "None of the issues stands out as the obvious duplicate hub; if forced, issue 43240 is the most broadly phrased, but it is still a distinct loss-passing bug rather than a representative superset.", + "best_pr_reason": null, + "soft_edge_verdicts": [ { "left": "issue:43668", "right": "issue:44031", "accept": false, - "reason": "Incorrect type hint and tokenizer regex warning regression are not the same issue." + "reason": "One is a ModernBERT type-hint bug; the other is a tokenizer regex warning regression. Different subsystems and failure modes." }, { "left": "issue:42936", "right": "issue:43240", "accept": false, - "reason": "ignore_value not used and kwargs not passed to loss calculation are different code paths." + "reason": "Mask2Former ignore_value handling and loss-kwargs forwarding are unrelated implementation bugs, even though both touch training code." }, { "left": "issue:43240", "right": "issue:46144", "accept": false, - "reason": "Loss-function kwargs handling and RoFormer attention interface are unrelated." + "reason": "Loss kwargs propagation and RoFormer attention interface usage affect different model paths and do not describe the same change." }, { "left": "issue:42936", "right": "issue:44031", "accept": false, - "reason": "Mask2Former ignore_value and tokenizer regex warnings do not share a concrete bug." + "reason": "Mask2Former ignore_value and tokenizer regex warnings are unrelated; no shared code path or fix." }, { "left": "issue:42936", "right": "issue:46144", "accept": false, - "reason": "Mask2Former ignore_value and RoFormer attention interface are unrelated." + "reason": "A segmentation loss-parameter bug is not the same as a RoFormer attention-interface bug." }, { "left": "issue:42936", "right": "issue:43813", "accept": false, - "reason": "A missing ignore_value use and a typo in quantization field naming are different issues." + "reason": "Ignore_value usage and a typo in a quantization attribute are different classes of issues with no concrete overlap." }, { "left": "issue:43240", "right": "issue:43668", "accept": false, - "reason": "Loss kwargs propagation and a ModernBERT type-hint bug are unrelated." + "reason": "Passing kwargs to loss calculation is unrelated to a ModernBERT config annotation problem." }, { "left": "issue:43668", "right": "issue:43813", "accept": false, - "reason": "Type-hint correctness and a quantization-operation typo are not the same bug." + "reason": "Both are small fixes, but they target different files and bugs: type hint vs typo." }, { "left": "issue:37934", "right": "issue:43668", "accept": false, - "reason": "A design question about RMS norm and an incorrect type hint are not duplicates." + "reason": "Llama4TextL2Norm/RMS norm question is unrelated to ModernBERT norm_eps typing." }, { "left": "issue:42936", "right": "issue:43668", "accept": false, - "reason": "Mask2Former ignore_value usage and ModernBERT norm_eps typing are unrelated." + "reason": "Different models and different bug types; no same underlying defect." } ] }, @@ -15980,12 +16162,12 @@ "issue:10519|issue:11005", "issue:11005|issue:11307" ], - "prepared_review_unit_hash": "5aba78bfc0ffbc64423f56d901a812967817508710faee8c1ec00e249d82570a", + "prepared_review_unit_hash": "de98ab6a0b469a88cefd82b7f8dbe4a49599f62e96fd510d40ae0cf65a31b8ab", "component_budget": { "node_count": 3, "item_count": 3, "soft_pair_count": 3, - "serialized_chars": 2876, + "serialized_chars": 2875, "estimated_input_tokens": 719, "estimated_eval_tokens": 1694 }, @@ -15993,7 +16175,7 @@ "node_count": 3, "item_count": 3, "soft_pair_count": 3, - "serialized_chars": 2876, + "serialized_chars": 2875, "estimated_input_tokens": 719, "estimated_eval_tokens": 1694 }, @@ -16002,30 +16184,30 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "The three issues are unrelated: truncation direction for token sequences, ReduceLROnPlateau-like scheduler behavior, and Wav2Vec2 word time offsets. There is no shared underlying bug or feature request, so this is not a duplicate cluster.", - "confidence": 0.99, - "canonical_issue_reason": "No clear canonical issue: the items are about different parts of the library and different requested behaviors.", + "summary": "These three issues are about different feature requests in different areas: token truncation behavior, learning-rate scheduling, and Wav2Vec2 word boundary offsets. I do not see any duplicates or a single coherent canonical issue here.", + "confidence": 0.98, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "Issue #10519 is the most self-contained and specific, but it still does not represent the other topics.", + "best_issue_reason": null, "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:10519", "right": "issue:11307", "accept": false, - "reason": "Truncation-from-beginning vs Wav2Vec2 word-offset timing are unrelated requests with different code paths and expected fixes." + "reason": "One is a tokenizer truncation option request; the other asks for word start/end timestamps in Wav2Vec2. Different API areas and different underlying problems." }, { "left": "issue:10519", "right": "issue:11005", "accept": false, - "reason": "Sequence truncation behavior and ReduceLROnPlateau-like scheduler functionality are completely different features." + "reason": "Token truncation behavior and ReduceLROnPlateau-like scheduler functionality are unrelated feature requests with no shared code-path or concrete bug." }, { "left": "issue:11005", "right": "issue:11307", "accept": false, - "reason": "Learning-rate scheduling behavior and Wav2Vec2 time-offset extraction do not describe the same bug or change." + "reason": "A learning-rate scheduler feature request is unrelated to extracting time offsets from Wav2Vec2 word boundaries." } ] }, @@ -16057,12 +16239,12 @@ "issue:13244|issue:16618", "issue:15980|issue:16618" ], - "prepared_review_unit_hash": "a98d5f56441154270f075864869317faa36b7ebe441fc1c734b5f704ff0be398", + "prepared_review_unit_hash": "5f276f9695cd09466d7695025896d5ce7e5e6173ce00f8622c4a3c02ca35051e", "component_budget": { "node_count": 3, "item_count": 3, "soft_pair_count": 2, - "serialized_chars": 2477, + "serialized_chars": 2479, "estimated_input_tokens": 620, "estimated_eval_tokens": 1496 }, @@ -16070,7 +16252,7 @@ "node_count": 3, "item_count": 3, "soft_pair_count": 2, - "serialized_chars": 2477, + "serialized_chars": 2479, "estimated_input_tokens": 620, "estimated_eval_tokens": 1496 }, @@ -16079,24 +16261,24 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Only one item is clearly unrelated (Tapas tokenization). The two model-loading issues are both about Hugging Face model download/load failures, but they describe different failure modes: private-model auth/error-message handling vs. a public model that cannot be loaded. I would not merge them as duplicates.", - "confidence": 0.86, - "canonical_issue_reason": "No single canonical duplicate stands out: 13244 is unrelated, and 15980 vs 16618 appear to be different model-download/loading problems rather than the same bug.", + "summary": "The cluster is not a true duplicate set. Issue 13244 is about Tapas tokenization matching TensorFlow code, while 15980 is about a bad login-related error for private model downloads, and 16618 is a separate model-loading failure for bert-base-uncased. The soft-similarity signals are too broad to treat any pair as the same underlying bug.", + "confidence": 0.91, + "canonical_issue_reason": "No single canonical issue is strongly supported; if one must be chosen, #16618 is the broadest model-loading report, but it does not match the others closely enough to serve as a true duplicate anchor.", "canonical_pr_reason": null, - "best_issue_reason": "15980 is the most concrete standalone bug report, with a specific failure mode (private model download while not logged in) and a likely targeted fix path.", + "best_issue_reason": "#16618 is the most central and broadly phrased model-loading issue in the set, but it still appears distinct from both the Tapas tokenization bug and the private-model login/error-message issue.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:13244", "right": "issue:16618", "accept": false, - "reason": "Different subsystems and symptoms: Tapas tokenization vs. loading bert-base-uncased. No shared concrete code-path bug." + "reason": "Tapas tokenization vs. bert-base-uncased loading are different code paths and different failure modes; only a vague NLP/model similarity is shared." }, { "left": "issue:15980", "right": "issue:16618", "accept": false, - "reason": "Both involve model download/load, but one is specifically about unauthenticated private-model access and the other is a public model load failure; they do not look like the same underlying bug." + "reason": "Both concern model download/load failures, but one is specifically about private models without login and the other is a public model load failure; not the same concrete bug." } ] }, @@ -16126,12 +16308,12 @@ "soft_pairs": [ "issue:26350|issue:33260" ], - "prepared_review_unit_hash": "26c3e9a2af8cdcc77d2d6ede62a8d169c6996c058d51b890bb10ea28aa024d75", + "prepared_review_unit_hash": "ab1a4f331af7c6d6534f6bdcd7a616d0221a85181649e4f63ca332c1dcb19f4c", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1590, + "serialized_chars": 1589, "estimated_input_tokens": 398, "estimated_eval_tokens": 1052 }, @@ -16139,7 +16321,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1590, + "serialized_chars": 1589, "estimated_input_tokens": 398, "estimated_eval_tokens": 1052 }, @@ -16148,18 +16330,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are two separate community-contribution requests for different features in different areas: Flash Attention 2 architecture support vs GGUF architecture support. They are not duplicates and should remain separate issues.", + "summary": "These are two separate community contribution tracking issues for different features: Flash Attention 2 support vs GGUF support. They share a broad theme of expanding architecture support, but they do not describe the same underlying change or bug.", "confidence": 0.98, - "canonical_issue_reason": "No single canonical issue fits both; the requests target different functionality and different support matrices.", + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "Neither issue is a better duplicate representative because they concern unrelated enhancements rather than the same underlying problem.", + "best_issue_reason": "Issue 26350 is the stronger representative issue overall because it has substantially higher discussion activity and appears more central within this cluster, but it is not a duplicate of 33260.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:26350", "right": "issue:33260", "accept": false, - "reason": "Different feature requests: Flash Attention 2 support and GGUF support are separate architecture-compatibility tasks, not the same underlying bug or change." + "reason": "Reject: the issues target different features and code paths. Flash Attention 2 support and GGUF support are unrelated contributions beyond both being architecture-expansion requests." } ] }, @@ -16207,12 +16389,12 @@ "issue:43834|issue:44365", "issue:44365|issue:44370" ], - "prepared_review_unit_hash": "44336f855f2538e06b920db94d8d0d1a27511887f9839c7eae641e87a73596de", + "prepared_review_unit_hash": "bc60e30a09a96aef241cce34c5e31df766a9e0cf2372eae5fc8422a29172b3e7", "component_budget": { "node_count": 6, "item_count": 6, "soft_pair_count": 15, - "serialized_chars": 8795, + "serialized_chars": 8794, "estimated_input_tokens": 2199, "estimated_eval_tokens": 4654 }, @@ -16220,7 +16402,7 @@ "node_count": 6, "item_count": 6, "soft_pair_count": 15, - "serialized_chars": 8795, + "serialized_chars": 8794, "estimated_input_tokens": 2199, "estimated_eval_tokens": 4654 }, @@ -16229,112 +16411,109 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Most issues are clearly separate i18n translation requests for different languages. The three placeholder-titled issues (43834, 44365, 44370) look like the same duplicated template request and are the only likely duplicates.", - "confidence": 0.89, - "canonical_issue_reason": "issue:43834 is the oldest of the near-identical placeholder-titled translation issues, so it is the best canonical representative for that duplicate set.", + "summary": "These are all documentation translation requests, but the explicitly named issues are for different languages, and the generic placeholder-title issues do not provide enough evidence to treat them as the same underlying request. No PRs are present.", + "confidence": 0.86, + "canonical_issue_reason": "No single canonical duplicate stands out: the named issues are language-specific requests, and the placeholder-title issues are too underspecified to safely collapse into one canonical issue.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43834 is the strongest overall issue candidate because it matches the duplicated template cluster while being the earliest instance.", + "best_issue_reason": "Issue 43834 is the closest representative only because it is the broad template-style translation issue, but it still should not subsume the language-specific requests.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:27088", "right": "issue:44370", "accept": false, - "reason": "Different translation requests: Turkish vs a placeholder template issue, not the same underlying task." + "reason": "Turkish docs translation vs generic placeholder translation; no evidence they are the same request." }, { "left": "issue:27088", "right": "issue:44365", "accept": false, - "reason": "Different translation requests: Turkish vs a placeholder template issue, not the same underlying task." + "reason": "Turkish docs translation vs generic placeholder translation; different underlying request is more likely." }, { "left": "issue:27088", "right": "issue:43834", "accept": false, - "reason": "Turkish-specific docs translation is distinct from the placeholder template issue." + "reason": "Named Turkish translation request vs underspecified template issue; not enough to merge." }, { "left": "issue:27088", "right": "issue:38435", "accept": false, - "reason": "Different languages and likely different translation work: Turkish vs Romanian." + "reason": "Different target languages: Turkish vs Romanian." }, { "left": "issue:27088", "right": "issue:28936", "accept": false, - "reason": "Different languages and likely different translation work: Turkish vs Spanish." + "reason": "Different target languages: Turkish vs Spanish." }, { "left": "issue:28936", "right": "issue:43834", "accept": false, - "reason": "Spanish-specific docs translation is distinct from the placeholder template issue." + "reason": "Spanish translation request vs generic placeholder issue; no concrete same-bug evidence." }, { "left": "issue:28936", "right": "issue:44365", "accept": false, - "reason": "Spanish-specific docs translation is distinct from the placeholder template issue." + "reason": "Spanish translation request vs generic placeholder issue; insufficient evidence of a single underlying issue." }, { "left": "issue:28936", "right": "issue:38435", "accept": false, - "reason": "Different languages and likely different translation work: Spanish vs Romanian." + "reason": "Different target languages: Spanish vs Romanian." }, { "left": "issue:28936", "right": "issue:44370", "accept": false, - "reason": "Spanish-specific docs translation is distinct from the placeholder template issue." + "reason": "Spanish translation request vs generic placeholder issue; do not merge on broad similarity alone." }, { "left": "issue:38435", "right": "issue:44370", "accept": false, - "reason": "Romanian-specific docs translation is distinct from the placeholder template issue." + "reason": "Romanian translation request vs generic placeholder issue; not enough evidence they are the same request." }, { "left": "issue:38435", "right": "issue:44365", "accept": false, - "reason": "Romanian-specific docs translation is distinct from the placeholder template issue." + "reason": "Romanian translation request vs generic placeholder issue; no explicit link or shared target." }, { "left": "issue:38435", "right": "issue:43834", "accept": false, - "reason": "Romanian-specific docs translation is distinct from the placeholder template issue." + "reason": "Romanian translation request vs underspecified template issue; too little evidence to merge." }, { "left": "issue:43834", "right": "issue:44370", - "accept": true, - "reason": "Both are the same placeholder-titled translation request and appear to be duplicated template issues." + "accept": false, + "reason": "Both are template-style titles, but the bodies/targets are not shown; strict triage should not assume they are duplicates." }, { "left": "issue:43834", "right": "issue:44365", - "accept": true, - "reason": "Both are the same placeholder-titled translation request and appear to be duplicated template issues." + "accept": false, + "reason": "Identical placeholder title is not enough to prove the same underlying translation request." }, { "left": "issue:44365", "right": "issue:44370", - "accept": true, - "reason": "Both are the same placeholder-titled translation request and appear to be duplicated template issues." + "accept": false, + "reason": "Both are generic placeholder translation issues, but there is no concrete evidence they target the same language or work item." } ] }, - "evaluator_result": { - "accept": true, - "feedback": "The summary is grounded in the packet: the three placeholder-titled issues are the only clearly duplicate-looking cluster (very high textual similarity and identical titles), and the language-specific issues are treated conservatively as separate translation requests. The canonical/oldest choice of issue:43834 is also supported by creation times." - }, - "evaluator_used": true, + "evaluator_result": null, + "evaluator_used": false, "retried": false, - "accepted_nontrivial_soft_edge": true, + "accepted_nontrivial_soft_edge": false, "error_kind": null, "error_message": null }, @@ -16406,250 +16585,250 @@ "issue:43837|issue:45177", "issue:45322|issue:46133" ], - "prepared_review_unit_hash": "52427afcfeef5f15f492d06a6fd29d8f6aa92ac49b6389ef4ec2f4d3e6760a9b", + "prepared_review_unit_hash": "4e38cc7d50b7e00681e72aebc9f2c16725af0212e2f2882789aa0f22d9e6203c", "component_budget": { "node_count": 16, "item_count": 16, "soft_pair_count": 36, - "serialized_chars": 21661, - "estimated_input_tokens": 5416, - "estimated_eval_tokens": 11088 + "serialized_chars": 21657, + "estimated_input_tokens": 5415, + "estimated_eval_tokens": 11086 }, "budget": { "node_count": 16, "item_count": 16, "soft_pair_count": 36, - "serialized_chars": 21661, - "estimated_input_tokens": 5416, - "estimated_eval_tokens": 11088 + "serialized_chars": 21657, + "estimated_input_tokens": 5415, + "estimated_eval_tokens": 11086 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These issues are mostly independent model-addition requests across different models and modalities; the similarities are generic wording, not the same underlying change.", - "confidence": 0.96, - "canonical_issue_reason": "No issue here is a true duplicate canonical: the cluster spans unrelated requests (BERT variants, detectors, audio/TTS/ASR, OCR, vision-language).", + "summary": "This cluster is a loose topical group of unrelated model-addition requests. They all resemble \u201cplease add support for model X,\u201d but they are not duplicates of the same underlying issue. No PRs are present.", + "confidence": 0.95, + "canonical_issue_reason": "issue:43671 is the broadest and most central representative of the cluster\u2019s generic model-support requests, but it is not a true duplicate of the others.", "canonical_pr_reason": null, - "best_issue_reason": "issue:42738 is the broadest umbrella-style request in the set, but it still does not represent the rest of the cluster well.", + "best_issue_reason": "issue:43671 is the best single representative because it is an open, discussion-backed model support request and sits near the center of the similarity graph.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:36817", "right": "issue:41211", "accept": false, - "reason": "EuroBert config support and DEIMv2 are unrelated models with different architecture needs." + "reason": "Different requested models (EuroBert vs DEIMv2); only the generic add-support framing overlaps." }, { "left": "issue:36817", "right": "issue:43671", "accept": false, - "reason": "EuroBert is a text model config request; Qwen3-TTS is an audio generation model request." + "reason": "EuroBert config support and Qwen3-TTS support are unrelated model additions." }, { "left": "issue:36817", "right": "issue:36879", "accept": false, - "reason": "EuroBert config and RF-DETR are different model families and code paths." + "reason": "EuroBert vs RF-DETR are different architectures and different feature requests." }, { "left": "issue:36879", "right": "issue:41211", "accept": false, - "reason": "RF-DETR and DEIMv2 are distinct detection models, not the same change." + "reason": "RF-DETR and DEIMv2 are distinct model integrations, not the same bug/change." }, { "left": "issue:36879", "right": "issue:43909", "accept": false, - "reason": "RF-DETR is a vision detector; LFM2.5 Audio is an audio model request." + "reason": "RF-DETR detection support and LFM2.5 Audio support are different domains." }, { "left": "issue:36879", "right": "issue:44121", "accept": false, - "reason": "RF-DETR and the sparse transformer request are different model architectures and targets." + "reason": "RF-DETR and OpenAI weight-sparse transformer are unrelated model requests." }, { "left": "issue:36879", "right": "issue:43837", "accept": false, - "reason": "RF-DETR vision detection and Qwen3-ASR speech recognition are unrelated." + "reason": "RF-DETR and Qwen3-ASR are different models with different code paths." }, { "left": "issue:36879", "right": "issue:43671", "accept": false, - "reason": "RF-DETR and Qwen3-TTS are different modalities and implementation areas." + "reason": "RF-DETR and Qwen3-TTS are unrelated additions despite both being model support requests." }, { "left": "issue:37015", "right": "issue:43837", "accept": false, - "reason": "NeoBERT and Qwen3-ASR are unrelated model requests." + "reason": "NeoBERT and Qwen3-ASR are unrelated model additions." }, { "left": "issue:37015", "right": "issue:44121", "accept": false, - "reason": "NeoBERT and circuit-sparsity transformer support are different code paths and model types." + "reason": "NeoBERT support and weight-sparse transformer support do not describe the same change." }, { "left": "issue:37015", "right": "issue:43909", "accept": false, - "reason": "NeoBERT is a text model; LFM2.5 Audio is an audio model request." + "reason": "NeoBERT and LFM2.5 Audio are distinct model requests." }, { "left": "issue:36879", "right": "issue:37015", "accept": false, - "reason": "RF-DETR and NeoBERT are different model families with no shared concrete bug." + "reason": "RF-DETR and NeoBERT are different model integrations." }, { "left": "issue:37015", "right": "issue:42503", "accept": false, - "reason": "NeoBERT and ModernVBERT are separate model additions, not one duplicate request." + "reason": "NeoBERT and ModernVBERT are different BERT-family requests, not the same model." }, { "left": "issue:41115", "right": "issue:43837", "accept": false, - "reason": "MiniCPM3 architecture support and Qwen3-ASR are distinct model requests." + "reason": "MiniCPM3 architecture support and Qwen3-ASR support are unrelated." }, { "left": "issue:41115", "right": "issue:44121", "accept": false, - "reason": "MiniCPM3 support and sparse-transformer support are different implementation changes." + "reason": "MiniCPM3 and the weight-sparse transformer request target different code paths." }, { "left": "issue:41115", "right": "issue:42503", "accept": false, - "reason": "MiniCPM3 and ModernVBERT are different model families." + "reason": "MiniCPM3 and ModernVBERT are separate model additions." }, { "left": "issue:41115", "right": "issue:45027", "accept": false, - "reason": "MiniCPM3 is not the same underlying request as Voxtral TTS support." + "reason": "MiniCPM3 and Voxtral TTS are different model families and tasks." }, { "left": "issue:41211", "right": "issue:43671", "accept": false, - "reason": "DEIMv2 and Qwen3-TTS are unrelated vision and audio model requests." + "reason": "DEIMv2 and Qwen3-TTS are unrelated model requests." }, { "left": "issue:42503", "right": "issue:43837", "accept": false, - "reason": "ModernVBERT and Qwen3-ASR are different models and modalities." + "reason": "ModernVBERT and Qwen3-ASR are different model additions." }, { "left": "issue:42503", "right": "issue:43671", "accept": false, - "reason": "ModernVBERT and Qwen3-TTS are unrelated requests." + "reason": "ModernVBERT and Qwen3-TTS are different models with different support needs." }, { "left": "issue:42503", "right": "issue:42738", "accept": false, - "reason": "ModernVBERT is a specific model request; BERT-like models with RoPE is a broad concept request, not the same item." + "reason": "ModernVBERT and BERT-like models with RoPE are related only broadly, not the same concrete change." }, { "left": "issue:42738", "right": "issue:45027", "accept": false, - "reason": "BERT-like models with RoPE and Voxtral TTS support are unrelated text-architecture vs audio-model requests." + "reason": "RoPE support for BERT-like models and Voxtral TTS support are unrelated changes." }, { "left": "issue:43671", "right": "issue:43837", "accept": false, - "reason": "Qwen3-TTS and Qwen3-ASR are related by family but are different concrete models and code paths." + "reason": "Qwen3-TTS and Qwen3-ASR are related product lines but different modalities and implementations." }, { "left": "issue:43671", "right": "issue:45027", "accept": false, - "reason": "Qwen3-TTS and Voxtral-4B-TTS are different TTS model requests." + "reason": "Qwen3-TTS and Voxtral TTS are distinct TTS model support requests." }, { "left": "issue:43671", "right": "issue:45177", "accept": false, - "reason": "Qwen3-TTS and DeepSeek-OCR2 are different modalities and model additions." + "reason": "Qwen3-TTS and DeepSeek-OCR2 are different model additions." }, { "left": "issue:43671", "right": "issue:44121", "accept": false, - "reason": "Qwen3-TTS and the sparse transformer request target different architectures and tasks." + "reason": "Qwen3-TTS support and weight-sparse transformer support are not the same underlying change." }, { "left": "issue:43671", "right": "issue:43909", "accept": false, - "reason": "Qwen3-TTS and LFM2.5 Audio are separate audio-model support requests." + "reason": "Qwen3-TTS and LFM2.5 Audio are different audio-model requests." }, { "left": "issue:43837", "right": "issue:43909", "accept": false, - "reason": "Qwen3-ASR and LFM2.5 Audio are different speech/audio model requests." + "reason": "Qwen3-ASR and LFM2.5 Audio are separate audio model integrations." }, { "left": "issue:43837", "right": "issue:45322", "accept": false, - "reason": "ASR support and EUPE vision-language encoder support are unrelated." + "reason": "Qwen3-ASR and EUPE are unrelated model requests." }, { "left": "issue:43837", "right": "issue:44121", "accept": false, - "reason": "Qwen3-ASR and weight-sparse transformer support are different concrete changes." + "reason": "Qwen3-ASR support and weight-sparse transformer support are different changes." }, { "left": "issue:43909", "right": "issue:44205", "accept": false, - "reason": "LFM2.5 Audio and SAM3-LiteText are different model domains." + "reason": "LFM2.5 Audio and SAM3-LiteText are different model families and tasks." }, { "left": "issue:44121", "right": "issue:45322", "accept": false, - "reason": "Sparse transformer support and EUPE are unrelated architecture requests." + "reason": "Weight-sparse transformer support and EUPE support are unrelated model requests." }, { "left": "issue:43837", "right": "issue:45027", "accept": false, - "reason": "Qwen3-ASR and Voxtral TTS are different speech model requests." + "reason": "Qwen3-ASR and Voxtral TTS are different speech-model additions." }, { "left": "issue:37015", "right": "issue:45027", "accept": false, - "reason": "NeoBERT and Voxtral TTS are unrelated text and audio model requests." + "reason": "NeoBERT and Voxtral TTS are unrelated model requests." }, { "left": "issue:43837", "right": "issue:45177", "accept": false, - "reason": "ASR support and DeepSeek-OCR2 are different tasks and models." + "reason": "Qwen3-ASR and DeepSeek-OCR2 are different model additions." }, { "left": "issue:45322", "right": "issue:46133", "accept": false, - "reason": "EUPE and TIPSv2 are different vision-language models/papers, not one duplicate change." + "reason": "EUPE and TIPSv2 are distinct vision-language model requests." } ] }, @@ -16679,12 +16858,12 @@ "soft_pairs": [ "issue:41084|issue:43037" ], - "prepared_review_unit_hash": "398e1de4b1b79e266eeebf7ba84e94289edf4ce24a442b9b7f47c5f1b633eea7", + "prepared_review_unit_hash": "37fca706d0e8b45ecbafe199df80c911dc3460adad84ad37fb0efcb1f4c24703", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1479, + "serialized_chars": 1480, "estimated_input_tokens": 370, "estimated_eval_tokens": 996 }, @@ -16692,7 +16871,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1479, + "serialized_chars": 1480, "estimated_input_tokens": 370, "estimated_eval_tokens": 996 }, @@ -16701,18 +16880,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These two issues are not duplicates: one is about Set decoding behavior, the other is a request for DeepSeek v3.2 model support. They concern different features and likely different code paths.", - "confidence": 0.96, - "canonical_issue_reason": "Neither issue is a clear canonical duplicate target because they describe unrelated problems: Set block decoding vs DeepSeek v3.2 support.", + "summary": "The two issues are unrelated: one is a block decoding setting change, the other is a request for DeepSeek v3.2 support. They should not be treated as duplicates or one cluster representative.", + "confidence": 0.98, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "No single issue is a better global canonical candidate; the topics are unrelated and neither subsumes the other.", + "best_issue_reason": null, "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:41084", "right": "issue:43037", "accept": false, - "reason": "Different underlying requests with no shared target or code path; similarity is only superficial from both being feature/bug reports." + "reason": "Different underlying requests: 'Set Block Decoding' is a decoding/configuration feature, while 'DeepSeek v3.2 support' is model support. No shared code-path or concrete bug overlap is evident." } ] }, @@ -16742,7 +16921,7 @@ "soft_pairs": [ "issue:43598|issue:43600" ], - "prepared_review_unit_hash": "4ec85c809cbca9ee82f65d4296a5cd826f6e2392aef3db18ed34c1673e0845ec", + "prepared_review_unit_hash": "7a2f54d8b939df308495286e3bd634788e7888f6f2d004c7ba890ac0cf813ddd", "component_budget": { "node_count": 2, "item_count": 2, @@ -16764,18 +16943,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are related Trainer customization issues, but they describe different problems: one is about `remove_unused_column` behavior, the other about FLOP estimation for custom trainers. They should not be treated as duplicates.", - "confidence": 0.92, - "canonical_issue_reason": "Issue #43598 is the more concrete and directly actionable bug report, focused on `remove_unused_column` customizability in Trainer.", + "summary": "The two issues are loosely related through Trainer/custom-trainer behavior, but they describe different problems: #43598 is about `remove_unused_column` customization, while #43600 is about FLOPS estimation for custom trainers. They should not be treated as duplicates.", + "confidence": 0.96, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "#43598 is the best representative issue because it names a specific code path and expected behavior change rather than a broader tooling concern.", + "best_issue_reason": "No issue is a clear canonical duplicate representative because the two reports target different Trainer code paths; if forced, #43598 is the more self-contained feature request, but it does not subsume #43600.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43598", "right": "issue:43600", "accept": false, - "reason": "They both involve custom trainers, but one is about column removal in Trainer while the other is about FLOP estimation; different underlying bugs, so not duplicates." + "reason": "Different underlying concerns: one changes column-removal behavior for customizability, the other changes FLOPS estimation for custom trainers. Shared Trainer context is too broad to merge them." } ] }, @@ -16805,7 +16984,7 @@ "soft_pairs": [ "issue:43684|issue:44405" ], - "prepared_review_unit_hash": "ec03569b0d1a716d30be33ef5004f7674bb3a049b1b86fbe66407f54900e7de3", + "prepared_review_unit_hash": "fdfd7cfed96e347f416dae52fae993f9a95da4f7b861acd420c98dca7a9a6a59", "component_budget": { "node_count": 2, "item_count": 2, @@ -16827,18 +17006,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are different feature requests for different Qwen variants and task heads, so they should not be deduplicated.", - "confidence": 0.96, - "canonical_issue_reason": "Issue 43684 is the broader registration request, but it is not the same underlying need as 44405, so there is no true canonical duplicate here.", + "summary": "These are separate feature requests for different model families and different auto-model heads, so they should not be clustered as duplicates.", + "confidence": 0.95, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "43684 is the better representative of the cluster because it is the more general model-registration request, while 44405 is a separate sequence-classification support request.", + "best_issue_reason": "No single best canonical issue: #43684 asks for Qwen3-Omni registration in AutoModel/AutoModelForConditionalGeneration, while #44405 asks for AutoModelForSequenceClassification support for Qwen3.5 (Qwen3_5Config). They target different code paths and tasks.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43684", "right": "issue:44405", "accept": false, - "reason": "Different model families and different capabilities: Qwen3-Omni AutoModel/AutoModelForConditionalGeneration registration versus Qwen3.5 AutoModelForSequenceClassification support." + "reason": "Different model variants and different auto-class heads; this is not the same underlying bug or change." } ] }, @@ -16875,80 +17054,77 @@ "issue:43845|issue:43859", "issue:43846|issue:43859" ], - "prepared_review_unit_hash": "2dde03940569ae68f7732f3e66f504dc91eeafba94bf52abc2a1068a17d9b61d", + "prepared_review_unit_hash": "0019635b2842297879147a3e434944583c548a51a1067d55309375c16fdce949", "component_budget": { "node_count": 4, "item_count": 4, "soft_pair_count": 6, - "serialized_chars": 4267, - "estimated_input_tokens": 1067, - "estimated_eval_tokens": 2390 + "serialized_chars": 4261, + "estimated_input_tokens": 1066, + "estimated_eval_tokens": 2388 }, "budget": { "node_count": 4, "item_count": 4, "soft_pair_count": 6, - "serialized_chars": 4267, - "estimated_input_tokens": 1067, - "estimated_eval_tokens": 2390 + "serialized_chars": 4261, + "estimated_input_tokens": 1066, + "estimated_eval_tokens": 2388 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "The three `huggingface` issues look like duplicates of the same vague report; `Fett` appears unrelated. I would keep the earliest `huggingface` issue as the canonical one.", - "confidence": 0.73, - "canonical_issue_reason": "Issue 43845 is the earliest of the three near-identical `huggingface` reports, so it is the best canonical placeholder for that duplicate set.", + "summary": "The cluster is too underspecified to confidently treat these as duplicates. The three \"huggingface\" issues may be related to each other by title, but there is no body or target evidence showing the same underlying bug, and \"Fett\" looks unrelated.", + "confidence": 0.31, + "canonical_issue_reason": "Issue 43845 is the earliest of the repeated \"huggingface\" titles and is the most plausible anchor if this cluster were later confirmed to be duplicates.", "canonical_pr_reason": null, - "best_issue_reason": "43845 is the best global issue candidate because it is the earliest and shares the exact same title as the other likely duplicates.", + "best_issue_reason": "43845 is the best issue candidate only because it is the earliest repeated title; however, the evidence is too weak to call it a true canonical duplicate target.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43835", "right": "issue:43845", "accept": false, - "reason": "Different titles and no shared context beyond generic similarity; no evidence they describe the same bug." + "reason": "Different titles and no body/target evidence tying \"Fett\" to the \"huggingface\" reports." }, { "left": "issue:43835", "right": "issue:43846", "accept": false, - "reason": "Different titles and no shared context beyond generic similarity; no evidence they describe the same bug." + "reason": "Different titles and no evidence of the same underlying problem." }, { "left": "issue:43835", "right": "issue:43859", "accept": false, - "reason": "Different titles and no shared context beyond generic similarity; no evidence they describe the same bug." + "reason": "Different titles and no evidence these describe the same bug." }, { "left": "issue:43845", "right": "issue:43846", - "accept": true, - "reason": "Same exact title, created minutes apart, and no distinguishing body details are visible; these look like the same issue." + "accept": false, + "reason": "Same short title, but there is no issue-body or target detail confirming the same underlying report." }, { "left": "issue:43845", "right": "issue:43859", - "accept": true, - "reason": "Same exact title with only timing differences and no other differentiating evidence; likely duplicate reports." + "accept": false, + "reason": "Matching title alone is not enough here; the issues are too underspecified to merge as duplicates." }, { "left": "issue:43846", "right": "issue:43859", - "accept": true, - "reason": "Same exact title with only timing differences and no other differentiating evidence; likely duplicate reports." + "accept": false, + "reason": "Matching title alone is insufficient to confirm the same bug." } ] }, - "evaluator_result": { - "accept": true, - "feedback": "The grouping is grounded in the packet: 43845/43846/43859 share the exact same title, and 43845 is the earliest among them, so using it as the canonical placeholder is reasonable. 43835 has a different title and weaker similarity, so treating it as separate is conservative. No mergeability concerns apply here since these are issues, not PRs." - }, - "evaluator_used": true, + "evaluator_result": null, + "evaluator_used": false, "retried": false, - "accepted_nontrivial_soft_edge": true, + "accepted_nontrivial_soft_edge": false, "error_kind": null, "error_message": null }, @@ -16982,88 +17158,88 @@ "issue:44572|issue:44937", "issue:44573|issue:44937" ], - "prepared_review_unit_hash": "cfbf533150aa458912f9ff4dcddb5aa239e317ea61d71006dcbeb93b280767e1", + "prepared_review_unit_hash": "984722bb2ca964ae12be9c7c15283bde548d490db181ca44dadf228f9221149c", "component_budget": { "node_count": 5, "item_count": 5, "soft_pair_count": 9, - "serialized_chars": 5791, - "estimated_input_tokens": 1448, - "estimated_eval_tokens": 3152 + "serialized_chars": 5788, + "estimated_input_tokens": 1447, + "estimated_eval_tokens": 3150 }, "budget": { "node_count": 5, "item_count": 5, "soft_pair_count": 9, - "serialized_chars": 5791, - "estimated_input_tokens": 1448, - "estimated_eval_tokens": 3152 + "serialized_chars": 5788, + "estimated_input_tokens": 1447, + "estimated_eval_tokens": 3150 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "This cluster appears to be low-signal spam/noise rather than a real duplicate set. The issues have unrelated titles and no bodies, links, filenames, or shared technical detail to support merging any pair.", - "confidence": 0.98, - "canonical_issue_reason": null, + "summary": "All items are low-signal and look like unrelated spam/placeholder issues rather than duplicates of one concrete bug or change. No soft pair should be merged.", + "confidence": 0.96, + "canonical_issue_reason": "issue:44418 is the only item with a somewhat descriptive title, but it still lacks concrete problem details and does not match the others as the same underlying report.", "canonical_pr_reason": null, - "best_issue_reason": null, + "best_issue_reason": "issue:44418 is the most informative representative by title alone, but it is still not a strong canonical candidate for a duplicate cluster.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44371", "right": "issue:44572", "accept": false, - "reason": "Both are generic spam placeholders with no substantive content; no evidence of the same underlying bug or request." + "reason": "Both are placeholder '' issues with no substantive bug description; no evidence they report the same problem." }, { "left": "issue:44371", "right": "issue:44573", "accept": false, - "reason": "Both are generic spam placeholders with no substantive content; no evidence of the same underlying bug or request." + "reason": "Same as above: generic spam-like placeholders, not a concrete shared issue." }, { "left": "issue:44371", "right": "issue:44418", "accept": false, - "reason": "One is spam and the other is a vague documentation suggestion; they do not describe the same issue." + "reason": "A spam placeholder and a vague documentation suggestion are not enough to establish the same underlying issue." }, { "left": "issue:44371", "right": "issue:44937", "accept": false, - "reason": "Spam placeholder versus an unrelated marketing-style title; no shared concrete problem." + "reason": "The titles are unrelated and there is no shared code-path or explicit linkage." }, { "left": "issue:44418", "right": "issue:44572", "accept": false, - "reason": "A documentation enhancement request and a spam placeholder are not the same underlying issue." + "reason": "Different vague topics with no concrete symptoms, targets, or evidence of duplication." }, { "left": "issue:44418", "right": "issue:44573", "accept": false, - "reason": "A documentation enhancement request and a spam placeholder are not the same underlying issue." + "reason": "The documentation suggestion does not look like the same report as the spam placeholder issue." }, { "left": "issue:44572", "right": "issue:44573", "accept": false, - "reason": "Both are spam placeholders, but there is no concrete bug/change to unify beyond generic similarity." + "reason": "Both are spam-like placeholders, but there is no basis to conclude they describe the same underlying bug or change." }, { "left": "issue:44572", "right": "issue:44937", "accept": false, - "reason": "Generic spam placeholder versus unrelated promotional title; no duplicate signal." + "reason": "Unrelated titles and no shared issue details indicate different topics." }, { "left": "issue:44573", "right": "issue:44937", "accept": false, - "reason": "Generic spam placeholder versus unrelated promotional title; no duplicate signal." + "reason": "No concrete overlap beyond being low-information closed issues." } ] }, @@ -17093,12 +17269,12 @@ "soft_pairs": [ "issue:45145|issue:45151" ], - "prepared_review_unit_hash": "8a4bbdc3807e809b336bfa2b03d09f348ebebb4de9df358ff2820e124c8bef17", + "prepared_review_unit_hash": "95571c546c479d44e639775f5e05db2caaffbfbaa98a0f38754315bf85ed912f", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1645, + "serialized_chars": 1646, "estimated_input_tokens": 412, "estimated_eval_tokens": 1080 }, @@ -17106,7 +17282,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1645, + "serialized_chars": 1646, "estimated_input_tokens": 412, "estimated_eval_tokens": 1080 }, @@ -17115,24 +17291,24 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "#45145 remains the best canonical representative because it was created first and already has inbound reference activity. The pair looks very likely related, but the available evidence is limited to identical titles, similar body sizes, and a soft-similarity signal rather than a direct body match.", - "confidence": 0.89, - "canonical_issue_reason": "#45145 is earlier and has the only inbound reference, so it is the best canonical issue for this near-duplicate pair.", + "summary": "The two closed issues are highly similar and likely duplicate the same report: identical titles, very similar body lengths, close timestamps, and a soft-similarity signal. #45145 is the better canonical issue because it was created first and already has an inbound reference.", + "confidence": 0.87, + "canonical_issue_reason": "#45145 is the best canonical anchor: it appeared first and already has an inbound reference, while the paired issue is highly similar.", "canonical_pr_reason": null, - "best_issue_reason": "#45145 is the strongest issue to keep as the representative record because it predates #45151 and already has linkage from another page.", + "best_issue_reason": "#45145 is the strongest representative of the cluster due to its earlier creation time and existing inbound reference, with no evidence that it is a different report.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:45145", "right": "issue:45151", "accept": true, - "reason": "These issues are probably duplicates: they have the same title and closely matching body lengths with a soft-similarity signal. The evidence does not prove identical bodies, so this should be treated as a probable duplicate rather than a fully confirmed one." + "reason": "The pair looks like the same underlying issue: identical titles, very close body sizes/timestamps, and a similarity signal. The evidence is suggestive rather than definitive, but sufficient to treat them as duplicates." } ] }, "evaluator_result": { "accept": false, - "feedback": "The canonical choice for #45145 is grounded: it was created earlier and has one inbound reference. However, the duplicate verdict is a bit too strong for the packet. The evidence shown is only identical titles, close body lengths, and a soft_similarity signal; it does not establish that the bodies are the same report. Please soften the language to indicate probable duplication rather than a confirmed true duplicate pair." + "feedback": "The pair looks plausibly duplicate, but the analyst overstates the evidence by calling it a definite \"true duplicate pair.\" The packet only supports identical titles, similar body length, close timestamps, and a soft-similarity signal; there are no explicit links or stronger deterministic indicators. The summary is mostly grounded, but the soft-edge verdict should be more conservative." }, "evaluator_used": true, "retried": true, @@ -17213,12 +17389,12 @@ "soft_pairs": [ "issue:45672|issue:45674" ], - "prepared_review_unit_hash": "c59828513e077647411c0c8a86f70ddc6d378937647ef75dae71832900b7fb48", + "prepared_review_unit_hash": "4d665438b390f53ff0576751f55adfa4c00a49f651492b172f1ded1b3a6aa3ad", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1667, + "serialized_chars": 1668, "estimated_input_tokens": 417, "estimated_eval_tokens": 1090 }, @@ -17226,7 +17402,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1667, + "serialized_chars": 1668, "estimated_input_tokens": 417, "estimated_eval_tokens": 1090 }, @@ -17235,18 +17411,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "The two issues both involve 4-bit inference, but they describe different failures and root causes: one is a Gemma4 audio-module torch.finfo TypeError on uint8 weights, the other is a BitsAndBytesConfig skip-modules regression that drops the default lm_head exclusion and triggers an AssertionError. They should not be merged as duplicates.", - "confidence": 0.94, - "canonical_issue_reason": null, + "summary": "Two open issues report different 4-bit inference failures in different parts of the stack: one in Gemma4 audio module dtype handling, the other in BitsAndBytesConfig module-exclusion logic. They are related by quantized inference, but not the same underlying bug.", + "confidence": 0.92, + "canonical_issue_reason": "Issue #45674 is the better representative because it describes a clear, config-driven regression with a concrete failure mode and reproduction path.", "canonical_pr_reason": null, - "best_issue_reason": "No single canonical issue fits both reports; if one must be kept as the representative, issue 45674 is the more general configuration regression, but it is still a separate bug from 45672.", + "best_issue_reason": "Issue #45674 is the strongest single issue in the cluster: it is specific, actionable, and centers on a concrete 4-bit inference regression.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:45672", "right": "issue:45674", "accept": false, - "reason": "Same broad 4-bit inference area, but different concrete code-path failures and different fixes; not the same underlying bug." + "reason": "Different failures and code paths: #45672 is a Gemma4 audio-module uint8/torch.finfo TypeError, while #45674 is a BitsAndBytesConfig lm_int8_skip_modules regression that clears the default lm_head exclusion and triggers an AssertionError. They should not be merged as duplicates." } ] }, @@ -17276,7 +17452,7 @@ "soft_pairs": [ "issue:45762|issue:45763" ], - "prepared_review_unit_hash": "9cd79cd70705cf70a67e5049700beb1bb36803fd222b5e0f68638be5a6491ea2", + "prepared_review_unit_hash": "8e9a399efd21ef390332339b6077c3df032430369ead22e53e32172cc706f7f3", "component_budget": { "node_count": 2, "item_count": 2, @@ -17298,7 +17474,7 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "The two closed issues look like unrelated, low-signal entries (personal names with dates) and do not describe the same bug or request. They should not be clustered as duplicates.", + "summary": "The two closed issues look unrelated: both are bare personal-name/date entries with no body, no links, and no shared bug or feature context. This is not a duplicate cluster.", "confidence": 0.96, "canonical_issue_reason": null, "canonical_pr_reason": null, @@ -17309,7 +17485,7 @@ "left": "issue:45762", "right": "issue:45763", "accept": false, - "reason": "No shared problem description, code path, or explicit linkage; the titles are just different personal-name/date strings and do not indicate the same underlying issue." + "reason": "Titles are different personal-name/date strings and there is no issue body or linked context showing the same underlying problem." } ] }, @@ -17339,7 +17515,7 @@ "soft_pairs": [ "issue:45764|issue:45765" ], - "prepared_review_unit_hash": "4abea14b379304a1a1b87a479314d695fb2878d12bb49123a90c3ceb81d4882e", + "prepared_review_unit_hash": "d486228ace9611605118b53c09e6d7daa7b5bbfffc48772f0b28c8fd96ac8230", "component_budget": { "node_count": 2, "item_count": 2, @@ -17361,18 +17537,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Both items are closed issues with only personal-name/date titles and no body or code references; they do not show evidence of the same bug or change, so this looks like unrelated noise/spam rather than a duplicate cluster.", - "confidence": 0.98, - "canonical_issue_reason": "Issue #45764 is marginally preferable as the canonical representative only because it appears first; there is no substantive issue content to distinguish them.", + "summary": "The two closed issues are unrelated personal-name/date entries with no description, references, or shared technical context. There is no evidence they describe the same bug, request, or change.", + "confidence": 0.99, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "Neither issue is a good canonical technical issue, but #45764 is the least arbitrary placeholder if one must be chosen.", + "best_issue_reason": null, "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:45764", "right": "issue:45765", "accept": false, - "reason": "The titles are different person/date strings with no shared problem statement, code path, or explicit linkage; they do not appear to be duplicates of the same underlying issue." + "reason": "The titles are different names and dates, and both issues lack body text or linking evidence; they do not appear to be the same underlying report." } ] }, @@ -17412,7 +17588,7 @@ "issue:45871|issue:45873", "issue:45872|issue:45873" ], - "prepared_review_unit_hash": "37498e845271dfe272d0cf32f4b601ee9437163c1475e8776872503f0083c264", + "prepared_review_unit_hash": "fe5b78ddf4fbb5ce8d6ec645192f544204d79042420d1536212cef440886aee4", "component_budget": { "node_count": 5, "item_count": 5, @@ -17434,69 +17610,69 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Two exact duplicate issue pairs are present: 45870/45871 and 45872/45873. The financial dataset item (45869) is related by theme but appears to be a different artifact, not the same underlying request.", - "confidence": 0.92, - "canonical_issue_reason": "Issue 45870 is the cleanest canonical representative of the exact duplicate pair because it has the clearer, more general title and is the earlier post.", + "summary": "Two high-confidence duplicate candidates appear: 45870/45871 and 45872/45873. 45869 looks like a separate dataset-promo issue. The evidence supports duplicate pairing mainly from identical titles and identical body lengths, but the packet does not expose body text, so the certainty should stay slightly qualified.", + "confidence": 0.89, + "canonical_issue_reason": "Issue 45870 is a reasonable canonical issue for the first duplicate pair because it is the earlier of the two identical-title reports and has slightly more discussion activity.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 45870 is the best overall issue candidate because it most directly represents the duplicate example request cluster without being overly specific.", + "best_issue_reason": "Issue 45870 is the best overall issue representative for this cluster because it is the earliest clear duplicate candidate and has the most activity among the example-prediction pair.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:45869", "right": "issue:45870", "accept": false, - "reason": "Same finance/domain theme, but one is a dataset bundle and the other is an example request; not the same underlying issue." + "reason": "Different titles and different apparent topic: dataset promo versus example prediction thread." }, { "left": "issue:45869", "right": "issue:45871", "accept": false, - "reason": "Different artifact types and scopes: dataset listing versus example prompt; too broad to treat as duplicates." + "reason": "Different titles and different apparent topic: dataset promo versus example prediction thread." }, { "left": "issue:45870", "right": "issue:45871", "accept": true, - "reason": "Exact same title and highly similar metadata; these are the same issue request." + "reason": "Same title, same body length, and close timestamps; strong duplicate signal." }, { "left": "issue:45870", "right": "issue:45872", "accept": false, - "reason": "Both are financial/DataForge related, but one asks for a prediction example while the other asks for a scraping service; different changes." + "reason": "Related financial/DataForge wording, but the titles describe different requests." }, { "left": "issue:45870", "right": "issue:45873", "accept": false, - "reason": "Same as 45872: related topic, but a different concrete request and not mergeable as one issue." + "reason": "Related financial/DataForge wording, but the titles describe different requests." }, { "left": "issue:45871", "right": "issue:45872", "accept": false, - "reason": "Identical to the prior comparison: duplicate titles do not match the scraping-service request." + "reason": "Different titles and different request shape: example prediction versus scraping-service topic." }, { "left": "issue:45871", "right": "issue:45873", "accept": false, - "reason": "Different concrete request despite shared finance/DataForge wording; not the same underlying bug/change." + "reason": "Different titles and different request shape: example prediction versus scraping-service topic." }, { "left": "issue:45872", "right": "issue:45873", "accept": true, - "reason": "Exact same title and metadata; these are duplicate copies of the same issue request." + "reason": "Same title, same body length, and close timestamps; strong duplicate signal." } ] }, "evaluator_result": { - "accept": true, - "feedback": "Grounded in the packet and appropriately conservative. The two exact-title, same-body-length pairs (45870/45871 and 45872/45873) are reasonable duplicate calls, and 45869 is correctly treated as related but distinct. No soft-edge verdict overstates the evidence." + "accept": false, + "feedback": "The grouping is directionally plausible, but several claims overstate the evidence available in the packet. The packet only provides titles, body lengths, and timestamps; it does not expose body text, so statements like \u201cexact duplicate,\u201d \u201cmatching body length/content pattern,\u201d and \u201cexact same title\u201d are only partly grounded. In particular, calling 45870/45871 and 45872/45873 definite duplicates is stronger than the evidence supports; \u2018likely duplicate\u2019 or \u2018high-confidence duplicate candidate\u2019 would be safer. The summary also says \u2018Two clear duplicate pairs appear,\u2019 which is more certain than the packet justifies. The 45869 separation is conservative and grounded." }, "evaluator_used": true, - "retried": false, + "retried": true, "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null @@ -17512,7 +17688,7 @@ "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-39895-15", + "cluster_id": "cluster-39895-13", "nodes": [ "pull_request:39895", "pull_request:41855", @@ -17526,207 +17702,170 @@ "pull_request:45170", "pull_request:45213", "pull_request:45655", - "pull_request:46048", - "pull_request:46067", - "pull_request:46189" + "pull_request:46048" ], "soft_pairs": [ "pull_request:39895|pull_request:43424", "pull_request:41855|pull_request:43823", "pull_request:41855|pull_request:43424", "pull_request:42668|pull_request:43823", - "pull_request:43251|pull_request:46067", "pull_request:43251|pull_request:43823", "pull_request:43251|pull_request:44827", - "pull_request:43424|pull_request:46067", "pull_request:43424|pull_request:44827", "pull_request:43424|pull_request:43823", "pull_request:43823|pull_request:45655", "pull_request:43823|pull_request:43995", - "pull_request:43823|pull_request:46067", "pull_request:43995|pull_request:45213", "pull_request:43995|pull_request:44827", - "pull_request:44827|pull_request:46189", - "pull_request:44827|pull_request:46067", "pull_request:44827|pull_request:45170", "pull_request:44827|pull_request:46048", + "pull_request:44827|pull_request:45655", "pull_request:44827|pull_request:45213", "pull_request:43823|pull_request:45213", - "pull_request:44827|pull_request:45655", "pull_request:43251|pull_request:43254" ], - "prepared_review_unit_hash": "0f834338daabd0381404e77c4277a24462186e6e41a10a371f891dfb8aaa6a16", + "prepared_review_unit_hash": "12ddc956583868684b2c7dcc90306d3dbae282d1d3d967612f01c080cd619b0d", "component_budget": { - "node_count": 15, - "item_count": 15, - "soft_pair_count": 23, - "serialized_chars": 40802, - "estimated_input_tokens": 10201, - "estimated_eval_tokens": 20658 + "node_count": 13, + "item_count": 13, + "soft_pair_count": 18, + "serialized_chars": 34775, + "estimated_input_tokens": 8694, + "estimated_eval_tokens": 17644 }, "budget": { - "node_count": 15, - "item_count": 15, - "soft_pair_count": 23, - "serialized_chars": 40802, - "estimated_input_tokens": 10201, - "estimated_eval_tokens": 20658 + "node_count": 13, + "item_count": 13, + "soft_pair_count": 18, + "serialized_chars": 34775, + "estimated_input_tokens": 8694, + "estimated_eval_tokens": 17644 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "This cluster is mostly unrelated PRs spanning different models, tests, docs, and infra. The only clear duplicate pair is the fixed_cross_entropy kwargs change: PRs 43251 and 43254 address the same loss-function bug, with 43251 the better canonical choice.", - "confidence": 0.94, + "summary": "Most items are unrelated PRs across different models/features. The only clear duplicate pair is the two `fixed_cross_entropy` PRs, which patch the same function for the same issue.", + "confidence": 0.96, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 43251 is the best canonical PR because it directly targets issue 43240 and is the more complete/explicit version of the same fixed_cross_entropy kwargs fix; 43254 is a near-duplicate of the same change.", + "canonical_pr_reason": "PR #43251 is the best canonical PR for the duplicate pair because it directly targets issue #43240 and presents the more complete, issue-linked fix for `fixed_cross_entropy`.", "best_issue_reason": null, - "best_pr_reason": "If one PR must represent the cluster, 43251 is the strongest anchor because it is the only clear duplicate match and has the clearest issue linkage; otherwise the cluster has no single global canonical PR due to broad topic divergence.", + "best_pr_reason": "PR #43251 is the strongest representative change in this set: it has an explicit issue target, a focused code-path fix, and matches the concrete bug described in the diff. ", "soft_edge_verdicts": [ { "left": "pull_request:39895", "right": "pull_request:43424", "accept": false, - "reason": "Videoprism model/docs addition vs an executorch dynamic-shape export test; different code paths and no shared underlying bug." + "reason": "Different features: new VideoPrism model package/docs vs an Executorch dynamic-shapes test. No shared code-path or mergeable overlap." }, { "left": "pull_request:41855", "right": "pull_request:43823", "accept": false, - "reason": "Mistral tokenizer helper methods vs adding a new MobileLLM model; unrelated functionality." + "reason": "Unrelated tokenizer API additions for Mistral common backend vs adding the MobileLLM model." }, { "left": "pull_request:41855", "right": "pull_request:43424", "accept": false, - "reason": "Tokenizer API additions vs executorch export testing; not the same change or bug." + "reason": "Tokenizer method addition vs Executorch export test; same broad repo area only, not the same bug/change." }, { "left": "pull_request:42668", "right": "pull_request:43823", "accept": false, - "reason": "Processor auto-loading robustness changes vs a new model implementation; too broad and unrelated." - }, - { - "left": "pull_request:43251", - "right": "pull_request:46067", - "accept": false, - "reason": "Cross-entropy kwargs fix vs base_model_prefix loading/conversion fix; different subsystems and code paths." + "reason": "Processor auto-loading robustness changes vs a new model implementation; too different to be duplicates." }, { "left": "pull_request:43251", "right": "pull_request:43823", "accept": false, - "reason": "Loss-function kwargs fix vs new model addition; no shared underlying issue." + "reason": "Loss-function kwargs fix vs adding MobileLLM; no shared underlying bug." }, { "left": "pull_request:43251", "right": "pull_request:44827", "accept": false, - "reason": "Cross-entropy helper bugfix vs Mistral4 test/model changes; unrelated." - }, - { - "left": "pull_request:43424", - "right": "pull_request:46067", - "accept": false, - "reason": "Executorch dynamic-shape export test vs conversion/loading prefix handling; unrelated." + "reason": "`cross_entropy` kwargs support fix is unrelated to Mistral4 test/model adjustments." }, { "left": "pull_request:43424", "right": "pull_request:44827", "accept": false, - "reason": "Export test coverage vs Mistral4 model/test fixes; different concerns." + "reason": "Executorch dynamic-shapes export test and Mistral4 test fixes address different concerns." }, { "left": "pull_request:43424", "right": "pull_request:43823", "accept": false, - "reason": "Executorch export testing vs adding MobileLLM; no common bug/change." + "reason": "Dynamic-shapes export coverage vs MobileLLM model addition are unrelated." }, { "left": "pull_request:43823", "right": "pull_request:45655", "accept": false, - "reason": "New MobileLLM model support vs config resolution-order fix; unrelated." + "reason": "New MobileLLM model vs config-resolution order fix; different code paths and intents." }, { "left": "pull_request:43823", "right": "pull_request:43995", "accept": false, - "reason": "MobileLLM addition vs Falcon output-collection refactor; different model families and code paths." - }, - { - "left": "pull_request:43823", - "right": "pull_request:46067", - "accept": false, - "reason": "MobileLLM model addition vs conversion prefix-loading fix; unrelated." + "reason": "MobileLLM model addition vs Falcon output-collection refactor; not the same change." }, { "left": "pull_request:43995", "right": "pull_request:45213", "accept": false, - "reason": "Falcon refactor vs a model-creation skill/meta PR; not the same underlying change." + "reason": "Falcon forward-signature refactor vs a model-creation skill draft; no overlap in the concrete bug/change." }, { "left": "pull_request:43995", "right": "pull_request:44827", "accept": false, - "reason": "Falcon output-interface refactor vs Mistral4 tests/fixes; unrelated." - }, - { - "left": "pull_request:44827", - "right": "pull_request:46189", - "accept": false, - "reason": "Mistral4 test/fix work vs Cohere2 MoE tensor-parallel plan updates; different models and problems." + "reason": "Falcon interface refactor and Mistral4 fixes are distinct model-specific changes." }, { "left": "pull_request:44827", - "right": "pull_request:46067", + "right": "pull_request:45170", "accept": false, - "reason": "Mistral4 tests/fixes vs conversion/loading prefix handling; no shared concrete bug." + "reason": "Mistral4 test/model fixes vs CLIP-like `layrnorm` rename correction; separate bugs." }, { "left": "pull_request:44827", - "right": "pull_request:45170", + "right": "pull_request:46048", "accept": false, - "reason": "Mistral4 test/fix work vs layrnorm-to-layernorm rename/weight mapping cleanup; unrelated." + "reason": "Mistral4 fixes vs adding a TDT loss kernel; unrelated subsystems." }, { "left": "pull_request:44827", - "right": "pull_request:46048", + "right": "pull_request:45655", "accept": false, - "reason": "Mistral4 tests/fixes vs TDT loss kernel integration; different subsystems." + "reason": "Mistral4 test/model fixes are unrelated to class config resolution order." }, { "left": "pull_request:44827", "right": "pull_request:45213", "accept": false, - "reason": "Mistral4 test/fix work vs model-creation skill/meta PR; unrelated." + "reason": "Model/test fixes vs a draft model-creation skill; not a duplicate change." }, { "left": "pull_request:43823", "right": "pull_request:45213", "accept": false, - "reason": "MobileLLM model addition vs a draft model-creation skill PR; no same bug/change." - }, - { - "left": "pull_request:44827", - "right": "pull_request:45655", - "accept": false, - "reason": "Mistral4 tests/fixes vs class config-resolution ordering fix; unrelated." + "reason": "New MobileLLM model addition vs model-creation skill draft; one is implementation, the other is tooling/docs." }, { "left": "pull_request:43251", "right": "pull_request:43254", "accept": true, - "reason": "Both PRs fix the same fixed_cross_entropy bug by adding support for passing through cross-entropy kwargs (weight/label_smoothing) to nn.functional.cross_entropy." + "reason": "Both PRs modify `fixed_cross_entropy` in `loss_utils.py` to pass through the same supported `cross_entropy` kwargs for the same issue #43240; they are the same underlying fix." } ] }, "evaluator_result": { "accept": true, - "feedback": "Grounded overall. The cluster is correctly characterized as mostly unrelated, and the 43251/43254 pairing is well supported by the shared issue target and identical fixed_cross_entropy kwargs fix. Minor caution: avoid calling 43251 the \u201cmore complete\u201d version unless you can point to a concrete diff advantage; the canonical-choice rationale is otherwise conservative enough." + "feedback": "The summary is grounded in the packet and stays conservative. The only accepted pair, PR #43251 and PR #43254, is well supported: both change `src/transformers/loss/loss_utils.py`, both target issue #43240, and both pass supported `cross_entropy` kwargs through the same helper. The other soft-edge verdicts are appropriately negative and do not overstate duplicate evidence." }, "evaluator_used": true, "retried": false, @@ -17753,7 +17892,7 @@ "soft_pairs": [ "pull_request:40546|pull_request:42781" ], - "prepared_review_unit_hash": "beeed39ef2a98081d8fab952f19cd05db2abff00b05e83a65b5c7e39504c5467", + "prepared_review_unit_hash": "c8fac57a1efdc76eb99f3247e320e4de735b1669ab9873136dc753462f6519cb", "component_budget": { "node_count": 2, "item_count": 2, @@ -17775,18 +17914,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are related VibeVoice PRs, but they are not the same change: #40546 adds the core VibeVoice implementation, while #42781 is a later draft for VibeVoice Realtime with extra tokenizer/generation work. I would keep them separate.", + "summary": "These are related VibeVoice PRs, but they are not duplicates: one adds the base VibeVoice implementation, while the other adds a separate realtime/acoustic-tokenizer extension on top of that family.", "confidence": 0.89, "canonical_issue_reason": null, - "canonical_pr_reason": "#40546 is the cleaner canonical representative of the core VibeVoice implementation: it is more mature, has substantial review activity, and centers on the base model integration rather than the later realtime extension.", + "canonical_pr_reason": "PR #40546 is the better canonical representative because it is the original non-draft VibeVoice implementation and covers the core model integration; #42781 is a later, narrower realtime-focused extension with additional tokenizer/generation work.", "best_issue_reason": null, - "best_pr_reason": "#40546 is the best standalone PR to represent the underlying VibeVoice addition; #42781 is a broader, draft realtime follow-up and does not look mergeable into the same single PR without changing scope.", + "best_pr_reason": "PR #40546 is the strongest overall representative of the cluster since it is broader, more mature, and closer to the primary VibeVoice landing; #42781 is a distinct follow-on change rather than the same PR.", "soft_edge_verdicts": [ { "left": "pull_request:40546", "right": "pull_request:42781", "accept": false, - "reason": "Shared filenames reflect the same model family, but #42781 adds VibeVoice Realtime and an acoustic tokenizer, so this is an extension of the original work rather than the same concrete PR change." + "reason": "They share the VibeVoice subsystem, but #42781 is a separate realtime/acoustic-tokenizer addition rather than the same concrete code-path fix or the same mergeable change as #40546." } ] }, @@ -17816,7 +17955,7 @@ "soft_pairs": [ "pull_request:40755|pull_request:41224" ], - "prepared_review_unit_hash": "087b8bfba874803eee7a338251fbebdfe9574580c9b7355cf5dc290706ea3e64", + "prepared_review_unit_hash": "d9b138efd317c52cd54378377030619618781764c8a96cefa0be603ff7f02336", "component_budget": { "node_count": 2, "item_count": 2, @@ -17838,18 +17977,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Two open PRs in different model families: one adds TimesFM covariate forecasting, the other adds DINOv3 image classification support. They share only a broad similarity of being feature additions with docs/tests, not the same underlying change.", - "confidence": 0.98, + "summary": "These two PRs are unrelated: one adds TimesFM covariate forecasting, the other adds DINOv3 image-classification support. They do not fix the same code path or represent one mergeable change.", + "confidence": 0.99, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 40755 is the more substantial and self-contained feature patch, with the larger diff and deeper implementation across the TimesFM forecasting stack.", + "canonical_pr_reason": "No canonical PR for the cluster; the PRs implement distinct features in different model families and are not duplicates.", "best_issue_reason": null, - "best_pr_reason": "PR 40755 is the best representative for the cluster only in a generic sense because it is the more expansive change; however, it is not a duplicate of PR 41224.", + "best_pr_reason": "No single best PR as a cluster representative because the items are heterogeneous and should not be grouped as one change.", "soft_edge_verdicts": [ { "left": "pull_request:40755", "right": "pull_request:41224", "accept": false, - "reason": "Different models and code paths: TimesFM covariate forecasting vs DINOv3 image classification support. The overlap is only superficial at the docs/tests level, so they should not be merged as duplicates." + "reason": "Different models, different features, and different files/code paths; they are not the same underlying bug or change." } ] }, @@ -17944,12 +18083,12 @@ "pull_request:41797|pull_request:45186", "pull_request:45186|pull_request:45630" ], - "prepared_review_unit_hash": "1d215039f0a21f7c374c833b6b9bd262ee35e78893bc1bc3c0ac9d6568d8397e", + "prepared_review_unit_hash": "682fdc1f74e4375464101dcc610536d03253489e616cf8791c6b0e4f6779ec6d", "component_budget": { "node_count": 3, "item_count": 3, "soft_pair_count": 2, - "serialized_chars": 10436, + "serialized_chars": 10435, "estimated_input_tokens": 2609, "estimated_eval_tokens": 5474 }, @@ -17957,7 +18096,7 @@ "node_count": 3, "item_count": 3, "soft_pair_count": 2, - "serialized_chars": 10436, + "serialized_chars": 10435, "estimated_input_tokens": 2609, "estimated_eval_tokens": 5474 }, @@ -17966,24 +18105,24 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are three separate model-addition PRs for different architectures (DeepSeek OCR, Isaac, and Kimi2-6), so they are not duplicates and should not be clustered together as the same change.", - "confidence": 0.99, + "summary": "These are three separate model-addition PRs (DeepSeek OCR, Isaac, Kimi2-6). They share common integration scaffolding, but each targets a different model and codepath, so they are not duplicates.", + "confidence": 0.98, "canonical_issue_reason": null, - "canonical_pr_reason": "No canonical PR: each PR introduces a different model integration with distinct files, tests, and codepaths.", + "canonical_pr_reason": null, "best_issue_reason": null, - "best_pr_reason": "No PR is a good cluster representative for deduplication because the changes target different models and are not mergeable into one concrete fix/change.", + "best_pr_reason": "Kimi2-6 is the broadest and most mature-looking integration here, with the highest review activity and the largest set of model/auto-registration changes, but it is still a distinct PR rather than a duplicate of the others.", "soft_edge_verdicts": [ { "left": "pull_request:41797", "right": "pull_request:45186", "accept": false, - "reason": "Different model additions (DeepSeek OCR vs Isaac) with separate implementations and tests; only broad shared scaffolding like auto registries/docs overlaps." + "reason": "Different models and implementations: DeepSeek OCR vs Isaac. The overlap is only in generic model-registration/docs boilerplate, not the same underlying change." }, { "left": "pull_request:45186", "right": "pull_request:45630", "accept": false, - "reason": "Different model additions (Isaac vs Kimi2-6) and distinct model codepaths; shared framework registration files are not enough to treat them as the same PR." + "reason": "Different model families and concrete integrations: Isaac vs Kimi2-6. Shared auto-mapping/docs scaffolding is too broad to treat as the same PR." } ] }, @@ -18013,7 +18152,7 @@ "soft_pairs": [ "pull_request:42975|pull_request:42976" ], - "prepared_review_unit_hash": "1adc326c895d2be58a51537b2d698e4493bdfcd52313ba295b041d4ba421f7c7", + "prepared_review_unit_hash": "08e51c95551a9519460377214bcdc8027bc0d4af16c3ae606be57ab1ad7fe9d9", "component_budget": { "node_count": 2, "item_count": 2, @@ -18035,18 +18174,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are related only at a broad maintenance level: both update GitHub Actions in workflow files, but they address different dependencies and change different code paths. They should not be treated as duplicates or merged into one PR.", - "confidence": 0.93, + "summary": "Both items are broad GitHub Actions upgrade PRs, but they target different action families and different concerns: Node 24 compatibility vs latest-version bumps for docker/build-push-action. They are related in scope yet not the same concrete change, so they should not be deduplicated.", + "confidence": 0.95, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 42975 is the better canonical PR because it is the broader workflow-maintenance change focused on GitHub Actions runtime compatibility across many workflows.", + "canonical_pr_reason": "PR 42975 is the better representative because its title and edits are more specifically about the compatibility-driven workflow action upgrade theme, which is the closer cluster anchor here.", "best_issue_reason": null, - "best_pr_reason": "PR 42975 is the best representative of the cluster since it has the wider scope and clearer unifying intent, while 42976 is a separate dependency bump for build-related actions.", + "best_pr_reason": "PR 42975 is the strongest single representative of the cluster because it touches many workflow actions in a clearly defined compatibility upgrade, while 42976 is a separate version-bump effort focused on docker build actions.", "soft_edge_verdicts": [ { "left": "pull_request:42975", "right": "pull_request:42976", "accept": false, - "reason": "Different underlying changes: 42975 upgrades general workflow actions for Node 24 compatibility, while 42976 bumps docker/build-push-action and a few other actions. They overlap in workflow files but do not fix the same concrete problem and would not plausibly be merged into one PR." + "reason": "Related workflow maintenance, but they update different actions for different purposes and would not naturally be merged into one concrete fix." } ] }, @@ -18085,76 +18224,76 @@ "pull_request:43451|pull_request:45490", "pull_request:45490|pull_request:46162" ], - "prepared_review_unit_hash": "1e70934457c9c93d81599e564902385d3a36446517f356e4730f750329b55a4d", + "prepared_review_unit_hash": "9a01b9fe7c7b4db8bb3d75736c2724a48e77859d31b1b6f644b71f629161cb22", "component_budget": { "node_count": 5, "item_count": 5, "soft_pair_count": 7, - "serialized_chars": 16485, - "estimated_input_tokens": 4122, - "estimated_eval_tokens": 8500 + "serialized_chars": 16554, + "estimated_input_tokens": 4139, + "estimated_eval_tokens": 8534 }, "budget": { "node_count": 5, "item_count": 5, "soft_pair_count": 7, - "serialized_chars": 16485, - "estimated_input_tokens": 4122, - "estimated_eval_tokens": 8500 + "serialized_chars": 16554, + "estimated_input_tokens": 4139, + "estimated_eval_tokens": 8534 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These PRs are mostly unrelated model additions or a small test-only tweak; they share only broad Transformers boilerplate like docs/auto registry updates, not the same underlying change or bug.", + "summary": "These PRs are all separate additions or test-only changes across different models and code paths; the shared files are mostly boilerplate model-registration/docs touchpoints, not evidence of duplicate underlying work.", "confidence": 0.97, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 43451 (Molmo2) is the most substantial and discussion-heavy submission, so it is the best anchor among this set even though it is not a duplicate of the others.", + "canonical_pr_reason": "PR 43451 is the most substantial and actively discussed item in the cluster, but it is not a duplicate of the others; it is just the best representative if one must be chosen.", "best_issue_reason": null, - "best_pr_reason": "PR 43451 is the strongest representative of the cluster because it has the broadest model implementation surface and the most review activity; the other PRs are separate model or test changes.", + "best_pr_reason": "PR 43451 has the broadest scope and highest review activity, making it the strongest standalone representative of the cluster, though not a duplicate target.", "soft_edge_verdicts": [ { "left": "pull_request:42978", "right": "pull_request:43448", "accept": false, - "reason": "Different model families (ViT NEPA vs Molmo) with separate implementations; overlap is only generic docs/auto-registration boilerplate." + "reason": "Different model additions: ViT NEPA vs Molmo. Overlap is limited to shared registration/docs scaffolding, not the same code-path change." }, { "left": "pull_request:42978", "right": "pull_request:43451", "accept": false, - "reason": "Different model families (ViT NEPA vs Molmo2); shared files are only common registry/docs paths, not the same code-path fix." + "reason": "Different model families and implementations: ViT NEPA vs Molmo2. Shared auto/docs edits are incidental boilerplate." }, { "left": "pull_request:42978", "right": "pull_request:45490", "accept": false, - "reason": "Unrelated additions for ViT NEPA and CTSM; no shared concrete bug or mergeable combined change." + "reason": "Completely different domains: a vision model addition vs a time-series model addition. No same bug or mergeable single change." }, { "left": "pull_request:43448", "right": "pull_request:43451", "accept": false, - "reason": "Molmo vs Molmo2 are related by naming, but they are distinct model implementations and not the same change." + "reason": "Molmo and Molmo2 are related names but are distinct model releases with separate configs, processors, and tests; not the same underlying change." }, { "left": "pull_request:43448", "right": "pull_request:45490", "accept": false, - "reason": "Different model additions (Molmo vs CTSM) with only broad transformer plumbing in common." + "reason": "Molmo model addition vs CTSM model addition; only broad model-registry files overlap, which is not enough to treat them as duplicates." }, { "left": "pull_request:43451", "right": "pull_request:45490", "accept": false, - "reason": "Molmo2 and CTSM are separate model ports; shared auto/model registry edits do not indicate one underlying fix." + "reason": "Molmo2 and CTSM are unrelated model additions touching different architectures and files; shared auto-registration edits are superficial." }, { "left": "pull_request:45490", "right": "pull_request:46162", "accept": false, - "reason": "CTSM model addition vs a tiny test annotation change for Florence2/LFM2-VL; completely different scope and code paths." + "reason": "CTSM model integration vs test annotation changes for Florence2/LFM2-VL. The latter is test maintenance, not the same code-path fix or feature." } ] }, @@ -18189,7 +18328,7 @@ "pull_request:43665|pull_request:46102", "pull_request:43665|pull_request:44815" ], - "prepared_review_unit_hash": "552a56ef3a33c44978d01532890db90aeec766cade8a9035bae576c45bcea7e6", + "prepared_review_unit_hash": "4155d50ded71537fe1e003f7367ed45759fc4545aef5582129c953827f68724f", "component_budget": { "node_count": 4, "item_count": 4, @@ -18211,36 +18350,36 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These four PRs are not duplicates of one another: they target unrelated areas (Whisper decoding, CLIP/ViT test stubs, FP8 dequant/loading, and FSDP tied-weights). The cluster is heterogeneous, with 44815 the most substantive standalone fix.", - "confidence": 0.95, + "summary": "These are not duplicates: the PRs target unrelated areas (Whisper decoding, CLIP/ViT test edits, FP8 dequantization, and FSDP tied-weights handling). The only commonality is generic \"fix\" wording and test touchpoints in some diffs.", + "confidence": 0.96, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 44815 is the strongest canonical anchor if one must be chosen: it has a coherent code fix across loading/FP8 paths and clear test coverage, unlike the other PRs which are either unrelated or appear to be test-only noise.", + "canonical_pr_reason": "pull_request:46102 is the clearest focused bug fix with a concrete code path (FSDP tied weights / model loading), but it is not a duplicate of the others.", "best_issue_reason": null, - "best_pr_reason": "PR 44815 is the best overall PR in the set because it addresses a concrete implementation bug in a specific code path and is materially more substantial than the test-only or unrelated PRs.", + "best_pr_reason": "pull_request:46102 is the strongest standalone PR: specific title, cohesive FSDP scope, and a focused fix path compared with the more generic or test-heavy alternatives.", "soft_edge_verdicts": [ { "left": "pull_request:43291", "right": "pull_request:43665", "accept": false, - "reason": "Whisper tokenizer decode logic and CLIP/ViT test edits are unrelated; they do not appear to fix the same bug or change." + "reason": "Different subsystems and problems: Whisper tokenizer decoding vs CLIP/ViT test edits. Superficial \"fix\" similarity only." }, { "left": "pull_request:43291", "right": "pull_request:44815", "accept": false, - "reason": "Whisper timestamp decoding is unrelated to FP8 dequant/loading and Mistral-related changes; different code paths and problem domains." + "reason": "Whisper timestamp decoding and FP8 dequant/model-loading changes are unrelated code paths; no same underlying bug." }, { "left": "pull_request:43665", "right": "pull_request:46102", "accept": false, - "reason": "The CLIP/ViT test-only edits are not the same underlying change as the FSDP tied-weights fix; shared test noise is not enough to merge them." + "reason": "Test-file edits for vision models are not the same issue as FSDP tied-weights/model-loading changes." }, { "left": "pull_request:43665", "right": "pull_request:44815", "accept": false, - "reason": "Although both are labeled like generic fixes and touch tests, they address unrelated subsystems and different bugs; they should not be merged as duplicates." + "reason": "These touch unrelated areas (vision tests vs FP8 quantization/loading). Shared generic wording is insufficient." } ] }, @@ -18277,73 +18416,73 @@ "pull_request:45932|pull_request:46100", "pull_request:46100|pull_request:46114" ], - "prepared_review_unit_hash": "53c476e0647c7ab85421cf6bc37d8f3a4abb50f9d282f6e2f86910978311f4d1", + "prepared_review_unit_hash": "7877d7166add60e71b11c9f8b09302c913da14e5c60f60271fa9f28e0c9320be", "component_budget": { "node_count": 5, "item_count": 5, "soft_pair_count": 5, - "serialized_chars": 11338, - "estimated_input_tokens": 2835, - "estimated_eval_tokens": 5926 + "serialized_chars": 11341, + "estimated_input_tokens": 2836, + "estimated_eval_tokens": 5928 }, "budget": { "node_count": 5, "item_count": 5, "soft_pair_count": 5, - "serialized_chars": 11338, - "estimated_input_tokens": 2835, - "estimated_eval_tokens": 5926 + "serialized_chars": 11341, + "estimated_input_tokens": 2836, + "estimated_eval_tokens": 5928 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Only one pair looks like a true duplicate: PRs 46100 and 46114 are effectively the same ViT test-migration change with identical edits to the same file. The other soft-similarity pairs share superficial test/workflow or bot-check traits, but they do not appear to fix or introduce the same concrete change.", - "confidence": 0.94, + "summary": "This cluster is mostly synthetic or test-only PRs. PR 45932 is the only substantive implementation change and is the best representative. The one true duplicate pair is 46100 and 46114, which have identical ViT test edits; the other soft-similarity pairs do not look like the same underlying change.", + "confidence": 0.82, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 46114 is the best representative of the only real duplicate pair: it has the same diff as 46100, the same failing-test additions in tests/models/vit/test_modeling_vit.py, and a clearer title indicating the forked OTEL migration test intent.", + "canonical_pr_reason": "PR 45932 is the best canonical PR because it contains the real multi-file implementation/config/CI change, while the others are either synthetic test PRs or unrelated workflow/test edits.", "best_issue_reason": null, - "best_pr_reason": "PR 46114 is the strongest global PR candidate because it is the clearest, most descriptive instance of the duplicated ViT test-migration change; PR 46100 is effectively equivalent, while the other PRs are unrelated test/bot/workflow probes.", + "best_pr_reason": "PR 45932 is the clearest representative of the cluster\u2019s substantive change set and is more useful as the cluster anchor than the synthetic duplicates.", "soft_edge_verdicts": [ { "left": "pull_request:43488", "right": "pull_request:46114", "accept": false, - "reason": "Both touch tests/models/vit/test_modeling_vit.py, but 43488 is a repo-bot/setup-format probe and 46114 is a forked OTEL test-migration PR; they are not the same underlying change." + "reason": "Both touch ViT tests, but 43488 also changes setup/dependency files and has a different purpose; the diffs are not the same underlying change." }, { "left": "pull_request:43488", "right": "pull_request:43532", "accept": false, - "reason": "These are different test/bot experiments: 43488 changes setup/versioning and dependency lists, while 43532 rewires CI workflows and testing utilities; shared test-file references are incidental." + "reason": "Shared test-file overlap is too weak: 43488 is setup/version/dependency-oriented, while 43532 is workflow/testing-utils oriented." }, { "left": "pull_request:43532", "right": "pull_request:46114", "accept": false, - "reason": "43532 is about CI workflow/show-diff plumbing, not the ViT test-migration change in 46114, so they do not target the same concrete code-path problem." + "reason": "43532 changes workflows and testing utilities, whereas 46114 is a forked ViT test-migration PR; they are not the same code change." }, { "left": "pull_request:45932", "right": "pull_request:46100", "accept": false, - "reason": "45932 is a broad Grafana/OTEL instrumentation and model-update PR across many files; 46100 is a tiny ViT test-migration edit. Same general testing theme, but not the same change or bug." + "reason": "These are different changes: 45932 updates model/config/CI code across many files, while 46100 only injects failing assertions into tests/models/vit/test_modeling_vit.py." }, { "left": "pull_request:46100", "right": "pull_request:46114", "accept": true, - "reason": "These PRs are near-identical: same file, same inserted failing assertions, same test-migration intent, and they could plausibly be merged into one PR." + "reason": "These appear to be the same synthetic ViT test PR: same file, same inserted assertions, and matching diff preview, so they could plausibly be merged as one PR." } ] }, "evaluator_result": { - "accept": true, - "feedback": "The output is grounded in the packet and stays conservative on the soft-edge pairs. The only clearly mergeable duplicate is PR 46100 and PR 46114, which have identical diffs to the same ViT test file. The other rejected pairs are reasonably distinguished as bot/workflow/setup experiments or a broad instrumentation PR versus a tiny test-migration PR, so there is no overstatement of evidence." + "accept": false, + "feedback": "The summary is not fully grounded in the packet. Most notably, the packet marks pull_request:45932 | pull_request:46100 as deterministic_accept=true, but the analyst says that pair is not the same underlying change and instead calls 46100 | 46114 the only true duplicate pair. That directly conflicts with the provided pair evidence. Also, the claim that 46100 and 46114 are the same synthetic PR is stronger than the packet supports, since their evidence is only soft_similarity, not deterministic acceptance. The representative choice of 45932 is fine, but the soft-edge verdicts and duplicate-pair summary need to be more conservative." }, "evaluator_used": true, - "retried": false, + "retried": true, "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null @@ -18367,12 +18506,12 @@ "soft_pairs": [ "pull_request:43612|pull_request:43613" ], - "prepared_review_unit_hash": "602b64391963f9071bae2975b8985ec4b03ddb051eb1b125b3f325b26695b36c", + "prepared_review_unit_hash": "b99469c6ee2b9bb25e897c96d8d68749c688fae68bad626cfd307cd1df2a7d67", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 6082, + "serialized_chars": 6083, "estimated_input_tokens": 1521, "estimated_eval_tokens": 3298 }, @@ -18380,7 +18519,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 6082, + "serialized_chars": 6083, "estimated_input_tokens": 1521, "estimated_eval_tokens": 3298 }, @@ -18389,18 +18528,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are related but distinct PRs: one adds a Promptable Concept Segmentation pipeline, the other adds a Promptable Visual Segmentation pipeline. They share scaffolding/docs/registry updates, but target different pipeline code paths and model families.", - "confidence": 0.96, + "summary": "Two feature PRs with overlapping scaffolding, but they implement different pipelines and different underlying segmentation tasks; they should not be merged as duplicates.", + "confidence": 0.94, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 43613 is the broader, more representative pipeline addition in this cluster, with integration across multiple segmentation model families and corresponding pipeline/tests updates.", + "canonical_pr_reason": "No canonical duplicate PR: 43612 adds Promptable Concept Segmentation, while 43613 adds Promptable Visual Segmentation; they target different code paths and model integrations.", "best_issue_reason": null, - "best_pr_reason": "PR 43613 is the stronger standalone change because it covers a wider concrete segmentation integration; PR 43612 is a separate narrower feature and not a duplicate.", + "best_pr_reason": "PR 43613 is the broader of the two feature additions and touches more model/test integration, so it is the better representative if one must be chosen, though it is still not a duplicate of 43612.", "soft_edge_verdicts": [ { "left": "pull_request:43612", "right": "pull_request:43613", "accept": false, - "reason": "Different features: concept segmentation vs visual segmentation. Shared docs/registration/test scaffolding is expected, but the underlying pipeline behavior and model targets differ, so they should not be merged as one PR." + "reason": "Different pipelines/features (concept segmentation vs visual segmentation) with different model and test coverage; shared docs/boilerplate files are not enough to treat them as the same change." } ] }, @@ -18430,7 +18569,7 @@ "soft_pairs": [ "pull_request:43651|pull_request:43663" ], - "prepared_review_unit_hash": "261b0d4826169dcaf7da6fb46f3e84e58f267c66e0825a3e80c5ed7d5a4b6d96", + "prepared_review_unit_hash": "cdc471dec3cf0047c73971c6e64d50ff26d19b9750dadea2aade31cc0ada433d", "component_budget": { "node_count": 2, "item_count": 2, @@ -18452,18 +18591,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Both PRs touch `trainer.py` and add custom-trainer extension points, but they solve different problems: one changes gradient-accumulation loss scaling, the other refactors signature-column filtering. They are not the same underlying change.", - "confidence": 0.94, + "summary": "Two open Trainer PRs touch the same file but implement different extension points: one changes gradient-accumulation loss scaling, the other refactors signature-column filtering for custom trainers. They are not the same underlying change.", + "confidence": 0.95, "canonical_issue_reason": null, - "canonical_pr_reason": "No canonical PR: these are separate trainer enhancements with different code paths and semantics, so they should not be collapsed into one duplicate.", + "canonical_pr_reason": "PR 43651 is the more concrete standalone fix: it changes loss scaling behavior in training_step and adds a supporting property. PR 43663 is a separate refactor for column-filtering extensibility.", "best_issue_reason": null, - "best_pr_reason": "PR 43651 is the more concrete bug-facing change and the more likely standalone fix; 43663 is a separate extensibility refactor.", + "best_pr_reason": "PR 43651 is the better representative because it has a narrowly defined behavior change and clearer bug-fix intent than the column-filtering refactor in 43663.", "soft_edge_verdicts": [ { "left": "pull_request:43651", "right": "pull_request:43663", "accept": false, - "reason": "Different trainer behaviors: 43651 controls loss scaling during gradient accumulation, while 43663 exposes signature-column filtering. Shared file alone is not enough to treat them as duplicates." + "reason": "Different code paths and goals: 43651 adjusts gradient accumulation loss normalization, while 43663 adds an overrideable helper for signature-column filtering. Shared file alone is not enough to treat them as duplicates." } ] }, @@ -18493,12 +18632,12 @@ "soft_pairs": [ "pull_request:43656|pull_request:43842" ], - "prepared_review_unit_hash": "b26170d4ab6321e25ef148691917b9ce54e312bb0a84092d030faf416950f305", + "prepared_review_unit_hash": "5963cc1b90f71e31551fbf24df588b6326f9a971b0e7be3981afc138a5c72c8d", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 3663, + "serialized_chars": 3664, "estimated_input_tokens": 916, "estimated_eval_tokens": 2088 }, @@ -18506,7 +18645,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 3663, + "serialized_chars": 3664, "estimated_input_tokens": 916, "estimated_eval_tokens": 2088 }, @@ -18515,27 +18654,24 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "The two PRs are related in the same file and both touch typing/import-time behavior in `src/transformers/cli/serve.py`, but they do not clearly fix the same bug. PR #43656 is a broader `from __future__ import annotations` / annotation-cleanup refactor, while PR #43842 is a narrower `TypeAdapter` import/typing fix tied to a different explicit issue target. They are similar, but not clearly mergeable into one bugfix PR.", - "confidence": 0.89, + "summary": "Two open PRs touch the same file and the same general TypeAdapter annotation problem in `src/transformers/cli/serve.py`, but they use different fixes and appear to target different concrete failure modes, so they should not be collapsed as duplicates.", + "confidence": 0.78, "canonical_issue_reason": null, - "canonical_pr_reason": "PR #43842 is the better canonical PR because it is the more focused fix for the `TypeAdapter`/`pydantic` NameError path, with a narrower scope and clearer bug target than the broader typing refactor in PR #43656.", + "canonical_pr_reason": "PR #43656 is the stronger canonical representative: it has the explicit issue link, more discussion/review activity, and a broader runtime-safe typing cleanup in the CLI serve code path.", "best_issue_reason": null, - "best_pr_reason": "PR #43842 is the best representative of the cluster since it addresses a specific runtime/import typing problem in the CLI with minimal changes, whereas PR #43656 is a wider annotation modernization that is less clearly the same fix.", + "best_pr_reason": "PR #43842 is the more narrowly targeted fix for the specific pydantic-absent `TypeAdapter` NameError, but it is less established than #43656 and appears to be an alternate implementation rather than the same change.", "soft_edge_verdicts": [ { "left": "pull_request:43656", "right": "pull_request:43842", "accept": false, - "reason": "They share `src/transformers/cli/serve.py` and both concern typing/import-time issues, but #43656 is a broader annotation refactor and #43842 is a narrower `TypeAdapter`/`pydantic` fix with a different explicit issue target. The overlap is similarity, not clear same-bug identity." + "reason": "Both are in `cli/serve.py` and mention `TypeAdapter`, but the diffs show different remedies (`from __future__ import annotations`/typing cleanup vs `TYPE_CHECKING` import plus quoted annotation) and different issue targets, so they do not look like the same concrete PR change." } ] }, - "evaluator_result": { - "accept": false, - "feedback": "The summary overstates the relationship. The packet only supports a loose similarity in `src/transformers/cli/serve.py`, not clear evidence that both PRs fix the same bug or are naturally mergeable into one bugfix PR. PR #43656 is a broader `from __future__ import annotations` / typing refactor, while PR #43842 is a narrower `TYPE_CHECKING`/`TypeAdapter` fix tied to a different explicit issue target. The conservatively grounded conclusion is that they are related, but not clearly the same fix." - }, - "evaluator_used": true, - "retried": true, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, "accepted_nontrivial_soft_edge": false, "error_kind": null, "error_message": null @@ -18678,8 +18814,8 @@ "pull_request:44098|pull_request:44101", "pull_request:44059|pull_request:44722", "pull_request:44018|pull_request:44071", - "pull_request:44017|pull_request:44722", "pull_request:44017|pull_request:44161", + "pull_request:44017|pull_request:44722", "pull_request:44017|pull_request:44073", "pull_request:44068|pull_request:44071", "pull_request:43997|pull_request:44073", @@ -18689,8 +18825,8 @@ "pull_request:44056|pull_request:44101", "pull_request:44059|pull_request:44071", "pull_request:44073|pull_request:44161", - "pull_request:44018|pull_request:44059", "pull_request:44059|pull_request:44068", + "pull_request:44018|pull_request:44059", "pull_request:44101|pull_request:44161", "pull_request:43997|pull_request:44017", "pull_request:43997|pull_request:44056", @@ -18706,8 +18842,8 @@ "pull_request:44010|pull_request:44073", "pull_request:44018|pull_request:44129", "pull_request:43997|pull_request:44074", - "pull_request:43997|pull_request:44129", "pull_request:44071|pull_request:44129", + "pull_request:43997|pull_request:44129", "pull_request:44010|pull_request:44018", "pull_request:43996|pull_request:44161", "pull_request:43996|pull_request:44017", @@ -18721,20 +18857,20 @@ "pull_request:44044|pull_request:44129", "pull_request:44044|pull_request:44154" ], - "prepared_review_unit_hash": "10ebb46c13e98957fdb20a25988d8d0304f7d9986cbee19f6ace094be30f872c", + "prepared_review_unit_hash": "75773f077841fe924b1af82b6bb723107ba20ff9dcaef128f3441637aeaa8bdc", "component_budget": { "node_count": 28, "item_count": 28, "soft_pair_count": 78, - "serialized_chars": 84897, - "estimated_input_tokens": 21225, - "estimated_eval_tokens": 42706 + "serialized_chars": 84901, + "estimated_input_tokens": 21226, + "estimated_eval_tokens": 42708 }, "budget": { "node_count": 27, "item_count": 27, "soft_pair_count": 72, - "serialized_chars": 75094, + "serialized_chars": 75096, "estimated_input_tokens": 18774, "estimated_eval_tokens": 37804 }, @@ -18743,400 +18879,352 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly independent per-model output-tracing refactors. I\u2019m only treating the strongest GPT-J/CodeGen lineage as duplicate-adjacent; same-issue or same-subsystem similarity alone is not enough for the rest.", - "confidence": 0.67, - "canonical_issue_reason": "Issue 43979 is the shared tracking issue explicitly referenced by the output-tracing refactors and best matches the cluster\u2019s common theme.", - "canonical_pr_reason": "PR 44722 is the clearest representative of the GPT-J/CodeGen refactor lineage: it spans the same core path and appears to be the later, broader revision of that change.", - "best_issue_reason": "43979 is the best global issue anchor because it is the explicit issue target recurring across the refactor PRs.", - "best_pr_reason": "44722 is the best global PR representative because it most directly captures the GPT-J/CodeGen standardized-output refactor and overlaps the same concrete code path.", + "summary": "This cluster is dominated by model-specific output-tracing refactors. The only clear duplicate groups are the GPT-Neo pair and the GPT-J family (including the later GPT-J/CodeGen follow-up); the rest are same-theme but different model/code paths.", + "confidence": 0.84, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 44722 is the strongest representative of the GPT-J output-tracing refactor cluster: it is the latest and most complete decorator-based migration on the same code path, with the clearest overlap to the earlier GPT-J PRs.", + "best_issue_reason": null, + "best_pr_reason": "PR 44722 best represents the duplicate cluster because it most completely captures the shared GPT-J output-tracing change and also subsumes the related CodeGen work.", "soft_edge_verdicts": [ { "left": "pull_request:43996", "right": "pull_request:44085", "accept": false, - "reason": "Different model families and code paths; only the broad output-tracing theme overlaps." + "reason": "Different model families and code paths; same output-tracing theme is too broad." }, { "left": "pull_request:43996", "right": "pull_request:44044", "accept": false, - "reason": "FNet/CVT versus DeBERTa; unrelated implementations." + "reason": "Unrelated models (FNet/CVT vs DeBERTa); not the same underlying change." }, { "left": "pull_request:44066", "right": "pull_request:44085", - "accept": false, - "reason": "Both are GPT-J-related, but 44066 is the broader GPTJ+CodeGen variant and the packet does not support calling this the same concrete change." + "accept": true, + "reason": "Both are GPT-J output-tracing refactors on the same model code path; the later PR is a broader variant, but the underlying change matches." }, { "left": "pull_request:44007", "right": "pull_request:44072", "accept": false, - "reason": "RegNet and EfficientNet are separate model implementations." + "reason": "Different vision models and different implementations; only the refactor pattern is shared." }, { "left": "pull_request:44072", "right": "pull_request:44722", "accept": false, - "reason": "EfficientNet refactor versus GPT-J/CodeGen refactor; same tracking issue is not enough." + "reason": "EfficientNet and GPT-J/CodeGen are unrelated; same refactor style does not make them duplicates." }, { "left": "pull_request:44066", "right": "pull_request:44072", "accept": false, - "reason": "GPTJ/CodeGen and EfficientNet are unrelated code paths." + "reason": "GPT-J/CodeGen vs EfficientNet; different code paths and model families." }, { "left": "pull_request:44013", "right": "pull_request:44044", "accept": false, - "reason": "MobileNetV2 versus DeBERTa; different model-specific changes." + "reason": "MobileNetV2 and DeBERTa are unrelated models." }, { "left": "pull_request:44066", "right": "pull_request:44086", "accept": false, - "reason": "GPTJ/CodeGen refactor versus MGP-STR refactor; different code paths." + "reason": "GPT-J/CodeGen vs MGP-STR; not the same bug/change." }, { "left": "pull_request:44066", "right": "pull_request:44071", "accept": false, - "reason": "GPTJ/CodeGen refactor versus MPT refactor; separate model code." + "reason": "GPT-J/CodeGen and MPT are different model implementations." }, { "left": "pull_request:44018", "right": "pull_request:44068", - "accept": false, - "reason": "Related GPT-Neo output-tracing refactors, but the packet does not give enough support to treat them as the same concrete change." + "accept": true, + "reason": "Same GPT-Neo output-tracing migration on the same model family and code path." }, { "left": "pull_request:44066", "right": "pull_request:44068", "accept": false, - "reason": "GPTJ/CodeGen versus GPT-Neo; similar theme, different code paths." + "reason": "GPT-J/CodeGen and GPT-Neo are unrelated model code paths." }, { "left": "pull_request:44018", "right": "pull_request:44066", "accept": false, - "reason": "GPT-Neo versus GPTJ/CodeGen are different refactors." + "reason": "Different model families (GPT-Neo vs GPT-J/CodeGen); same issue target is not enough." }, { "left": "pull_request:44019", "right": "pull_request:44722", "accept": false, - "reason": "ResNet refactor and GPTJ/CodeGen refactor are unrelated implementations." + "reason": "ResNet and GPT-J/CodeGen are unrelated; no shared concrete bug/change." }, { "left": "pull_request:44086", "right": "pull_request:44722", "accept": false, - "reason": "MGP-STR and GPTJ/CodeGen touch different model paths." + "reason": "MGP-STR and GPT-J/CodeGen are unrelated code paths." }, { "left": "pull_request:44019", "right": "pull_request:44085", "accept": false, - "reason": "ResNet versus GPT-J; unrelated code paths." + "reason": "ResNet and GPT-J are unrelated model implementations." }, { "left": "pull_request:44018", "right": "pull_request:44019", "accept": false, - "reason": "GPT-Neo and ResNet are unrelated model implementations." + "reason": "GPT-Neo and ResNet are unrelated; only the refactor theme overlaps." }, { "left": "pull_request:44019", "right": "pull_request:44071", "accept": false, - "reason": "ResNet versus MPT; same broad pattern, not the same change." + "reason": "ResNet and MPT are unrelated models and code paths." }, { "left": "pull_request:44068", "right": "pull_request:44116", "accept": false, - "reason": "GPT-Neo versus Flaubert/XLM; different model families." + "reason": "GPT-Neo and Flaubert are unrelated model families." }, { "left": "pull_request:44068", "right": "pull_request:44722", "accept": false, - "reason": "GPT-Neo and GPTJ/CodeGen are different model code paths." + "reason": "GPT-Neo and GPT-J/CodeGen are different model families and refactors." }, { "left": "pull_request:44018", "right": "pull_request:44116", "accept": false, - "reason": "GPT-Neo and Flaubert/XLM are unrelated refactors." + "reason": "GPT-Neo and Flaubert are unrelated code paths." }, { "left": "pull_request:44071", "right": "pull_request:44722", "accept": false, - "reason": "MPT and GPTJ/CodeGen are different model implementations." + "reason": "MPT and GPT-J/CodeGen are unrelated model implementations." }, { "left": "pull_request:44018", "right": "pull_request:44086", "accept": false, - "reason": "GPT-Neo versus MGP-STR are separate model code paths." + "reason": "GPT-Neo and MGP-STR are unrelated; same output-tracing pattern is too broad." }, { "left": "pull_request:44018", "right": "pull_request:44085", "accept": false, - "reason": "GPT-Neo and GPT-J are not the same fix." + "reason": "GPT-Neo and GPT-J are different model code paths." }, { "left": "pull_request:44068", "right": "pull_request:44086", "accept": false, - "reason": "GPT-Neo and MGP-STR are separate model-specific changes." + "reason": "GPT-Neo and MGP-STR are unrelated." }, { "left": "pull_request:44071", - "right": "pull_request:44085", + "right": "pull_request:44086", "accept": false, - "reason": "MPT and GPT-J are different model implementations." + "reason": "MPT and MGP-STR are unrelated model refactors." }, { "left": "pull_request:44059", "right": "pull_request:44086", "accept": false, - "reason": "GPT2 refactor versus MGP-STR refactor; not the same change." + "reason": "GPT2 and MGP-STR are unrelated code paths." }, { "left": "pull_request:44085", "right": "pull_request:44086", "accept": false, - "reason": "GPT-J and MGP-STR are different model code paths." + "reason": "GPT-J and MGP-STR are unrelated models." }, { "left": "pull_request:44098", "right": "pull_request:44101", "accept": false, - "reason": "ViLT and Flaubert/XLM refactors are unrelated." + "reason": "ViLT and Flaubert/XLM are unrelated model families." }, { "left": "pull_request:44059", "right": "pull_request:44722", "accept": false, - "reason": "GPT2 and GPTJ/CodeGen are separate model refactors." + "reason": "GPT2 and GPT-J/CodeGen are different model implementations." }, { "left": "pull_request:44018", "right": "pull_request:44071", "accept": false, - "reason": "GPT-Neo and MPT are different model implementations." + "reason": "GPT-Neo and MPT are unrelated code paths." }, { - "left": "pull_request:44017", - "right": "pull_request:44722", - "accept": false, - "reason": "Segformer refactor versus GPTJ/CodeGen refactor; unrelated code paths." - }, - { - "left": "pull_request:44017", - "right": "pull_request:44161", + "left": "pull_request:44059", + "right": "pull_request:44068", "accept": false, - "reason": "Segformer and LongT5 are different model families." + "reason": "GPT2 and GPT-Neo are different model families." }, { - "left": "pull_request:44017", - "right": "pull_request:44073", + "left": "pull_request:44071", + "right": "pull_request:44085", "accept": false, - "reason": "Segformer and VisualBert are not the same underlying change." + "reason": "MPT and GPT-J are unrelated models." }, { - "left": "pull_request:44068", + "left": "pull_request:44059", "right": "pull_request:44071", "accept": false, - "reason": "GPT-Neo and MPT are distinct model implementations." - }, - { - "left": "pull_request:43997", - "right": "pull_request:44073", - "accept": false, - "reason": "RegNet and VisualBert are unrelated model refactors." - }, - { - "left": "pull_request:44071", - "right": "pull_request:44076", - "accept": false, - "reason": "MPT and ImageGPT touch different code paths." + "reason": "GPT2 and MPT are unrelated code paths." }, { "left": "pull_request:44073", - "right": "pull_request:44101", + "right": "pull_request:44161", "accept": false, - "reason": "VisualBert and Flaubert/XLM are separate model-specific changes." + "reason": "VisualBERT and LongT5 are unrelated models." }, { - "left": "pull_request:44017", - "right": "pull_request:44071", + "left": "pull_request:44018", + "right": "pull_request:44059", "accept": false, - "reason": "Segformer versus MPT; same pattern, different models." + "reason": "GPT-Neo and GPT2 are different model implementations." }, { - "left": "pull_request:44056", - "right": "pull_request:44101", + "left": "pull_request:44101", + "right": "pull_request:44161", "accept": false, - "reason": "MPNet refactor versus Flaubert/XLM refactor; unrelated code paths." + "reason": "Flaubert/XLM and LongT5 are unrelated code paths." }, { "left": "pull_request:43997", "right": "pull_request:44017", "accept": false, - "reason": "RegNet and Segformer are unrelated model implementations." + "reason": "RegNet and SegFormer are different vision models." }, { "left": "pull_request:43997", "right": "pull_request:44056", "accept": false, - "reason": "RegNet versus MPNet; different code paths." + "reason": "RegNet and MPNet are unrelated models." }, { "left": "pull_request:44071", "right": "pull_request:44076", "accept": false, - "reason": "MPT and ImageGPT are separate model refactors." - }, - { - "left": "pull_request:44010", - "right": "pull_request:44071", - "accept": false, - "reason": "SqueezeBert and MPT are different model code paths." - }, - { - "left": "pull_request:44010", - "right": "pull_request:44073", - "accept": false, - "reason": "SqueezeBert versus VisualBert; not the same fix." - }, - { - "left": "pull_request:44018", - "right": "pull_request:44129", - "accept": false, - "reason": "GPT-Neo refactor versus SpeechT5 refactor; separate implementations." + "reason": "MPT and ImageGPT are unrelated model implementations." }, { "left": "pull_request:44074", "right": "pull_request:44101", "accept": false, - "reason": "TextNet and Flaubert/XLM are different model families." + "reason": "TextNet and Flaubert/XLM are unrelated code paths." }, { "left": "pull_request:44056", "right": "pull_request:44161", "accept": false, - "reason": "MPNet refactor versus LongT5 refactor; unrelated code paths." + "reason": "MPNet and LongT5 are unrelated models." }, { "left": "pull_request:43997", "right": "pull_request:44018", "accept": false, - "reason": "RegNet and GPT-Neo are unrelated model implementations." + "reason": "RegNet and GPT-Neo are unrelated model families." }, { "left": "pull_request:43997", "right": "pull_request:44161", "accept": false, - "reason": "RegNet and LongT5 are separate model-specific changes." + "reason": "RegNet and LongT5 are unrelated code paths." }, { "left": "pull_request:44074", "right": "pull_request:44161", "accept": false, - "reason": "TextNet and LongT5 are different model refactors." + "reason": "TextNet and LongT5 are unrelated models." }, { - "left": "pull_request:43997", - "right": "pull_request:44071", + "left": "pull_request:44018", + "right": "pull_request:44161", "accept": false, - "reason": "RegNet and MPT are different model code paths." + "reason": "GPT-Neo and LongT5 are unrelated code paths." }, { - "left": "pull_request:44010", - "right": "pull_request:44018", + "left": "pull_request:43997", + "right": "pull_request:44071", "accept": false, - "reason": "SqueezeBert and GPT-Neo are unrelated model implementations." + "reason": "RegNet and MPT are unrelated models." }, { - "left": "pull_request:43996", - "right": "pull_request:44161", + "left": "pull_request:44010", + "right": "pull_request:44071", "accept": false, - "reason": "FNet/CVT refactor versus LongT5 refactor; different code paths." + "reason": "SqueezeBert and MPT are unrelated model implementations." }, { - "left": "pull_request:43996", - "right": "pull_request:44017", + "left": "pull_request:44010", + "right": "pull_request:44073", "accept": false, - "reason": "FNet/CVT and Segformer are separate model refactors." + "reason": "SqueezeBert and VisualBERT are different models." }, { - "left": "pull_request:44010", + "left": "pull_request:44018", "right": "pull_request:44044", "accept": false, - "reason": "SqueezeBert and DeBERTa are different model implementations." + "reason": "GPT-Neo and DeBERTa are unrelated code paths." }, { "left": "pull_request:44044", "right": "pull_request:44074", "accept": false, - "reason": "DeBERTa and TextNet are unrelated model-specific changes." + "reason": "DeBERTa and TextNet are unrelated models." }, { "left": "pull_request:44066", "right": "pull_request:44722", "accept": true, - "reason": "Same GPTJ/CodeGen output-tracing refactor family; both touch the same core GPT-J path and 44722 is the later broader revision." + "reason": "Same GPT-J/CodeGen output-tracing refactor on overlapping code paths; the later PR looks like a broader rework of the same change." }, { "left": "pull_request:44085", "right": "pull_request:44722", "accept": true, - "reason": "Same GPT-J output-tracing refactor on the same file; 44722 is a later compatible revision of that change." + "reason": "Both target the GPT-J output-tracing migration on the same underlying code path, with the later PR expanding the refactor." }, { "left": "pull_request:44085", "right": "pull_request:44129", "accept": false, - "reason": "GPT-J and SpeechT5 are unrelated model code paths." - }, - { - "left": "pull_request:43996", - "right": "pull_request:44007", - "accept": false, - "reason": "FNet/CVT and RegNet are different model families." - }, - { - "left": "pull_request:43996", - "right": "pull_request:44129", - "accept": false, - "reason": "FNet/CVT and SpeechT5 are unrelated refactors." + "reason": "GPT-J and SpeechT5 are unrelated models." }, { "left": "pull_request:44044", "right": "pull_request:44129", "accept": false, - "reason": "DeBERTa and SpeechT5 are different model-specific changes." + "reason": "DeBERTa and SpeechT5 are unrelated code paths." }, { "left": "pull_request:44044", "right": "pull_request:44154", "accept": false, - "reason": "DeBERTa and Vits are separate model code paths." + "reason": "DeBERTa and VITS are unrelated model implementations." } ] }, "evaluator_result": { - "accept": false, - "feedback": "The high-level summary is mostly grounded, but the soft-edge judgments are not conservative enough. In the packet, 44066\u219444085 and 44018\u219444068 are explicitly marked deterministic_accept=false, yet the analyst accepted them as the same underlying fix. That overstates the evidence. The packet\u2019s stronger GPT-J linkage is 44085\u219444722 (and 44066\u219444722), not 44066\u219444085. Because the accepted pairings are not sufficiently supported by the packet, I can\u2019t accept this output." + "accept": true, + "feedback": "The cluster summary is conservative and grounded in the packet. The accepted soft edges are limited to the clearly mergeable GPT-Neo and GPT-J/CodeGen follow-up chains, and the rejected edges avoid overgeneralizing across unrelated model families/code paths." }, "evaluator_used": true, - "retried": true, - "accepted_nontrivial_soft_edge": false, + "retried": false, + "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null }, @@ -19170,70 +19258,70 @@ "pull_request:43997|pull_request:44010", "pull_request:44010|pull_request:44129" ], - "prepared_review_unit_hash": "0dc12ea5b0e52d6958c3091e2cc13a67e16a77b8623e3b062456d37c4af66a5d", + "prepared_review_unit_hash": "247ede06c9acf6503c18e57190c12d34cbd51f224c2527df5a6b22a969a1bc22", "component_budget": { "node_count": 28, "item_count": 28, "soft_pair_count": 78, - "serialized_chars": 84897, - "estimated_input_tokens": 21225, - "estimated_eval_tokens": 42706 + "serialized_chars": 84901, + "estimated_input_tokens": 21226, + "estimated_eval_tokens": 42708 }, "budget": { "node_count": 8, "item_count": 8, "soft_pair_count": 6, - "serialized_chars": 17124, - "estimated_input_tokens": 4281, - "estimated_eval_tokens": 8818 + "serialized_chars": 17126, + "estimated_input_tokens": 4282, + "estimated_eval_tokens": 8820 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "These are model-specific PRs for migrating different architectures to standardized output tracing / capture_outputs. They share a theme but are not the same underlying change and should not be deduplicated as one PR cluster.", - "confidence": 0.92, + "summary": "These PRs share a broad theme of migrating model implementations to standardized output tracing/capture hooks, but they are model-specific refactors rather than duplicates of the same concrete code path. I would keep one representative PR for the cluster theme, but reject all soft pair merges.", + "confidence": 0.88, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 44129 is the strongest canonical representative: it has the clearest concrete refactor, an explicit issue target, and a substantial output-tracing change in a single model file.", + "canonical_pr_reason": "44073 is a good canonical PR for the cluster because it clearly represents the recurring standardized-output refactor pattern (adding can_return_tuple/capture_outputs and model-specific output hooks) without being dominated by an unrelated issue target.", "best_issue_reason": null, - "best_pr_reason": "PR 44129 is the best overall representative because it is the most concrete and traceable to an explicit issue, while the others are similar but model-specific refactors without a shared code path.", + "best_pr_reason": "44129 is the strongest standalone PR in the set because it has an explicit issue target and a concrete model-specific output-tracing refactor, making it the most actionable single change even though it is not a duplicate of the others.", "soft_edge_verdicts": [ { "left": "pull_request:44024", "right": "pull_request:44076", "accept": false, - "reason": "Both are output-tracing migrations, but for different models (FocalNet vs ImageGPT) and different code paths; they are not the same underlying bug/change." + "reason": "Both are output-tracing refactors, but one is FocalNet and the other is ImageGPT; they touch different model code paths and are not the same fix." }, { "left": "pull_request:44073", "right": "pull_request:44074", "accept": false, - "reason": "These touch VisualBert vs TextNet, so they are separate model-specific refactors rather than one mergeable fix." + "reason": "VisualBert vs TextNet are separate implementations; this is the same general migration style, not one mergeable duplicate PR." }, { "left": "pull_request:44056", "right": "pull_request:44073", "accept": false, - "reason": "MPNet and VisualBert are unrelated model implementations; the similarity is only the shared decorator/refactor pattern." + "reason": "MPNet and VisualBert both adopt capture_outputs, but they are distinct model-specific changes with different hooks and forward paths." }, { "left": "pull_request:44056", "right": "pull_request:44074", "accept": false, - "reason": "MPNet and TextNet are different model code paths, so this is not a duplicate change." + "reason": "These are parallel refactors in different models, not the same underlying bug or change." }, { "left": "pull_request:43997", "right": "pull_request:44010", "accept": false, - "reason": "RegNet and SqueezeBert are distinct architectures; they share a migration pattern but not the same concrete fix." + "reason": "RegNet and SqueezeBert are unrelated model code paths; similarity is only the shared output-tracing migration pattern." }, { "left": "pull_request:44010", "right": "pull_request:44129", "accept": false, - "reason": "SqueezeBert and SpeechT5 modify different implementations and output behaviors; the shared theme is too broad to treat as one PR." + "reason": "SqueezeBert and SpeechT5 are different model implementations, and the explicit issue target on SpeechT5 does not make these the same PR." } ] }, @@ -19319,292 +19407,334 @@ "pull_request:44002|pull_request:46106", "pull_request:43998|pull_request:46106" ], - "prepared_review_unit_hash": "116358f8eb1cacf7d841768427bcd6f40f1b6fcbd15356a3e5a0ad8b25442e71", + "prepared_review_unit_hash": "6181569103897973301d948050265797e80e0ed8571c1a212309bf20cf1128a2", "component_budget": { "node_count": 14, "item_count": 14, "soft_pair_count": 45, - "serialized_chars": 43782, - "estimated_input_tokens": 10946, - "estimated_eval_tokens": 22148 + "serialized_chars": 43785, + "estimated_input_tokens": 10947, + "estimated_eval_tokens": 22150 }, "budget": { "node_count": 14, "item_count": 14, "soft_pair_count": 45, - "serialized_chars": 43782, - "estimated_input_tokens": 10946, - "estimated_eval_tokens": 22148 + "serialized_chars": 43785, + "estimated_input_tokens": 10947, + "estimated_eval_tokens": 22150 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "This cluster is dominated by similar-but-separate model-specific output-tracing refactors, plus one unrelated docs toctree fix. They are not duplicate fixes for the same concrete code path, so the cluster should not be merged as a duplicate set.", - "confidence": 0.92, + "summary": "These PRs are mostly independent, model-specific output-tracing refactors that share a naming pattern but not the same concrete change. The only unrelated item is the docs toctree fix. I found no duplicate-worthy pairings here under a strict PR policy.", + "confidence": 0.9, "canonical_issue_reason": null, - "canonical_pr_reason": "PR #44025 is the best representative of the main cluster theme: a focused output-tracing refactor applied to closely related DepthAnything model files, with a clear, self-contained code change.", + "canonical_pr_reason": "#44003 is the broadest and most representative output-tracing refactor in the set, touching two related model implementations and the shared output-capturing plumbing.", "best_issue_reason": null, - "best_pr_reason": "PR #44025 is the strongest global representative because it is clearly scoped, model-specific, and closest to the shared output-tracing refactor pattern seen across the cluster.", + "best_pr_reason": "#44003 is the strongest cluster representative: it is a substantive code change, spans multiple related files, and best matches the cluster\u2019s core theme of output-tracing refactoring.", "soft_edge_verdicts": [ { "left": "pull_request:43998", "right": "pull_request:44002", "accept": false, - "reason": "Different models and code paths (timm_backbone vs upernet); same refactor theme is not enough for a duplicate PR merge." + "reason": "Different model families and files; same refactor theme but not the same concrete change." }, { "left": "pull_request:43998", "right": "pull_request:44000", "accept": false, - "reason": "Separate model implementations with no shared concrete bug or change." + "reason": "Different model families and files; same refactor theme but not the same concrete change." }, { "left": "pull_request:43998", "right": "pull_request:44001", "accept": false, - "reason": "Different model families; both touch output tracing but not the same fix." + "reason": "Different model families and files; same refactor theme but not the same concrete change." }, { "left": "pull_request:43998", "right": "pull_request:44003", "accept": false, - "reason": "Unrelated model code paths; no evidence they could be merged into one PR." + "reason": "Different model families and files; same refactor theme but not the same concrete change." + }, + { + "left": "pull_request:43998", + "right": "pull_request:44004", + "accept": false, + "reason": "Different code paths and files; both mention output tracing, but they are separate model changes." }, { "left": "pull_request:43998", "right": "pull_request:44029", "accept": false, - "reason": "Different architectures and files; only a broad refactor pattern is shared." + "reason": "Different model families and files; same refactor theme but not the same concrete change." + }, + { + "left": "pull_request:43998", + "right": "pull_request:46106", + "accept": false, + "reason": "Unrelated docs toctree change, not an output-tracing refactor." }, { "left": "pull_request:43999", "right": "pull_request:44001", "accept": false, - "reason": "MobileNetV1 vs UnivNet are unrelated model paths despite similar refactor wording." + "reason": "Different model families and files; same refactor theme but not the same concrete change." }, { "left": "pull_request:43999", "right": "pull_request:44002", "accept": false, - "reason": "Different model code paths; not the same underlying change." + "reason": "Different model families and files; same refactor theme but not the same concrete change." }, { "left": "pull_request:43999", "right": "pull_request:44025", "accept": false, - "reason": "MobileNetV1 and DepthAnything are separate changes, not one duplicate fix." + "reason": "Different model families and files; same refactor theme but not the same concrete change." }, { "left": "pull_request:43999", "right": "pull_request:44026", "accept": false, - "reason": "Different model-specific refactors; no shared concrete bug." + "reason": "Different model families and files; same refactor theme but not the same concrete change." }, { "left": "pull_request:43999", "right": "pull_request:44027", "accept": false, - "reason": "Separate model wrappers and different code paths." + "reason": "Different model families and files; same refactor theme but not the same concrete change." }, { "left": "pull_request:44000", "right": "pull_request:44001", "accept": false, - "reason": "Vision-text dual encoder vs UnivNet are unrelated model implementations." + "reason": "Different model families and files; same refactor theme but not the same concrete change." }, { "left": "pull_request:44000", "right": "pull_request:44002", "accept": false, - "reason": "Different model families; same subsystem theme only." + "reason": "Different model families and files; same refactor theme but not the same concrete change." }, { "left": "pull_request:44000", "right": "pull_request:44003", "accept": false, - "reason": "No shared concrete code path; these are model-specific refactors." + "reason": "Different model families and files; same refactor theme but not the same concrete change." + }, + { + "left": "pull_request:44000", + "right": "pull_request:44004", + "accept": false, + "reason": "Different code paths and files; both mention output tracing, but they are separate model changes." }, { "left": "pull_request:44000", "right": "pull_request:44025", "accept": false, - "reason": "Different models and different affected files; not mergeable as one PR." + "reason": "Different model families and files; same refactor theme but not the same concrete change." }, { "left": "pull_request:44001", "right": "pull_request:44002", "accept": false, - "reason": "UnivNet and UperNet are unrelated code paths." + "reason": "Different model families and files; same refactor theme but not the same concrete change." }, { "left": "pull_request:44001", "right": "pull_request:44003", "accept": false, - "reason": "Different model implementations; no same underlying bug/change." + "reason": "Different model families and files; same refactor theme but not the same concrete change." + }, + { + "left": "pull_request:44001", + "right": "pull_request:44004", + "accept": false, + "reason": "Different code paths and files; both mention output tracing, but they are separate model changes." }, { "left": "pull_request:44001", "right": "pull_request:44025", "accept": false, - "reason": "Separate model-specific output-tracing updates, not the same fix." + "reason": "Different model families and files; same refactor theme but not the same concrete change." }, { "left": "pull_request:44001", - "right": "pull_request:44004", + "right": "pull_request:46106", "accept": false, - "reason": "Unrelated models and code paths; similarity is only at a high level." + "reason": "Unrelated docs toctree change, not an output-tracing refactor." }, { "left": "pull_request:44002", "right": "pull_request:44003", "accept": false, - "reason": "UperNet and FalconMamba do not share the same concrete code path." + "reason": "Different model families and files; same refactor theme but not the same concrete change." + }, + { + "left": "pull_request:44002", + "right": "pull_request:44004", + "accept": false, + "reason": "Different code paths and files; both mention output tracing, but they are separate model changes." + }, + { + "left": "pull_request:44002", + "right": "pull_request:44029", + "accept": false, + "reason": "Different model families and files; same refactor theme but not the same concrete change." }, { "left": "pull_request:44002", + "right": "pull_request:46106", + "accept": false, + "reason": "Unrelated docs toctree change, not an output-tracing refactor." + }, + { + "left": "pull_request:44003", "right": "pull_request:44004", "accept": false, - "reason": "Different model code paths; not a duplicate PR." + "reason": "Different model families and files; same refactor theme but not the same concrete change." }, { "left": "pull_request:44003", "right": "pull_request:44028", "accept": false, - "reason": "FalconMamba and SuperPoint are distinct models with different fixes." + "reason": "Different model families and files; same refactor theme but not the same concrete change." }, { "left": "pull_request:44003", - "right": "pull_request:44004", + "right": "pull_request:44030", "accept": false, - "reason": "Separate model implementations; no evidence of the same underlying bug." + "reason": "Different model families and files; same refactor theme but not the same concrete change." }, { "left": "pull_request:43998", "right": "pull_request:44004", "accept": false, - "reason": "TimmBackbone and CodeGen are unrelated code paths; same refactor label is too broad." + "reason": "Different code paths and files; both mention output tracing, but they are separate model changes." + }, + { + "left": "pull_request:44001", + "right": "pull_request:44004", + "accept": false, + "reason": "Different code paths and files; both mention output tracing, but they are separate model changes." + }, + { + "left": "pull_request:44002", + "right": "pull_request:44004", + "accept": false, + "reason": "Different code paths and files; both mention output tracing, but they are separate model changes." + }, + { + "left": "pull_request:44000", + "right": "pull_request:44004", + "accept": false, + "reason": "Different code paths and files; both mention output tracing, but they are separate model changes." }, { "left": "pull_request:44025", "right": "pull_request:44026", "accept": false, - "reason": "DepthAnything vs VisionEncoderDecoder are different model wrappers and different concrete changes." + "reason": "Different model families and files; same refactor theme but not the same concrete change." }, { "left": "pull_request:44025", "right": "pull_request:44027", "accept": false, - "reason": "Different model code paths; not the same duplicate fix." + "reason": "Different model families and files; same refactor theme but not the same concrete change." }, { "left": "pull_request:44025", "right": "pull_request:44028", "accept": false, - "reason": "DepthAnything and SuperPoint are unrelated implementations." + "reason": "Different model families and files; same refactor theme but not the same concrete change." }, { "left": "pull_request:44025", "right": "pull_request:44029", "accept": false, - "reason": "DepthAnything and RWKV are separate model-specific changes." + "reason": "Different model families and files; same refactor theme but not the same concrete change." }, { "left": "pull_request:44025", "right": "pull_request:44030", "accept": false, - "reason": "DepthAnything and DPR do not fix the same concrete path." + "reason": "Different model families and files; same refactor theme but not the same concrete change." }, { "left": "pull_request:44026", "right": "pull_request:44027", "accept": false, - "reason": "VisionEncoderDecoder and SpeechEncoderDecoder are similar wrappers but still separate code paths and changes." + "reason": "Different model families and files; same refactor theme but not the same concrete change." }, { "left": "pull_request:44026", "right": "pull_request:44028", "accept": false, - "reason": "Different models; not one mergeable fix." + "reason": "Different model families and files; same refactor theme but not the same concrete change." }, { "left": "pull_request:44026", "right": "pull_request:44029", "accept": false, - "reason": "Unrelated model implementations with no shared concrete bug." + "reason": "Different model families and files; same refactor theme but not the same concrete change." }, { "left": "pull_request:44026", "right": "pull_request:44030", "accept": false, - "reason": "Different model families and affected code." + "reason": "Different model families and files; same refactor theme but not the same concrete change." }, { - "left": "pull_request:44027", - "right": "pull_request:44028", + "left": "pull_request:44026", + "right": "pull_request:46106", "accept": false, - "reason": "SpeechEncoderDecoder vs SuperPoint are unrelated code paths." + "reason": "Unrelated docs toctree change, not an output-tracing refactor." }, { "left": "pull_request:44027", - "right": "pull_request:44029", + "right": "pull_request:44028", "accept": false, - "reason": "Different model-specific refactors; not the same underlying change." + "reason": "Different model families and files; same refactor theme but not the same concrete change." }, { "left": "pull_request:44027", - "right": "pull_request:44030", - "accept": false, - "reason": "SpeechEncoderDecoder and DPR do not share the same concrete bug." - }, - { - "left": "pull_request:44028", "right": "pull_request:44029", "accept": false, - "reason": "SuperPoint and RWKV are unrelated model implementations." - }, - { - "left": "pull_request:44028", - "right": "pull_request:44030", - "accept": false, - "reason": "Different code paths; same broad theme is insufficient." + "reason": "Different model families and files; same refactor theme but not the same concrete change." }, { - "left": "pull_request:44029", + "left": "pull_request:44027", "right": "pull_request:44030", "accept": false, - "reason": "RWKV and DPR are distinct model-specific changes." + "reason": "Different model families and files; same refactor theme but not the same concrete change." }, { "left": "pull_request:44027", "right": "pull_request:46106", "accept": false, - "reason": "SpeechEncoderDecoder refactor and docs toctree cleanup are unrelated." - }, - { - "left": "pull_request:44026", - "right": "pull_request:46106", - "accept": false, - "reason": "VisionEncoderDecoder code change vs documentation tree fix are not duplicates." + "reason": "Unrelated docs toctree change, not an output-tracing refactor." }, { - "left": "pull_request:44001", - "right": "pull_request:46106", + "left": "pull_request:44028", + "right": "pull_request:44029", "accept": false, - "reason": "UnivNet model refactor and docs-only toctree change are unrelated." + "reason": "Different model families and files; same refactor theme but not the same concrete change." }, { - "left": "pull_request:44002", - "right": "pull_request:46106", + "left": "pull_request:44028", + "right": "pull_request:44030", "accept": false, - "reason": "UperNet code change and a docs navigation fix are different issues." + "reason": "Different model families and files; same refactor theme but not the same concrete change." }, { - "left": "pull_request:43998", - "right": "pull_request:46106", + "left": "pull_request:44029", + "right": "pull_request:44030", "accept": false, - "reason": "TimmBackbone model refactor and docs toctree edit are unrelated." + "reason": "Different model families and files; same refactor theme but not the same concrete change." } ] }, @@ -19637,12 +19767,12 @@ "pull_request:44662|pull_request:45075", "pull_request:45075|pull_request:45401" ], - "prepared_review_unit_hash": "10c1262a64e84657f1304e710b565faf244915dda21343144f665183ec104c79", + "prepared_review_unit_hash": "7625955c16f8b35bb35b75d769b140bf1fb46062d2064b99f7419714bcd66749", "component_budget": { "node_count": 3, "item_count": 3, "soft_pair_count": 3, - "serialized_chars": 10912, + "serialized_chars": 10911, "estimated_input_tokens": 2728, "estimated_eval_tokens": 5712 }, @@ -19650,7 +19780,7 @@ "node_count": 3, "item_count": 3, "soft_pair_count": 3, - "serialized_chars": 10912, + "serialized_chars": 10911, "estimated_input_tokens": 2728, "estimated_eval_tokens": 5712 }, @@ -19659,30 +19789,30 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "All three PRs are separate model-addition integrations (PenguinVL, DeepSeek-OCR-2, and Voxtral TTS). They share only generic Transformers plumbing and docs registration, not the same concrete code change, so none should be clustered as duplicates.", - "confidence": 0.97, + "summary": "These are three separate model-addition PRs for unrelated models; the overlap is mostly boilerplate docs, auto-registry wiring, and tests, not a shared bug or the same change.", + "confidence": 0.98, "canonical_issue_reason": null, - "canonical_pr_reason": "No single canonical PR: these PRs implement different models with different architecture/codepaths and do not subsume one another.", + "canonical_pr_reason": null, "best_issue_reason": null, - "best_pr_reason": "No global best PR for this cluster; 45075 is the most self-contained implementation, but it is still only one distinct model integration and not representative of the others.", + "best_pr_reason": "PR 45075 is the strongest representative if one must be picked: it is a substantial, non-draft model integration with high review activity and broad implementation coverage, but it is still not a duplicate of the others.", "soft_edge_verdicts": [ { "left": "pull_request:44662", "right": "pull_request:45401", "accept": false, - "reason": "Different model integrations with distinct model files and behavior; overlap is only generic auto-registration/docs, so they are not the same underlying change." + "reason": "Different model integrations (PenguinVL vs Voxtral TTS); shared files are only generic auto-registry/docs plumbing, not the same underlying code-path change." }, { "left": "pull_request:44662", "right": "pull_request:45075", "accept": false, - "reason": "Both are model-addition PRs, but for different models and codepaths. Shared files are boilerplate registration and docs, not a duplicate fix." + "reason": "Both add separate vision-language models, but they target different models and implementations; shared files are generic registration/docs changes, so this is not a duplicate PR." }, { "left": "pull_request:45075", "right": "pull_request:45401", "accept": false, - "reason": "DeepSeek-OCR-2 and Voxtral TTS are unrelated model implementations; shared Transformers registry files are too generic to treat as the same PR." + "reason": "DeepSeek-OCR-2 and Voxtral TTS are unrelated model additions with distinct architecture and code paths; overlap is limited to standard Transformers integration scaffolding." } ] }, @@ -19712,12 +19842,12 @@ "soft_pairs": [ "pull_request:44979|pull_request:45363" ], - "prepared_review_unit_hash": "6e95a85a72987f7271d6fa935f96c8c836ff610ba9203c8815e40bbbfc6afeb5", + "prepared_review_unit_hash": "596f4bbf30794e0eb641bf5f03fcdd76a6804a515b83d58d1708bb3a3ade686a", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4408, + "serialized_chars": 4407, "estimated_input_tokens": 1102, "estimated_eval_tokens": 2460 }, @@ -19725,7 +19855,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4408, + "serialized_chars": 4407, "estimated_input_tokens": 1102, "estimated_eval_tokens": 2460 }, @@ -19734,18 +19864,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These PRs are related to model/module fusion, but they are not duplicates: #44979 adds a general Module Fusion API, while #45363 implements a specific n-to-1 kernel fusion path via KernelConfig and related integration changes.", - "confidence": 0.89, + "summary": "These PRs are related to kernel/module fusion, but they target different abstractions and code paths: one introduces a generic Module Fusion API, while the other adds n-to-1 fusion through KernelConfig and integration hooks. They are not duplicates.", + "confidence": 0.86, "canonical_issue_reason": null, - "canonical_pr_reason": "#45363 is the more concrete end-user feature PR, with implementation across kernel-config, integrations, and a demo example; #44979 is a separate foundational API proposal.", + "canonical_pr_reason": "PR #44979 is the cleaner core API proposal: it adds a new `module_fusion.py` module plus tests, centered on the fusion abstraction itself. It is more canonical for the general feature area than the more specialized kernel-config integration PR.", "best_issue_reason": null, - "best_pr_reason": "#45363 best represents the cluster because it contains the more complete, user-facing fusion workflow rather than only the lower-level API scaffolding.", + "best_pr_reason": "PR #45363 is the most complete user-facing implementation in this cluster: it wires fusion into `modeling_utils`, `integrations`, and `KernelConfig`, and includes a runnable example. It is the strongest standalone PR, though it addresses a different mechanism than #44979.", "soft_edge_verdicts": [ { "left": "pull_request:44979", "right": "pull_request:45363", "accept": false, - "reason": "They share the broad fusion theme, but #44979 is a generic Module Fusion API and #45363 is a specific KernelConfig-based n-to-1 fusion implementation; they do not appear to fix the same concrete code-path problem." + "reason": "They both concern fusion, but #44979 is a generic module fusion API while #45363 implements n-to-1 kernel fusion via `KernelConfig` and integration plumbing. Different concrete mechanisms and merge targets." } ] }, @@ -19775,7 +19905,7 @@ "soft_pairs": [ "pull_request:45086|pull_request:45317" ], - "prepared_review_unit_hash": "573a3df150b8eae76d1b88a117b9195d8b376508903eeea87a71a15651c5fd95", + "prepared_review_unit_hash": "7f07bf70f36bbd57c88a72d1d42bacce5b6a391334cfcd97cf2c3b03a1ee0798", "component_budget": { "node_count": 2, "item_count": 2, @@ -19797,18 +19927,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Both pull requests implement the same Mistral regex patch in the same file, with 45317 adding accompanying test coverage. They are duplicate fixes for the same AttributeError and could be merged into one PR.", - "confidence": 0.98, + "summary": "Both PRs fix the same _patch_mistral_regex AttributeError by switching from backend_tokenizer.pre_tokenizer to tokenizer.pre_tokenizer; 45317 is the more complete later patch with tests.", + "confidence": 0.99, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 45317 is the stronger canonical PR because it contains the same code fix as 45086 plus test coverage, making it the more complete representation of the change.", + "canonical_pr_reason": "pull_request:45317 is the stronger canonical PR: it targets the same bug as 45086, adds tests, and appears to be the fuller implementation of the same fix.", "best_issue_reason": null, - "best_pr_reason": "PR 45317 is the best overall PR: same concrete fix, same target issue, and it includes tests in addition to the code change.", + "best_pr_reason": "pull_request:45317 is the best representative PR because it covers the same concrete code-path bug and includes test coverage, making it the most complete version of the fix.", "soft_edge_verdicts": [ { "left": "pull_request:45086", "right": "pull_request:45317", "accept": true, - "reason": "Same underlying AttributeError fix in the same code path and same file; 45317 is essentially 45086 plus tests, so they are mergeable as one change." + "reason": "Same underlying AttributeError fix in the same file and code path; the diff is effectively the same change, with 45317 adding tests." } ] }, @@ -19838,7 +19968,7 @@ "soft_pairs": [ "pull_request:45115|pull_request:45569" ], - "prepared_review_unit_hash": "daf2f2c2958ff8dcac52e9fa7614f220a05ebb1ab429e3bab87bcb600599a5ed", + "prepared_review_unit_hash": "aedcce9b4cec296758b56f74354b0392bf5007111dadac7f3e7377e232196dc6", "component_budget": { "node_count": 2, "item_count": 2, @@ -19860,18 +19990,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Both PRs are about NemotronH, but they target different changes: #45115 is a focused refactor/conversion mapping update toward GraniteMoeHybrid-compatible naming, while #45569 is a larger model-organization PR that splits NemotronH into dense/sparse variants and updates docs/tests. They overlap in shared NemotronH files, but they are not the same underlying change.", - "confidence": 0.89, + "summary": "Two PRs touch Nemotron H files, but they are different changes: one is a base-model refactor/conversion-mapping update, the other is a broader split into NemotronH dense/sparse models with docs and tests. They should not be treated as duplicates.", + "confidence": 0.82, "canonical_issue_reason": null, - "canonical_pr_reason": "PR #45569 is the better canonical representative because it covers the broader architectural rework and the new dense/sparse model split, whereas #45115 is a narrower refactor/conversion-specific change.", + "canonical_pr_reason": "PR 45569 is the better canonical PR because it is the broader, more complete implementation of the Nemotron H model split (dense/sparse wrappers, docs, auto mappings, tests), while PR 45115 is a narrower refactor focused on inheritance and conversion mapping.", "best_issue_reason": null, - "best_pr_reason": "PR #45569 is the best overall representative of the cluster\u2019s main theme since it introduces the larger NemotronH model reorganization and related API/docs updates; #45115 is too narrow to serve as the primary reference.", + "best_pr_reason": "PR 45569 is the strongest representative of the cluster: it contains the main user-facing model restructuring and supporting documentation/tests. PR 45115 is a separate, narrower code-path refactor and not a duplicate.", "soft_edge_verdicts": [ { "left": "pull_request:45115", "right": "pull_request:45569", "accept": false, - "reason": "Shared NemotronH files are not enough here: #45115 is about conversion/key renaming and inheritance refactoring, while #45569 adds separate dense/sparse model families plus docs/tests. They would not plausibly be merged as one PR." + "reason": "They overlap on Nemotron H base files, but the concrete changes differ: 45115 is a refactor/conversion-mapping and inheritance update, while 45569 adds dense/sparse model variants, docs, auto mappings, and tests. Not the same underlying fix and not a single PR candidate." } ] }, @@ -19901,7 +20031,7 @@ "soft_pairs": [ "pull_request:45332|pull_request:45333" ], - "prepared_review_unit_hash": "9dec5aced3e3d96db003ea77fe3ee4478ef0c724b291b6447b7fd4e35aeaf059", + "prepared_review_unit_hash": "71fc5528e0b8f779659676eed86bec33dfa58af2ac4ef81de8bf0c7448f04f9a", "component_budget": { "node_count": 2, "item_count": 2, @@ -19923,18 +20053,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Two related but distinct feature PRs: one adds heterogeneous per-layer configuration plumbing, the other adds broader heterogeneous model/runtime support. They overlap in config utilities and tests, but they are not the same concrete change.", - "confidence": 0.82, + "summary": "Two related but distinct feature PRs: one introduces heterogeneous per-layer config plumbing, the other adds broader heterogeneous model/runtime support on top. They share infrastructure files but are not the same concrete change, so they should not be merged as duplicates.", + "confidence": 0.92, "canonical_issue_reason": null, - "canonical_pr_reason": "PR #45332 is the broader umbrella change: it covers per-layer config handling plus downstream modeling/cache support, so it is the most representative PR in this pair.", + "canonical_pr_reason": "PR #45332 is the broader umbrella change and the better representative of the cluster: it extends the feature into cache, masking, and model code paths, whereas #45333 is limited to per-layer config plumbing.", "best_issue_reason": null, - "best_pr_reason": "PR #45332 is the best single representative because it subsumes the config work and extends it into the actual model code paths, making it the more complete change.", + "best_pr_reason": "PR #45332 is the strongest single PR to represent this cluster because it covers the larger end-to-end heterogeneous model support effort and contains the most complete implementation surface.", "soft_edge_verdicts": [ { "left": "pull_request:45332", "right": "pull_request:45333", "accept": false, - "reason": "They share heterogeneity/config infrastructure, but #45333 is specifically per-layer config support while #45332 is broader model/runtime support. Related sequence of changes, not the same underlying PR-sized fix." + "reason": "Related feature work, but not the same underlying change. #45333 adds per-layer config support; #45332 adds model/runtime support that builds on heterogeneous config. They could land in sequence, not as one duplicate PR." } ] }, @@ -19964,7 +20094,7 @@ "soft_pairs": [ "pull_request:45415|pull_request:45425" ], - "prepared_review_unit_hash": "e7c15a8ba6893da7be2f132c18b594e5c54f172a5bb63d88b0fc92e4a93cc072", + "prepared_review_unit_hash": "b22e43170fd6bbee7deecd104c7ab5a6d10424e633b1442eb78c052cb1776955", "component_budget": { "node_count": 2, "item_count": 2, @@ -19986,18 +20116,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are related to typing cleanup in Transformers, but they target different scopes and code paths: one is a broad type-checking sweep across many files, while the other adds specific typing protocols/helpers for PEFT/modeling utilities. Not a duplicate pair.", - "confidence": 0.9, + "summary": "Both PRs are typing-focused, but they are different changes: #45415 is a broad type-checking sweep across many modules plus pyproject config, while #45425 adds specific typing helpers/protocols for modeling_utils/PEFT-related code. The overlap is incidental, not the same underlying fix.", + "confidence": 0.91, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 45415 is the broader, more central typing sweep across the repository and is the better representative of this cluster than the narrower modeling_utils/PEFT typing work in 45425.", + "canonical_pr_reason": "PR #45425 is the more self-contained, focused typing change; PR #45415 is a much broader repo-wide type-checking sweep.", "best_issue_reason": null, - "best_pr_reason": "PR 45415 is the best single representative because it covers the wider type-checking initiative and touches many of the same areas only incidentally, whereas 45425 is a focused support patch.", + "best_pr_reason": "PR #45425 is the best representative of the cluster because it makes a concrete typing-infrastructure change in a narrower code path, whereas #45415 is a large mixed typing cleanup.", "soft_edge_verdicts": [ { "left": "pull_request:45415", "right": "pull_request:45425", "accept": false, - "reason": "Both are typing-related, but they fix different problems: 45415 is a broad type-checking pass across many modules, while 45425 adds specific typing protocols and aliases for modeling_utils/PEFT integration. They are not the same underlying change and would not plausibly merge into one PR." + "reason": "Shared files are incidental; #45415 is a broad type-checking sweep, while #45425 adds specific typing protocols/helpers for different code paths. They are not the same concrete change and would not plausibly merge into one PR." } ] }, @@ -20031,52 +20161,52 @@ "pull_request:46109|pull_request:46151", "pull_request:46135|pull_request:46151" ], - "prepared_review_unit_hash": "2e96f8d9f5856a738b128a63b625f5a5cc8386dbe0b8077dff3832cfb53ab426", + "prepared_review_unit_hash": "00963de295bad7b69ebeffc1381f45996e149fc4a321a2c49074ac12e1f8d83d", "component_budget": { "node_count": 4, "item_count": 4, "soft_pair_count": 3, - "serialized_chars": 9910, - "estimated_input_tokens": 2478, - "estimated_eval_tokens": 5212 + "serialized_chars": 9913, + "estimated_input_tokens": 2479, + "estimated_eval_tokens": 5214 }, "budget": { "node_count": 4, "item_count": 4, "soft_pair_count": 3, - "serialized_chars": 9910, - "estimated_input_tokens": 2478, - "estimated_eval_tokens": 5212 + "serialized_chars": 9913, + "estimated_input_tokens": 2479, + "estimated_eval_tokens": 5214 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are all test-focused PRs, but they target different subsystems and changes: model-parallel fixes, distributed trainer test normalization, Kosmos2.5 integration expectations, and token-classification pipeline expectations. They are not duplicates of the same underlying bug/change.", - "confidence": 0.89, + "summary": "These PRs are not duplicates of one another: they cover different code paths and different kinds of changes (AltCLIP/ChineseCLIP model-parallel bugfixes, distributed trainer test parity tweaks, Kosmos2.5 integration-test expectations, and pipeline token-classification expectations).", + "confidence": 0.96, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 46151 is the most self-contained and concrete change in the set: a narrow expectations update in one pipeline test file. The others are broader, or address different code paths entirely.", + "canonical_pr_reason": "No single canonical PR: the cluster mixes unrelated fixes, so there is no one artifact that subsumes the others.", "best_issue_reason": null, - "best_pr_reason": "46151 is the best representative if one PR must be chosen, because it is a small, specific test expectation change with a clear scope; however, it is not the same change as the other PRs.", + "best_pr_reason": "PR 45487 is the strongest standalone change because it fixes a concrete runtime bug in model code; the others are narrower test-expected-output adjustments in unrelated areas.", "soft_edge_verdicts": [ { "left": "pull_request:45487", "right": "pull_request:46151", "accept": false, - "reason": "45487 is a model-parallel bug fix for AltCLIP/ChineseCLIP model code, while 46151 only adds pipeline token-classification test expectations. Different underlying behavior and code paths." + "reason": "Different features and code paths: AltCLIP/ChineseCLIP model-parallel device handling vs token-classification pipeline test expectations." }, { "left": "pull_request:46109", "right": "pull_request:46151", "accept": false, - "reason": "46109 changes distributed trainer test assertions for device naming; 46151 changes token-classification pipeline expectations. They share a testing theme but not the same bug or change." + "reason": "Both are test-only, but they target different subsystems and issues: distributed trainer device-string parity vs pipeline token-classification expectations." }, { "left": "pull_request:46135", "right": "pull_request:46151", "accept": false, - "reason": "46135 adds XPU expectations to Kosmos2.5 integration tests, while 46151 adjusts token-classification pipeline expectations. Different models, tests, and underlying issues." + "reason": "Kosmos2.5 integration-test expectations are unrelated to token-classification pipeline expectations; not the same underlying change." } ] }, @@ -20108,12 +20238,12 @@ "pull_request:45513|pull_request:46084", "pull_request:46084|pull_request:46098" ], - "prepared_review_unit_hash": "c1f95559d29c0dcf17da07e9ea7d06d6b5367299d4d35c0dbca2eccaa723fc8e", + "prepared_review_unit_hash": "5a4046297a4240a7284b1773aaf360039955d995c9829af3ec0b8cc726524609", "component_budget": { "node_count": 3, "item_count": 3, "soft_pair_count": 2, - "serialized_chars": 7427, + "serialized_chars": 7426, "estimated_input_tokens": 1857, "estimated_eval_tokens": 3970 }, @@ -20121,7 +20251,7 @@ "node_count": 3, "item_count": 3, "soft_pair_count": 2, - "serialized_chars": 7427, + "serialized_chars": 7426, "estimated_input_tokens": 1857, "estimated_eval_tokens": 3970 }, @@ -20130,30 +20260,30 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Two PRs are about the same Mamba2 cache bug for `seq_len > 1`; the Qwen3.5 PR is unrelated and should stay separate.", - "confidence": 0.84, + "summary": "Two PRs are about the same Mamba2 `use_cache`/`seq_len > 1` bug; the Qwen3.5 PR is a different model and code path. PR #46098 is the best representative because it is the most complete Mamba2 fix and explicitly targets the tracked issue.", + "confidence": 0.92, "canonical_issue_reason": null, - "canonical_pr_reason": "PR #46098 is the clearest representative of the shared Mamba2 fix: it explicitly targets issue #46032, describes the incorrect `use_cache` behavior for `seq_len > 1`, and includes the broader code/test/doc coverage.", + "canonical_pr_reason": "PR #46098 is the cleanest representative of the duplicated Mamba2 bug fix: it explicitly targets issue #46032, covers the seq_len>1 cached path comprehensively, and includes tests/docs. PR #46084 is a narrower variant of the same fix, while #45513 is unrelated.", "best_issue_reason": null, - "best_pr_reason": "PR #46098 is the best overall PR candidate because it is the most explicit and complete version of the Mamba2 fix, with the issue link, broader CPU/GPU framing, and added documentation/tests.", + "best_pr_reason": "PR #46098 is the best global PR candidate because it directly addresses the tracked Mamba2 cache bug with the broadest fix surface and validation, making it the most complete canonical change among the set.", "soft_edge_verdicts": [ { "left": "pull_request:45513", "right": "pull_request:46084", "accept": false, - "reason": "Different models and different bugs: Qwen3.5 GDN linear attention cached forward vs Mamba2 `use_cache` handling. Same general theme, but not the same concrete change." + "reason": "Different models and bugs: Qwen3.5 linear-attention cached forward vs. Mamba2 `use_cache` handling. Similar symptom, but not the same concrete code-path change and not mergeable into one PR." }, { "left": "pull_request:46084", "right": "pull_request:46098", "accept": true, - "reason": "Both target the same Mamba2 `use_cache` bug for `seq_len > 1` on the same code path and issue #46032, so they look like duplicate fixes for the same underlying problem." + "reason": "Same underlying Mamba2 `use_cache` with `seq_len > 1` bug on the same code path, with #46098 essentially extending/subsuming the narrower fix in #46084." } ] }, "evaluator_result": { "accept": true, - "feedback": "The summary is grounded in the packet: PR #46084 and #46098 both target issue #46032 and the same Mamba2 `use_cache`/`seq_len > 1` bug, while #45513 is a different Qwen3.5 change. The soft-edge decisions are conservative enough, and the PR pair looks mergeable as one bugfix family." + "feedback": "The summary is grounded in the packet: 45513 is clearly a different Qwen3.5-related PR, while 46084 and 46098 share the same Mamba2 issue target and files. The soft-edge verdicts are conservative, and treating 46084 and 46098 as mergeable for the same bugfix is supported by the evidence." }, "evaluator_used": true, "retried": false, @@ -20180,7 +20310,7 @@ "soft_pairs": [ "pull_request:45548|pull_request:45662" ], - "prepared_review_unit_hash": "0ffec0bf1ea23f4e8b2fa6d40330cbda3bc8abcdd5714340007067484ef5c7ad", + "prepared_review_unit_hash": "9555e949d36abd8cbf394499ae8b8d9b20a049de2a1d36b6be7fe651e85be000", "component_budget": { "node_count": 2, "item_count": 2, @@ -20202,18 +20332,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Both PRs are about expert parallelism, but they fix different bugs on different loading paths. They are related, not duplicates.", - "confidence": 0.94, + "summary": "Two open PRs touch expert-parallel loading, but they fix different bugs: one is about preserving distributed_config / EP state when loading via accelerate+ZeRO-3, the other is about preventing expert weights from being overwritten by rank-0 broadcast under FSDP2.", + "confidence": 0.91, "canonical_issue_reason": null, - "canonical_pr_reason": "No issue links are present. If a single representative PR is needed, #45662 is the stronger canonical pick because it targets a specific expert-parallel corruption path and makes the broader code-path fix.", + "canonical_pr_reason": "PR 45662 is the better cluster representative because it addresses a concrete EP/FSDP2 overwrite path with broader code impact; PR 45548 is a separate, narrower accelerate/ZeRO-3 configuration-loading fix.", "best_issue_reason": null, - "best_pr_reason": "#45662 is the better standalone PR: it addresses a concrete rank-0 broadcast / post-shard wrapping bug with a clear failure mode. #45548 is narrower and mainly about preserving distributed_config and a helper property for loading/ZeRO-3.", + "best_pr_reason": "PR 45662 is the strongest standalone PR in this set: it targets the actual expert-parameter overwrite failure mode across multiple loading modules, while 45548 is a smaller serialization/loading adjustment.", "soft_edge_verdicts": [ { "left": "pull_request:45548", "right": "pull_request:45662", "accept": false, - "reason": "Both involve EP/loading, but #45548 fixes distributed_config serialization and has_ep for accelerate/ZeRO-3 loading, while #45662 fixes experts being overwritten by rank-0 broadcast via post-shard wrapping/FSDP2. Different underlying bugs and merge scopes." + "reason": "Both are about expert parallelism, but the underlying bugs differ: 45548 fixes EP state/config handling during accelerate+ZeRO-3 loading, while 45662 fixes experts being overwritten by rank-0 broadcast in FSDP2. They would not be merged into one PR." } ] }, @@ -20243,7 +20373,7 @@ "soft_pairs": [ "pull_request:45639|pull_request:45645" ], - "prepared_review_unit_hash": "a95dd27513b9136da6f4925b98a659420f56df680981ac75417c799be5b68e75", + "prepared_review_unit_hash": "cc6329d876355e0560bfba5d39a6ed1d4b8816e5eaac13c5a613de81142a9761", "component_budget": { "node_count": 2, "item_count": 2, @@ -20265,18 +20395,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Both PRs address the same xdist-safe captured_info logging problem; PR 45645 is the broader, more complete fix and subsumes 45639's core change.", - "confidence": 0.9, + "summary": "Both pull requests address the same xdist-related captured_info artifact collision problem; PR 45645 is the broader, more complete version that also updates CI log collection. PR 45639 looks like a narrower subset of the same fix.", + "confidence": 0.88, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 45645 is the better canonical representative because it includes the core testing_utils fix plus the CI workflow/log-handling updates needed to preserve and aggregate the xdist-generated artifacts.", + "canonical_pr_reason": "PR 45645 is the better canonical candidate because it includes the core xdist-safe output-file change plus the CI workflow adjustment to collect multiple captured_info files, making it the more complete fix.", "best_issue_reason": null, - "best_pr_reason": "PR 45645 best captures the underlying change set; PR 45639 is a narrower subset focused only on the testing_utils side.", + "best_pr_reason": "PR 45645 is the best overall PR because it addresses the underlying collision bug and also preserves debug-log visibility in CI, whereas PR 45639 only covers the output-file side of the change.", "soft_edge_verdicts": [ { "left": "pull_request:45639", "right": "pull_request:45645", "accept": true, - "reason": "Accept: both target the same concrete bug in captured_info output under xdist, and 45639 is essentially a subset of 45645's broader fix." + "reason": "Both PRs target the same concrete bug: xdist collisions for captured_info artifacts in testing_utils. PR 45645 appears to supersede the narrower implementation in 45639 with the same core fix plus CI handling." } ] }, @@ -20364,63 +20494,85 @@ "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-46069-3", + "cluster_id": "cluster-46069-4", "nodes": [ "pull_request:46069", "pull_request:46073", - "pull_request:46154" + "pull_request:46154", + "pull_request:46196" ], "soft_pairs": [ "pull_request:46069|pull_request:46073", "pull_request:46069|pull_request:46154", - "pull_request:46073|pull_request:46154" + "pull_request:46069|pull_request:46196", + "pull_request:46073|pull_request:46154", + "pull_request:46073|pull_request:46196", + "pull_request:46154|pull_request:46196" ], - "prepared_review_unit_hash": "4d961c108bea1ad061e9003cadc6e5e74eb0c291da0432d0a99d20354b180f14", + "prepared_review_unit_hash": "c749d9042e9101fc8ed99af5adbaebb8183df06ea9e3584be5091188b08b45b5", "component_budget": { - "node_count": 3, - "item_count": 3, - "soft_pair_count": 3, - "serialized_chars": 7264, - "estimated_input_tokens": 1816, - "estimated_eval_tokens": 3888 + "node_count": 4, + "item_count": 4, + "soft_pair_count": 6, + "serialized_chars": 10494, + "estimated_input_tokens": 2624, + "estimated_eval_tokens": 5504 }, "budget": { - "node_count": 3, - "item_count": 3, - "soft_pair_count": 3, - "serialized_chars": 7264, - "estimated_input_tokens": 1816, - "estimated_eval_tokens": 3888 + "node_count": 4, + "item_count": 4, + "soft_pair_count": 6, + "serialized_chars": 10494, + "estimated_input_tokens": 2624, + "estimated_eval_tokens": 5504 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are three model-specific attention-dispatch PRs. Two are broad SDPA-support changes for different architectures (CTRL and T5), and the third is a ShieldGemma2-specific attn_implementation dispatch PR. They are similar in theme but not duplicates of the same bug/change.", - "confidence": 0.94, + "summary": "These PRs all add SDPA/attention-dispatch support, but each targets a different model family and code path, so they are related feature ports rather than duplicates.", + "confidence": 0.98, "canonical_issue_reason": null, - "canonical_pr_reason": "No single canonical PR: each PR targets a different model implementation and code path, so they are not interchangeable duplicates.", + "canonical_pr_reason": "No single canonical PR: each change is model-specific (ShieldGemma2, CTRL, T5, SeamlessM4T/v2) and would not be merged into one concrete fix.", "best_issue_reason": null, - "best_pr_reason": "pull_request:46154 is the strongest representative because it has the clearest, most complete SDPA dispatch change and the broadest accompanying test updates, but it is still model-specific rather than a duplicate of the others.", + "best_pr_reason": "PR 46196 is the broadest representative because it covers both SeamlessM4T and SeamlessM4T v2, but it is still not a duplicate of the other model-specific dispatch PRs.", "soft_edge_verdicts": [ { "left": "pull_request:46069", "right": "pull_request:46073", "accept": false, - "reason": "Different model classes and different attention dispatch code paths (ShieldGemma2 vs CTRL). They share a feature theme, but not the same underlying fix." + "reason": "Same general SDPA theme, but ShieldGemma2 and CTRL are different model implementations with separate dispatch paths." }, { "left": "pull_request:46069", "right": "pull_request:46154", "accept": false, - "reason": "ShieldGemma2 attn_implementation dispatch is unrelated to T5 SDPA dispatch beyond the generic attention-feature theme; not the same concrete change." + "reason": "Both add attention-dispatch support, but ShieldGemma2 and T5 are unrelated code paths and not one concrete bug fix." + }, + { + "left": "pull_request:46069", + "right": "pull_request:46196", + "accept": false, + "reason": "ShieldGemma2 image-classification attention dispatch is distinct from SeamlessM4T/v2 seq2seq attention dispatch." }, { "left": "pull_request:46073", "right": "pull_request:46154", "accept": false, - "reason": "Both add SDPA dispatch support, but for different architectures with separate implementations and tests. They are similar feature work, not a single mergeable PR." + "reason": "CTRL and T5 are separate model families; this is the same feature class, not the same underlying PR change." + }, + { + "left": "pull_request:46073", + "right": "pull_request:46196", + "accept": false, + "reason": "CTRL SDPA dispatch and SeamlessM4T/v2 SDPA dispatch touch different architectures and cannot plausibly be one merged fix." + }, + { + "left": "pull_request:46154", + "right": "pull_request:46196", + "accept": false, + "reason": "T5 and SeamlessM4T/v2 both add SDPA support, but they are independent model-specific implementations, not duplicates." } ] }, diff --git a/analysis/current/manifest.json b/analysis/current/manifest.json index af421a01ae01d1163aaef5640f18ce661ebde644..0137c8cd77e1b39a0fa66fc4af898987a072cb81 100644 --- a/analysis/current/manifest.json +++ b/analysis/current/manifest.json @@ -1,8 +1,8 @@ { - "analysis_id": "hybrid-model-20260525t060031z", + "analysis_id": "hybrid-model-20260525t120039z", "archived_artifacts": { - "hybrid": "snapshots/20260525T060031Z/analysis-runs/hybrid-model-20260525t060031z/analysis-report-hybrid.json", - "hybrid_reviews": "snapshots/20260525T060031Z/analysis-runs/hybrid-model-20260525t060031z/analysis-report-hybrid.llm-reviews.json" + "hybrid": "snapshots/20260525T120039Z/analysis-runs/hybrid-model-20260525t120039z/analysis-report-hybrid.json", + "hybrid_reviews": "snapshots/20260525T120039Z/analysis-runs/hybrid-model-20260525t120039z/analysis-report-hybrid.llm-reviews.json" }, "artifacts": { "hybrid": "analysis/current/analysis-report-hybrid.json", @@ -10,9 +10,9 @@ }, "channel": "canonical", "model": null, - "published_at": "2026-05-25T06:10:07Z", + "published_at": "2026-05-25T12:15:45Z", "repo": "huggingface/transformers", "schema_version": 1, - "snapshot_id": "20260525T060031Z", + "snapshot_id": "20260525T120039Z", "variant": "hybrid" } diff --git a/snapshots/20260525T120039Z/analysis-runs/hybrid-model-20260525t120039z/analysis-report-hybrid.json b/snapshots/20260525T120039Z/analysis-runs/hybrid-model-20260525t120039z/analysis-report-hybrid.json new file mode 100644 index 0000000000000000000000000000000000000000..b16ba19c0541933795ddabfbeb0df3141d3cbd10 --- /dev/null +++ b/snapshots/20260525T120039Z/analysis-runs/hybrid-model-20260525t120039z/analysis-report-hybrid.json @@ -0,0 +1,2156 @@ +{ + "schema_version": "1.0", + "repo": "huggingface/transformers", + "snapshot_id": "20260525T120039Z", + "generated_at": "2026-05-25T12:15:43Z", + "evidence_quality": "full", + "llm_enrichment": true, + "meta_bugs": [ + { + "cluster_id": "cluster-43979-11", + "summary": "Cluster of 1 issues and 10 PRs centered on issue #43979.", + "status": "open", + "confidence": 0.8, + "canonical_issue_number": 43979, + "canonical_pr_number": 44007, + "issue_numbers": [ + 43979 + ], + "pr_numbers": [ + 43996, + 44007, + 44013, + 44044, + 44066, + 44072, + 44085, + 44129, + 44154, + 44722 + ], + "evidence_types": [ + "closing_reference", + "shared_issue_target", + "soft_similarity" + ], + "pr_comparisons": [ + { + "left_pr_number": 43996, + "right_pr_number": 44007, + "code_similarity": 0.179, + "size_similarity": 0.576, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.429, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 43996, + "right_pr_number": 44013, + "code_similarity": 0.122, + "size_similarity": 0.318, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.392, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 43996, + "right_pr_number": 44044, + "code_similarity": 0.245, + "size_similarity": 0.864, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.479, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 43996, + "right_pr_number": 44066, + "code_similarity": 0.225, + "size_similarity": 0.818, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.408, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 43996, + "right_pr_number": 44072, + "code_similarity": 0.14, + "size_similarity": 0.303, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.528, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 43996, + "right_pr_number": 44085, + "code_similarity": 0.216, + "size_similarity": 0.783, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.398, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 43996, + "right_pr_number": 44129, + "code_similarity": 0.163, + "size_similarity": 0.643, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.229, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 43996, + "right_pr_number": 44154, + "code_similarity": 0.153, + "size_similarity": 0.535, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.31, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 43996, + "right_pr_number": 44722, + "code_similarity": 0.225, + "size_similarity": 0.848, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.368, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 44007, + "right_pr_number": 44013, + "code_similarity": 0.19, + "size_similarity": 0.553, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.531, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 44007, + "right_pr_number": 44044, + "code_similarity": 0.186, + "size_similarity": 0.667, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.354, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 44007, + "right_pr_number": 44066, + "code_similarity": 0.188, + "size_similarity": 0.704, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.315, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 44007, + "right_pr_number": 44072, + "code_similarity": 0.212, + "size_similarity": 0.526, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.708, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 44007, + "right_pr_number": 44085, + "code_similarity": 0.195, + "size_similarity": 0.735, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.318, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 44007, + "right_pr_number": 44129, + "code_similarity": 0.103, + "size_similarity": 0.37, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.191, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 44007, + "right_pr_number": 44154, + "code_similarity": 0.238, + "size_similarity": 0.93, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.344, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 44007, + "right_pr_number": 44722, + "code_similarity": 0.178, + "size_similarity": 0.679, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.28, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 44013, + "right_pr_number": 44044, + "code_similarity": 0.126, + "size_similarity": 0.368, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.351, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 44013, + "right_pr_number": 44066, + "code_similarity": 0.127, + "size_similarity": 0.389, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.325, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 44013, + "right_pr_number": 44072, + "code_similarity": 0.29, + "size_similarity": 0.952, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.667, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 44013, + "right_pr_number": 44085, + "code_similarity": 0.131, + "size_similarity": 0.406, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.329, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 44013, + "right_pr_number": 44129, + "code_similarity": 0.07, + "size_similarity": 0.205, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.192, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 44013, + "right_pr_number": 44154, + "code_similarity": 0.177, + "size_similarity": 0.594, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.389, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 44013, + "right_pr_number": 44722, + "code_similarity": 0.118, + "size_similarity": 0.375, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.287, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 44044, + "right_pr_number": 44066, + "code_similarity": 0.25, + "size_similarity": 0.947, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.404, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 44044, + "right_pr_number": 44072, + "code_similarity": 0.136, + "size_similarity": 0.351, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.442, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 44044, + "right_pr_number": 44085, + "code_similarity": 0.24, + "size_similarity": 0.906, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.394, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 44044, + "right_pr_number": 44129, + "code_similarity": 0.147, + "size_similarity": 0.555, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.243, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 44044, + "right_pr_number": 44154, + "code_similarity": 0.17, + "size_similarity": 0.62, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.306, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 44044, + "right_pr_number": 44722, + "code_similarity": 0.257, + "size_similarity": 0.982, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.402, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 44066, + "right_pr_number": 44072, + "code_similarity": 0.133, + "size_similarity": 0.37, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.393, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 44066, + "right_pr_number": 44085, + "code_similarity": 0.763, + "size_similarity": 0.957, + "file_overlap": 0.5, + "area_overlap": 0.825, + "patch_similarity": 0.887, + "shared_filenames": [ + "src/transformers/models/gptj/modeling_gptj.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/models/gptj/modeling_gptj.py", + "left_ranges": [ + [ + 33, + 41 + ], + [ + 174, + 181 + ], + [ + 250, + 257 + ], + [ + 398, + 419 + ], + [ + 425, + 434 + ], + [ + 459, + 466 + ], + [ + 471, + 488 + ], + [ + 518, + 540 + ], + [ + 554, + 560 + ], + [ + 566, + 575 + ], + [ + 580, + 598 + ], + [ + 601, + 606 + ], + [ + 634, + 640 + ], + [ + 646, + 653 + ], + [ + 658, + 674 + ], + [ + 717, + 722 + ], + [ + 738, + 744 + ], + [ + 749, + 772 + ], + [ + 790, + 795 + ] + ], + "right_ranges": [ + [ + 33, + 40 + ], + [ + 173, + 180 + ], + [ + 249, + 256 + ], + [ + 397, + 405 + ], + [ + 408, + 419 + ], + [ + 425, + 434 + ], + [ + 459, + 465 + ], + [ + 470, + 489 + ], + [ + 517, + 539 + ], + [ + 553, + 559 + ], + [ + 565, + 574 + ], + [ + 579, + 597 + ], + [ + 600, + 611 + ], + [ + 633, + 639 + ], + [ + 645, + 652 + ], + [ + 657, + 673 + ], + [ + 716, + 728 + ], + [ + 737, + 743 + ], + [ + 748, + 771 + ], + [ + 789, + 794 + ] + ] + } + ] + }, + { + "left_pr_number": 44066, + "right_pr_number": 44129, + "code_similarity": 0.145, + "size_similarity": 0.526, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.263, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 44066, + "right_pr_number": 44154, + "code_similarity": 0.174, + "size_similarity": 0.654, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.286, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 44066, + "right_pr_number": 44722, + "code_similarity": 0.907, + "size_similarity": 0.964, + "file_overlap": 1.0, + "area_overlap": 0.808, + "patch_similarity": 0.874, + "shared_filenames": [ + "src/transformers/models/codegen/modeling_codegen.py", + "src/transformers/models/gptj/modeling_gptj.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/models/codegen/modeling_codegen.py", + "left_ranges": [ + [ + 245, + 266 + ] + ], + "right_ranges": [ + [ + 228, + 234 + ] + ] + }, + { + "filename": "src/transformers/models/gptj/modeling_gptj.py", + "left_ranges": [ + [ + 33, + 41 + ], + [ + 174, + 181 + ], + [ + 250, + 257 + ], + [ + 398, + 419 + ], + [ + 425, + 434 + ], + [ + 459, + 466 + ], + [ + 471, + 488 + ], + [ + 518, + 540 + ], + [ + 554, + 560 + ], + [ + 566, + 575 + ], + [ + 580, + 598 + ], + [ + 601, + 606 + ], + [ + 634, + 640 + ], + [ + 646, + 653 + ], + [ + 658, + 674 + ], + [ + 717, + 722 + ], + [ + 738, + 744 + ], + [ + 749, + 772 + ], + [ + 790, + 795 + ] + ], + "right_ranges": [ + [ + 33, + 42 + ], + [ + 175, + 182 + ], + [ + 251, + 258 + ], + [ + 399, + 420 + ], + [ + 426, + 435 + ], + [ + 460, + 467 + ], + [ + 472, + 489 + ], + [ + 519, + 540 + ], + [ + 554, + 560 + ], + [ + 566, + 575 + ], + [ + 580, + 598 + ], + [ + 601, + 612 + ], + [ + 634, + 640 + ], + [ + 646, + 653 + ], + [ + 658, + 663 + ], + [ + 666, + 674 + ], + [ + 717, + 722 + ], + [ + 738, + 744 + ], + [ + 749, + 772 + ], + [ + 790, + 795 + ] + ] + } + ] + }, + { + "left_pr_number": 44072, + "right_pr_number": 44085, + "code_similarity": 0.137, + "size_similarity": 0.387, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.398, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 44072, + "right_pr_number": 44129, + "code_similarity": 0.074, + "size_similarity": 0.195, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.231, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 44072, + "right_pr_number": 44154, + "code_similarity": 0.175, + "size_similarity": 0.566, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.414, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 44072, + "right_pr_number": 44722, + "code_similarity": 0.124, + "size_similarity": 0.357, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.347, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 44085, + "right_pr_number": 44129, + "code_similarity": 0.141, + "size_similarity": 0.503, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.272, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 44085, + "right_pr_number": 44154, + "code_similarity": 0.18, + "size_similarity": 0.684, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.289, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 44085, + "right_pr_number": 44722, + "code_similarity": 0.728, + "size_similarity": 0.923, + "file_overlap": 0.5, + "area_overlap": 0.791, + "patch_similarity": 0.78, + "shared_filenames": [ + "src/transformers/models/gptj/modeling_gptj.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/models/gptj/modeling_gptj.py", + "left_ranges": [ + [ + 33, + 40 + ], + [ + 173, + 180 + ], + [ + 249, + 256 + ], + [ + 397, + 405 + ], + [ + 408, + 419 + ], + [ + 425, + 434 + ], + [ + 459, + 465 + ], + [ + 470, + 489 + ], + [ + 517, + 539 + ], + [ + 553, + 559 + ], + [ + 565, + 574 + ], + [ + 579, + 597 + ], + [ + 600, + 611 + ], + [ + 633, + 639 + ], + [ + 645, + 652 + ], + [ + 657, + 673 + ], + [ + 716, + 728 + ], + [ + 737, + 743 + ], + [ + 748, + 771 + ], + [ + 789, + 794 + ] + ], + "right_ranges": [ + [ + 33, + 42 + ], + [ + 175, + 182 + ], + [ + 251, + 258 + ], + [ + 399, + 420 + ], + [ + 426, + 435 + ], + [ + 460, + 467 + ], + [ + 472, + 489 + ], + [ + 519, + 540 + ], + [ + 554, + 560 + ], + [ + 566, + 575 + ], + [ + 580, + 598 + ], + [ + 601, + 612 + ], + [ + 634, + 640 + ], + [ + 646, + 653 + ], + [ + 658, + 663 + ], + [ + 666, + 674 + ], + [ + 717, + 722 + ], + [ + 738, + 744 + ], + [ + 749, + 772 + ], + [ + 790, + 795 + ] + ] + } + ] + }, + { + "left_pr_number": 44129, + "right_pr_number": 44154, + "code_similarity": 0.099, + "size_similarity": 0.344, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.199, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 44129, + "right_pr_number": 44722, + "code_similarity": 0.146, + "size_similarity": 0.545, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.247, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 44154, + "right_pr_number": 44722, + "code_similarity": 0.164, + "size_similarity": 0.631, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.255, + "shared_filenames": [], + "shared_file_areas": [] + } + ] + }, + { + "cluster_id": "cluster-41211-3", + "summary": "Cluster of 1 issues and 2 PRs centered on issue #41211.", + "status": "open", + "confidence": 0.55, + "canonical_issue_number": 41211, + "canonical_pr_number": 44339, + "issue_numbers": [ + 41211 + ], + "pr_numbers": [ + 41356, + 44339 + ], + "evidence_types": [ + "closing_reference" + ], + "pr_comparisons": [ + { + "left_pr_number": 41356, + "right_pr_number": 44339, + "code_similarity": 0.155, + "size_similarity": 0.094, + "file_overlap": 0.273, + "area_overlap": 0.078, + "patch_similarity": 0.183, + "shared_filenames": [ + "docs/source/en/model_doc/deimv2.md", + "src/transformers/models/deimv2/__init__.py", + "src/transformers/models/deimv2/configuration_deimv2.py", + "src/transformers/models/deimv2/modeling_deimv2.py", + "tests/models/deimv2/__init__.py", + "tests/models/deimv2/test_modeling_deimv2.py" + ], + "shared_file_areas": [ + { + "filename": "docs/source/en/model_doc/deimv2.md", + "left_ranges": [ + [ + 1, + 132 + ] + ], + "right_ranges": [ + [ + 1, + 65 + ] + ] + }, + { + "filename": "src/transformers/models/deimv2/__init__.py", + "left_ranges": [ + [ + 1, + 15 + ] + ], + "right_ranges": [ + [ + 1, + 29 + ] + ] + }, + { + "filename": "src/transformers/models/deimv2/configuration_deimv2.py", + "left_ranges": [ + [ + 1, + 74 + ] + ], + "right_ranges": [ + [ + 1, + 266 + ] + ] + }, + { + "filename": "tests/models/deimv2/test_modeling_deimv2.py", + "left_ranges": [ + [ + 1, + 15 + ] + ], + "right_ranges": [ + [ + 1, + 1734 + ] + ] + } + ] + } + ] + }, + { + "cluster_id": "cluster-43240-3", + "summary": "Cluster of 1 issues and 2 PRs centered on issue #43240.", + "status": "open", + "confidence": 0.75, + "canonical_issue_number": 43240, + "canonical_pr_number": 43251, + "issue_numbers": [ + 43240 + ], + "pr_numbers": [ + 43251, + 43254 + ], + "evidence_types": [ + "closing_reference", + "shared_issue_target" + ], + "pr_comparisons": [ + { + "left_pr_number": 43251, + "right_pr_number": 43254, + "code_similarity": 0.794, + "size_similarity": 0.64, + "file_overlap": 1.0, + "area_overlap": 0.667, + "patch_similarity": 0.882, + "shared_filenames": [ + "src/transformers/loss/loss_utils.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/loss/loss_utils.py", + "left_ranges": [ + [ + 30, + 50 + ] + ], + "right_ranges": [ + [ + 30, + 43 + ] + ] + } + ] + } + ] + }, + { + "cluster_id": "cluster-46032-3", + "summary": "Cluster of 1 issues and 2 PRs centered on issue #46032.", + "status": "open", + "confidence": 0.75, + "canonical_issue_number": 46032, + "canonical_pr_number": 46084, + "issue_numbers": [ + 46032 + ], + "pr_numbers": [ + 46084, + 46098 + ], + "evidence_types": [ + "closing_reference", + "shared_issue_target" + ], + "pr_comparisons": [ + { + "left_pr_number": 46084, + "right_pr_number": 46098, + "code_similarity": 0.43, + "size_similarity": 0.5, + "file_overlap": 0.667, + "area_overlap": 0.226, + "patch_similarity": 0.339, + "shared_filenames": [ + "src/transformers/models/mamba2/modeling_mamba2.py", + "tests/models/mamba2/test_modeling_mamba2.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/models/mamba2/modeling_mamba2.py", + "left_ranges": [ + [ + 256, + 262 + ], + [ + 303, + 353 + ], + [ + 461, + 467 + ], + [ + 471, + 491 + ], + [ + 496, + 502 + ], + [ + 599, + 609 + ] + ], + "right_ranges": [ + [ + 255, + 373 + ], + [ + 481, + 495 + ] + ] + }, + { + "filename": "tests/models/mamba2/test_modeling_mamba2.py", + "left_ranges": [ + [ + 206, + 236 + ], + [ + 285, + 322 + ] + ], + "right_ranges": [ + [ + 476, + 542 + ] + ] + } + ] + } + ] + }, + { + "cluster_id": "cluster-41115-3", + "summary": "Cluster of 1 issues and 2 PRs centered on issue #41115.", + "status": "open", + "confidence": 0.55, + "canonical_issue_number": 41115, + "canonical_pr_number": 45613, + "issue_numbers": [ + 41115 + ], + "pr_numbers": [ + 41116, + 45613 + ], + "evidence_types": [ + "closing_reference" + ], + "pr_comparisons": [ + { + "left_pr_number": 41116, + "right_pr_number": 45613, + "code_similarity": 0.805, + "size_similarity": 0.885, + "file_overlap": 0.75, + "area_overlap": 0.854, + "patch_similarity": 0.693, + "shared_filenames": [ + "docs/source/en/_toctree.yml", + "docs/source/en/model_doc/minicpm3.md", + "src/transformers/models/auto/modeling_auto.py", + "src/transformers/models/minicpm3/__init__.py", + "src/transformers/models/minicpm3/configuration_minicpm3.py", + "src/transformers/models/minicpm3/modeling_minicpm3.py", + "src/transformers/models/minicpm3/modular_minicpm3.py", + "tests/models/minicpm3/__init__.py", + "tests/models/minicpm3/test_modeling_minicpm3.py" + ], + "shared_file_areas": [ + { + "filename": "docs/source/en/_toctree.yml", + "left_ranges": [ + [ + 713, + 720 + ] + ], + "right_ranges": [ + [ + 713, + 720 + ] + ] + }, + { + "filename": "docs/source/en/model_doc/minicpm3.md", + "left_ranges": [ + [ + 1, + 67 + ] + ], + "right_ranges": [ + [ + 1, + 45 + ] + ] + }, + { + "filename": "src/transformers/models/auto/modeling_auto.py", + "left_ranges": [ + [ + 285, + 291 + ], + [ + 700, + 706 + ], + [ + 1298, + 1304 + ] + ], + "right_ranges": [ + [ + 699, + 705 + ], + [ + 1301, + 1307 + ] + ] + }, + { + "filename": "src/transformers/models/minicpm3/__init__.py", + "left_ranges": [ + [ + 1, + 29 + ] + ], + "right_ranges": [ + [ + 1, + 29 + ] + ] + }, + { + "filename": "src/transformers/models/minicpm3/configuration_minicpm3.py", + "left_ranges": [ + [ + 1, + 141 + ] + ], + "right_ranges": [ + [ + 1, + 126 + ] + ] + }, + { + "filename": "src/transformers/models/minicpm3/modeling_minicpm3.py", + "left_ranges": [ + [ + 1, + 544 + ] + ], + "right_ranges": [ + [ + 1, + 522 + ] + ] + }, + { + "filename": "src/transformers/models/minicpm3/modular_minicpm3.py", + "left_ranges": [ + [ + 1, + 444 + ] + ], + "right_ranges": [ + [ + 1, + 342 + ] + ] + }, + { + "filename": "tests/models/minicpm3/test_modeling_minicpm3.py", + "left_ranges": [ + [ + 1, + 109 + ] + ], + "right_ranges": [ + [ + 1, + 136 + ] + ] + } + ] + } + ] + }, + { + "cluster_id": "cluster-45647-3", + "summary": "Cluster of 1 issues and 2 PRs centered on issue #45647.", + "status": "open", + "confidence": 0.75, + "canonical_issue_number": 45647, + "canonical_pr_number": 45738, + "issue_numbers": [ + 45647 + ], + "pr_numbers": [ + 45737, + 45738 + ], + "evidence_types": [ + "closing_reference", + "shared_issue_target" + ], + "pr_comparisons": [ + { + "left_pr_number": 45737, + "right_pr_number": 45738, + "code_similarity": 0.526, + "size_similarity": 0.11, + "file_overlap": 0.5, + "area_overlap": 1.0, + "patch_similarity": 0.024, + "shared_filenames": [ + "src/transformers/models/musicgen_melody/modeling_musicgen_melody.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/models/musicgen_melody/modeling_musicgen_melody.py", + "left_ranges": [ + [ + 491, + 497 + ] + ], + "right_ranges": [ + [ + 491, + 497 + ] + ] + } + ] + } + ] + }, + { + "cluster_id": "cluster-45081-3", + "summary": "Cluster of 1 issues and 2 PRs centered on issue #45081.", + "status": "open", + "confidence": 0.8, + "canonical_issue_number": 45081, + "canonical_pr_number": 45317, + "issue_numbers": [ + 45081 + ], + "pr_numbers": [ + 45086, + 45317 + ], + "evidence_types": [ + "closing_reference", + "shared_issue_target", + "soft_similarity" + ], + "pr_comparisons": [ + { + "left_pr_number": 45086, + "right_pr_number": 45317, + "code_similarity": 0.589, + "size_similarity": 0.34, + "file_overlap": 0.5, + "area_overlap": 1.0, + "patch_similarity": 0.136, + "shared_filenames": [ + "src/transformers/tokenization_utils_tokenizers.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/tokenization_utils_tokenizers.py", + "left_ranges": [ + [ + 1360, + 1370 + ], + [ + 1374, + 1380 + ] + ], + "right_ranges": [ + [ + 1360, + 1370 + ], + [ + 1374, + 1380 + ] + ] + } + ] + } + ] + }, + { + "cluster_id": "cluster-45561-3", + "summary": "Cluster of 1 issues and 2 PRs centered on issue #45561.", + "status": "open", + "confidence": 0.8, + "canonical_issue_number": 45561, + "canonical_pr_number": 45645, + "issue_numbers": [ + 45561 + ], + "pr_numbers": [ + 45639, + 45645 + ], + "evidence_types": [ + "closing_reference", + "shared_issue_target", + "soft_similarity" + ], + "pr_comparisons": [ + { + "left_pr_number": 45639, + "right_pr_number": 45645, + "code_similarity": 0.532, + "size_similarity": 0.619, + "file_overlap": 0.5, + "area_overlap": 0.543, + "patch_similarity": 0.451, + "shared_filenames": [ + "src/transformers/testing_utils.py", + "tests/utils/test_testing_utils.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/testing_utils.py", + "left_ranges": [ + [ + 3529, + 3536 + ], + [ + 3753, + 3787 + ] + ], + "right_ranges": [ + [ + 3525, + 3558 + ], + [ + 3782, + 3788 + ] + ] + }, + { + "filename": "tests/utils/test_testing_utils.py", + "left_ranges": [ + [ + 1, + 86 + ] + ], + "right_ranges": [ + [ + 1, + 114 + ] + ] + } + ] + } + ] + }, + { + "cluster_id": "cluster-43698-3", + "summary": "Cluster of 1 issues and 2 PRs centered on issue #43698.", + "status": "open", + "confidence": 0.75, + "canonical_issue_number": 43698, + "canonical_pr_number": 43779, + "issue_numbers": [ + 43698 + ], + "pr_numbers": [ + 43779, + 43816 + ], + "evidence_types": [ + "closing_reference", + "shared_issue_target" + ], + "pr_comparisons": [ + { + "left_pr_number": 43779, + "right_pr_number": 43816, + "code_similarity": 0.418, + "size_similarity": 0.538, + "file_overlap": 1.0, + "area_overlap": 0.02, + "patch_similarity": 0.02, + "shared_filenames": [ + "src/transformers/integrations/integration_utils.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/integrations/integration_utils.py", + "left_ranges": [ + [ + 2227, + 2241 + ], + [ + 2303, + 2309 + ] + ], + "right_ranges": [ + [ + 2278, + 2291 + ], + [ + 2309, + 2322 + ] + ] + } + ] + } + ] + }, + { + "cluster_id": "cluster-46100-2", + "summary": "Cluster of 2 related pull requests linked by soft_similarity.", + "status": "open", + "confidence": 0.5, + "canonical_issue_number": null, + "canonical_pr_number": 46100, + "issue_numbers": [], + "pr_numbers": [ + 46100, + 46114 + ], + "evidence_types": [ + "soft_similarity" + ], + "pr_comparisons": [ + { + "left_pr_number": 46100, + "right_pr_number": 46114, + "code_similarity": 1.0, + "size_similarity": 1.0, + "file_overlap": 1.0, + "area_overlap": 1.0, + "patch_similarity": 1.0, + "shared_filenames": [ + "tests/models/vit/test_modeling_vit.py" + ], + "shared_file_areas": [ + { + "filename": "tests/models/vit/test_modeling_vit.py", + "left_ranges": [ + [ + 216, + 226 + ] + ], + "right_ranges": [ + [ + 216, + 226 + ] + ] + } + ] + } + ] + }, + { + "cluster_id": "cluster-44018-2", + "summary": "Cluster of 2 related pull requests linked by soft_similarity.", + "status": "open", + "confidence": 0.5, + "canonical_issue_number": null, + "canonical_pr_number": 44068, + "issue_numbers": [], + "pr_numbers": [ + 44018, + 44068 + ], + "evidence_types": [ + "soft_similarity" + ], + "pr_comparisons": [ + { + "left_pr_number": 44018, + "right_pr_number": 44068, + "code_similarity": 0.766, + "size_similarity": 0.939, + "file_overlap": 1.0, + "area_overlap": 0.425, + "patch_similarity": 0.866, + "shared_filenames": [ + "src/transformers/models/gpt_neo/modeling_gpt_neo.py", + "tests/models/gpt_neo/test_modeling_gpt_neo.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/models/gpt_neo/modeling_gpt_neo.py", + "left_ranges": [ + [ + 26, + 46 + ], + [ + 138, + 143 + ], + [ + 182, + 187 + ], + [ + 283, + 295 + ], + [ + 327, + 341 + ], + [ + 348, + 354 + ], + [ + 360, + 369 + ], + [ + 399, + 406 + ], + [ + 411, + 419 + ], + [ + 428, + 433 + ], + [ + 472, + 492 + ], + [ + 507, + 513 + ], + [ + 519, + 528 + ], + [ + 541, + 559 + ], + [ + 562, + 567 + ], + [ + 595, + 601 + ], + [ + 607, + 614 + ], + [ + 627, + 643 + ], + [ + 685, + 690 + ], + [ + 708, + 714 + ], + [ + 720, + 727 + ], + [ + 740, + 757 + ], + [ + 761, + 766 + ], + [ + 780, + 786 + ], + [ + 791, + 798 + ], + [ + 807, + 822 + ], + [ + 840, + 845 + ] + ], + "right_ranges": [ + [ + 26, + 31 + ], + [ + 34, + 44 + ], + [ + 136, + 141 + ], + [ + 180, + 185 + ], + [ + 281, + 293 + ], + [ + 325, + 339 + ], + [ + 346, + 352 + ], + [ + 358, + 367 + ], + [ + 397, + 404 + ], + [ + 409, + 417 + ], + [ + 426, + 434 + ], + [ + 463, + 483 + ], + [ + 498, + 504 + ], + [ + 510, + 519 + ], + [ + 532, + 550 + ], + [ + 553, + 564 + ], + [ + 586, + 592 + ], + [ + 598, + 605 + ], + [ + 618, + 634 + ], + [ + 676, + 688 + ], + [ + 699, + 705 + ], + [ + 711, + 718 + ], + [ + 731, + 748 + ], + [ + 752, + 762 + ], + [ + 771, + 777 + ], + [ + 782, + 789 + ], + [ + 798, + 813 + ], + [ + 831, + 836 + ] + ] + }, + { + "filename": "tests/models/gpt_neo/test_modeling_gpt_neo.py", + "left_ranges": [ + [ + 458, + 464 + ] + ], + "right_ranges": [ + [ + 458, + 464 + ] + ] + } + ] + } + ] + } + ], + "duplicate_issues": [], + "duplicate_prs": [ + { + "cluster_id": "cluster-41115-3", + "canonical_pr_number": 45613, + "duplicate_pr_numbers": [ + 41116 + ], + "target_issue_number": 41115, + "reason": "PRs in cluster-41115-3 are treated as duplicates because they converge on issue #41115 with closing_reference evidence." + }, + { + "cluster_id": "cluster-41211-3", + "canonical_pr_number": 44339, + "duplicate_pr_numbers": [ + 41356 + ], + "target_issue_number": 41211, + "reason": "PRs in cluster-41211-3 are treated as duplicates because they converge on issue #41211 with closing_reference evidence." + }, + { + "cluster_id": "cluster-43240-3", + "canonical_pr_number": 43251, + "duplicate_pr_numbers": [ + 43254 + ], + "target_issue_number": 43240, + "reason": "PRs in cluster-43240-3 are treated as duplicates because they converge on issue #43240 with closing_reference, shared_issue_target evidence." + }, + { + "cluster_id": "cluster-43698-3", + "canonical_pr_number": 43779, + "duplicate_pr_numbers": [ + 43816 + ], + "target_issue_number": 43698, + "reason": "PRs in cluster-43698-3 are treated as duplicates because they converge on issue #43698 with closing_reference, shared_issue_target evidence." + }, + { + "cluster_id": "cluster-43979-11", + "canonical_pr_number": 44007, + "duplicate_pr_numbers": [ + 43996, + 44013, + 44044, + 44066, + 44072, + 44085, + 44129, + 44154, + 44722 + ], + "target_issue_number": 43979, + "reason": "PRs in cluster-43979-11 are treated as duplicates because they converge on issue #43979 with closing_reference, shared_issue_target, soft_similarity evidence." + }, + { + "cluster_id": "cluster-44018-2", + "canonical_pr_number": 44068, + "duplicate_pr_numbers": [ + 44018 + ], + "target_issue_number": null, + "reason": "PRs in cluster-44018-2 are treated as duplicates because they share soft_similarity evidence." + }, + { + "cluster_id": "cluster-45081-3", + "canonical_pr_number": 45317, + "duplicate_pr_numbers": [ + 45086 + ], + "target_issue_number": 45081, + "reason": "PRs in cluster-45081-3 are treated as duplicates because they converge on issue #45081 with closing_reference, shared_issue_target, soft_similarity evidence." + }, + { + "cluster_id": "cluster-45561-3", + "canonical_pr_number": 45645, + "duplicate_pr_numbers": [ + 45639 + ], + "target_issue_number": 45561, + "reason": "PRs in cluster-45561-3 are treated as duplicates because they converge on issue #45561 with closing_reference, shared_issue_target, soft_similarity evidence." + }, + { + "cluster_id": "cluster-45647-3", + "canonical_pr_number": 45738, + "duplicate_pr_numbers": [ + 45737 + ], + "target_issue_number": 45647, + "reason": "PRs in cluster-45647-3 are treated as duplicates because they converge on issue #45647 with closing_reference, shared_issue_target evidence." + }, + { + "cluster_id": "cluster-46032-3", + "canonical_pr_number": 46084, + "duplicate_pr_numbers": [ + 46098 + ], + "target_issue_number": 46032, + "reason": "PRs in cluster-46032-3 are treated as duplicates because they converge on issue #46032 with closing_reference, shared_issue_target evidence." + }, + { + "cluster_id": "cluster-46100-2", + "canonical_pr_number": 46100, + "duplicate_pr_numbers": [ + 46114 + ], + "target_issue_number": null, + "reason": "PRs in cluster-46100-2 are treated as duplicates because they share soft_similarity evidence." + } + ], + "best_issue": { + "cluster_id": "cluster-45647-3", + "issue_number": 45647, + "reason": "Issue #45647 is the strongest global issue candidate because it is open, belongs to a cluster with 3 artifacts, and carries 12 discussion comments plus 8 inbound references.", + "score": 162.0 + }, + "best_pr": { + "cluster_id": "cluster-43979-11", + "pr_number": 44007, + "reason": "PR #44007 is the strongest global PR candidate because it is open, belongs to a cluster with 11 artifacts, links to 1 issues, and carries 2 review/discussion events.", + "score": 238.886 + } +} diff --git a/snapshots/20260525T120039Z/analysis-runs/hybrid-model-20260525t120039z/analysis-report-hybrid.llm-reviews.json b/snapshots/20260525T120039Z/analysis-runs/hybrid-model-20260525t120039z/analysis-report-hybrid.llm-reviews.json new file mode 100644 index 0000000000000000000000000000000000000000..8cc599bb51da47d87dc3be7d13c2807c44127c4a --- /dev/null +++ b/snapshots/20260525T120039Z/analysis-runs/hybrid-model-20260525t120039z/analysis-report-hybrid.llm-reviews.json @@ -0,0 +1,20587 @@ +{ + "schema_version": "1.0", + "repo": "huggingface/transformers", + "snapshot_id": "20260525T120039Z", + "generated_at": "2026-05-25T12:15:43Z", + "model": "gpt-5.4-mini?service_tier=flex", + "reviews": [ + { + "label": "issue", + "component_index": 1, + "component_count": 17, + "review_unit_index": 1, + "review_unit_count": 50, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-428", + "nodes": [ + "issue:24643", + "issue:30277", + "issue:34634", + "issue:34928", + "issue:35545", + "issue:36010", + "issue:36331", + "issue:39290", + "issue:41720", + "issue:41762", + "issue:42915", + "issue:43039", + "issue:43425", + "issue:43541", + "issue:43572", + "issue:43596", + "issue:43638", + "issue:43716", + "issue:43827", + "issue:43828", + "issue:43866", + "issue:43901", + "issue:43975", + "issue:44292", + "issue:44322", + "issue:44457", + "issue:44509", + "issue:44512", + "issue:44560", + "issue:44805", + "issue:44841", + "issue:44863", + "issue:44918", + "issue:45081", + "issue:45161", + "issue:45237", + "issue:45464", + "issue:45507", + "issue:45685", + "issue:45812", + "issue:45823", + "issue:45859", + "issue:45874", + "issue:45901", + "issue:45902", + "issue:45920", + "issue:46015", + "issue:46165" + ], + "soft_pairs": [ + "issue:44322|issue:45464", + "issue:45237|issue:45685", + "issue:43716|issue:45237", + "issue:43828|issue:45237", + "issue:45237|issue:45507", + "issue:44292|issue:45237", + "issue:30277|issue:35545", + "issue:35545|issue:42915", + "issue:35545|issue:41720", + "issue:45901|issue:46015", + "issue:44509|issue:44512", + "issue:35545|issue:36010", + "issue:44805|issue:45823", + "issue:44918|issue:46165", + "issue:43638|issue:44805", + "issue:43638|issue:46165", + "issue:43596|issue:44805", + "issue:44805|issue:45161", + "issue:44560|issue:46165", + "issue:43596|issue:46165", + "issue:44805|issue:44918", + "issue:43572|issue:46165", + "issue:34634|issue:35545", + "issue:30277|issue:43638", + "issue:43596|issue:43638", + "issue:43975|issue:44457", + "issue:43039|issue:45161", + "issue:44292|issue:44457", + "issue:34928|issue:44805", + "issue:24643|issue:30277", + "issue:45859|issue:45902", + "issue:43638|issue:45161", + "issue:34928|issue:36331", + "issue:43596|issue:45161", + "issue:43716|issue:45507", + "issue:45823|issue:45874", + "issue:39290|issue:45081", + "issue:43425|issue:44292", + "issue:34928|issue:41720", + "issue:43828|issue:45161", + "issue:44918|issue:45161", + "issue:43828|issue:44292", + "issue:43716|issue:45685", + "issue:43541|issue:43716", + "issue:44292|issue:44918", + "issue:43716|issue:44292", + "issue:45507|issue:45685", + "issue:43039|issue:44918", + "issue:41762|issue:43638", + "issue:43827|issue:44512", + "issue:43716|issue:44560", + "issue:43828|issue:45685", + "issue:44918|issue:45685", + "issue:43039|issue:43828", + "issue:30277|issue:36331", + "issue:45237|issue:46015", + "issue:34928|issue:36010", + "issue:43039|issue:44805", + "issue:41762|issue:43596", + "issue:39290|issue:41720", + "issue:43901|issue:44512", + "issue:45237|issue:45901", + "issue:44841|issue:45823", + "issue:43039|issue:44560", + "issue:43975|issue:46015", + "issue:45812|issue:45920", + "issue:44863|issue:46015", + "issue:43866|issue:44863", + "issue:44292|issue:46015" + ], + "prepared_review_unit_hash": "f9259cba0923024ffa657b5f2d2363fbc928d0316cbe6b0d68e1f33f51b3e507", + "component_budget": { + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 69, + "serialized_chars": 48403, + "estimated_input_tokens": 12101, + "estimated_eval_tokens": 24458 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "Mostly heterogeneous cluster. The only clear duplicate subgroups are the DeepSpeed ZeRO-3 empty-index failures, the AutoTokenizer token-ID regression, and possibly the Qwen3_5Moe rotary-embedding device/materialization bug. Most other soft edges are different bugs or unrelated docs/issues.", + "confidence": 0.58, + "canonical_issue_reason": "No single issue covers the whole set; issue 45920 is the best canonical anchor because it is the broadest report of the tokenizer-ID regression and generalizes beyond one model family.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 45920 is the strongest overall representative: broad symptom, current regression framing, and it already aggregates the tokenizer bug across multiple models.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44322", + "right": "issue:45464", + "accept": false, + "reason": "Different failures: a config attribute error vs a streaming chat/completions bug." + }, + { + "left": "issue:45237", + "right": "issue:45685", + "accept": false, + "reason": "Different backends and code paths: AMD GPU runtime failure vs MPS histogram int issue." + }, + { + "left": "issue:43716", + "right": "issue:45237", + "accept": false, + "reason": "Image-preprocessor dtype mismatch is not the same bug as the AMD GPU device failure." + }, + { + "left": "issue:43828", + "right": "issue:45237", + "accept": false, + "reason": "Autocast dtype mismatch and AMD GPU device failure are separate problems." + }, + { + "left": "issue:45237", + "right": "issue:45507", + "accept": false, + "reason": "Unrelated GPT-OSS AMD issue vs a GraniteMoE invalid-method call." + }, + { + "left": "issue:44292", + "right": "issue:45237", + "accept": false, + "reason": "Different model/back-end failures; no shared underlying bug." + }, + { + "left": "issue:30277", + "right": "issue:35545", + "accept": false, + "reason": "DeepSpeed collective mismatch and ONNX export failure are unrelated." + }, + { + "left": "issue:35545", + "right": "issue:42915", + "accept": false, + "reason": "ONNX export error vs FineGrainedFP8Config failure are different code paths." + }, + { + "left": "issue:35545", + "right": "issue:41720", + "accept": false, + "reason": "Export-time ONNX issue is unrelated to the Qwen3 auto-device-map CUDA assert." + }, + { + "left": "issue:45901", + "right": "issue:46015", + "accept": false, + "reason": "Table-QA runtime crash is unrelated to outdated RAG docs." + }, + { + "left": "issue:44509", + "right": "issue:44512", + "accept": false, + "reason": "Both are v5 docs regressions, but they concern different removed features and are not the same concrete fix." + }, + { + "left": "issue:35545", + "right": "issue:36010", + "accept": false, + "reason": "ModernBERT ONNX export and GenerationMixin import failure are unrelated." + }, + { + "left": "issue:44805", + "right": "issue:45823", + "accept": false, + "reason": "Mask/tensor shape mismatch is not the same as Gemma4 PLE device placement." + }, + { + "left": "issue:44918", + "right": "issue:46165", + "accept": false, + "reason": "TRL SFT input-embedding unpacking failure and training-resume/trust-remote-code issue are different bugs." + }, + { + "left": "issue:43638", + "right": "issue:44805", + "accept": false, + "reason": "ZeRO-3 empty-index loading error is unrelated to the mask/index shape mismatch." + }, + { + "left": "issue:44805", + "right": "issue:45161", + "accept": false, + "reason": "Different failures: shape mismatch vs GPT-OSS tensor-parallel behavior." + }, + { + "left": "issue:43638", + "right": "issue:46165", + "accept": false, + "reason": "ZeRO-3 Bert loading bug is unrelated to resume-training with trust-remote-code." + }, + { + "left": "issue:43596", + "right": "issue:44805", + "accept": false, + "reason": "Bert ZeRO-3 index error and mask/tensor shape mismatch are different problems." + }, + { + "left": "issue:44805", + "right": "issue:44918", + "accept": false, + "reason": "The tensor-mask mismatch is not the same as the Qwen3.5 input-embeddings/TRL issue." + }, + { + "left": "issue:43572", + "right": "issue:46165", + "accept": false, + "reason": "StableLm config pad-token regression is unrelated to training resume behavior." + }, + { + "left": "issue:34634", + "right": "issue:35545", + "accept": false, + "reason": "BarkProcessor voice_preset and ModernBERT ONNX export are unrelated." + }, + { + "left": "issue:30277", + "right": "issue:43638", + "accept": false, + "reason": "Collective-mismatch DeepSpeed bug is different from the ZeRO-3 empty-index loading bug." + }, + { + "left": "issue:43596", + "right": "issue:43638", + "accept": true, + "reason": "Same ZeRO-3 index-0-out-of-bounds loading failure on Bert; one is just a non-pretrained reproduction." + }, + { + "left": "issue:43975", + "right": "issue:44457", + "accept": false, + "reason": "Detokenization regression and LoRA save/load mismatch are unrelated." + }, + { + "left": "issue:43039", + "right": "issue:45161", + "accept": false, + "reason": "Liger cross-entropy being called and GPT-OSS tensor-parallel failure are not the same bug." + }, + { + "left": "issue:44292", + "right": "issue:44457", + "accept": false, + "reason": "Qwen-3 NVFP4 runtime error and LoRA merge/save mismatch are unrelated." + }, + { + "left": "issue:34928", + "right": "issue:44805", + "accept": false, + "reason": "Activation-checkpointing/FSDP size recomputation bug is different from the mask-shape error." + }, + { + "left": "issue:24643", + "right": "issue:30277", + "accept": false, + "reason": "DeepSpeed weight-dimension error and collective mismatch are different failures." + }, + { + "left": "issue:45859", + "right": "issue:45902", + "accept": true, + "reason": "Same Qwen3_5Moe rotary-embedding state/device-transition bug; one reports CPU offload, the other meta-to-empty materialization." + }, + { + "left": "issue:43638", + "right": "issue:45161", + "accept": false, + "reason": "ZeRO-3 Bert loading bug is unrelated to GPT-OSS tensor-parallel behavior." + }, + { + "left": "issue:34928", + "right": "issue:36331", + "accept": false, + "reason": "FSDP activation-checkpointing bug and CustomTrainer signature breakage are different issues." + }, + { + "left": "issue:43596", + "right": "issue:45161", + "accept": false, + "reason": "Bert ZeRO-3 loading failure is unrelated to GPT-OSS tensor-parallel failure." + }, + { + "left": "issue:43716", + "right": "issue:45507", + "accept": false, + "reason": "Image dtype mismatch and GraniteMoE invalid-method call are unrelated." + }, + { + "left": "issue:45823", + "right": "issue:45874", + "accept": false, + "reason": "Gemma4 device-map mismatch and expensive reverse-embedding lookup are different forward-path bugs." + }, + { + "left": "issue:39290", + "right": "issue:45081", + "accept": false, + "reason": "Gemma3 config/vLLM attribute error is unrelated to the Mistral regex patch crash." + }, + { + "left": "issue:43425", + "right": "issue:44292", + "accept": false, + "reason": "Torch version incompatibility and Qwen-3 NVFP4 runtime failure are not the same bug." + }, + { + "left": "issue:34928", + "right": "issue:41720", + "accept": false, + "reason": "FSDP checkpointing issue and Qwen3 auto-device-map CUDA assert are unrelated." + }, + { + "left": "issue:43828", + "right": "issue:45161", + "accept": false, + "reason": "Autocast dtype mismatch and GPT-OSS tensor-parallel failure are different problems." + }, + { + "left": "issue:44918", + "right": "issue:45161", + "accept": false, + "reason": "Qwen3.5/TRl input-embedding issue is unrelated to GPT-OSS tensor-parallel behavior." + }, + { + "left": "issue:43828", + "right": "issue:44292", + "accept": false, + "reason": "Autocast dtype mismatch and Qwen-3 NVFP4 runtime error are unrelated." + }, + { + "left": "issue:43716", + "right": "issue:45685", + "accept": false, + "reason": "Mistral image-preprocessor dtype mismatch and MPS histogram issue are unrelated." + }, + { + "left": "issue:43541", + "right": "issue:43716", + "accept": false, + "reason": "Mixtral torch-dynamo grouped_mm error and image dtype mismatch are different bugs." + }, + { + "left": "issue:44292", + "right": "issue:44918", + "accept": false, + "reason": "Qwen-3 NVFP4 runtime error and Qwen3.5/TRl input-embedding failure are unrelated." + }, + { + "left": "issue:43716", + "right": "issue:44292", + "accept": false, + "reason": "Image preprocessor dtype mismatch is not the same as the Qwen-3 NVFP4 issue." + }, + { + "left": "issue:45507", + "right": "issue:45685", + "accept": false, + "reason": "GraniteMoE invalid-method call and MPS histogram int error are unrelated." + }, + { + "left": "issue:43039", + "right": "issue:44918", + "accept": false, + "reason": "Liger cross-entropy routing and Qwen3.5/TRl input embedding failure are different bugs." + }, + { + "left": "issue:41762", + "right": "issue:43638", + "accept": true, + "reason": "Same ZeRO-3 empty-index loading error; Gemma3 and Bert are just different reproductions of the same root cause." + }, + { + "left": "issue:43827", + "right": "issue:44512", + "accept": false, + "reason": "Both are docs regressions in v5, but they refer to different removed features and not one shared fix." + }, + { + "left": "issue:43716", + "right": "issue:44560", + "accept": false, + "reason": "Mistral dtype mismatch and Qwen3-vl video StopIteration are unrelated." + }, + { + "left": "issue:43828", + "right": "issue:45685", + "accept": false, + "reason": "Autocast dtype mismatch and MPS histogram int failure are different backend issues." + }, + { + "left": "issue:44918", + "right": "issue:45685", + "accept": false, + "reason": "Qwen3.5/TRl input-embedding failure is unrelated to the MPS histogram issue." + }, + { + "left": "issue:43039", + "right": "issue:43828", + "accept": false, + "reason": "Liger cross-entropy routing and autocast dtype mismatch are not the same bug." + }, + { + "left": "issue:30277", + "right": "issue:36331", + "accept": false, + "reason": "DeepSpeed collective mismatch and CustomTrainer signature breakage are unrelated." + }, + { + "left": "issue:45237", + "right": "issue:46015", + "accept": false, + "reason": "AMD GPU runtime failure and outdated RAG docs are unrelated." + }, + { + "left": "issue:34928", + "right": "issue:36010", + "accept": false, + "reason": "FSDP checkpointing bug and GenerationMixin import error are unrelated." + }, + { + "left": "issue:43039", + "right": "issue:44805", + "accept": false, + "reason": "Liger kernel cross-entropy issue and tensor mask/index mismatch are different failures." + }, + { + "left": "issue:41762", + "right": "issue:43596", + "accept": true, + "reason": "Same ZeRO-3 index-0-out-of-bounds loading failure; both are the same empty-tensor path with different model repros." + }, + { + "left": "issue:39290", + "right": "issue:41720", + "accept": false, + "reason": "Gemma3/vLLM config attribute error and Qwen3 CUDA assert are unrelated." + }, + { + "left": "issue:43901", + "right": "issue:44512", + "accept": false, + "reason": "TextClassificationPipeline docs drift and run-command docs drift are separate documentation fixes." + }, + { + "left": "issue:45237", + "right": "issue:45901", + "accept": false, + "reason": "GPT-OSS AMD GPU failure and table-QA crash are unrelated." + }, + { + "left": "issue:44841", + "right": "issue:45823", + "accept": false, + "reason": "Voxtral processor failure and Gemma4 device-map mismatch are different issues." + }, + { + "left": "issue:43039", + "right": "issue:44560", + "accept": false, + "reason": "Liger kernel issue and Qwen3-vl video StopIteration are unrelated." + }, + { + "left": "issue:43975", + "right": "issue:46015", + "accept": false, + "reason": "Detokenization regression and outdated RAG docs are unrelated." + }, + { + "left": "issue:45812", + "right": "issue:45920", + "accept": true, + "reason": "Same tokenizer-ID regression; 45920 is the broader report spanning multiple model families." + }, + { + "left": "issue:44863", + "right": "issue:46015", + "accept": false, + "reason": "NemotronH checkpoint loading and RAG docs are unrelated." + }, + { + "left": "issue:43866", + "right": "issue:44863", + "accept": false, + "reason": "Checkpoint corruption and NemotronH load failure are different bugs." + }, + { + "left": "issue:44292", + "right": "issue:46015", + "accept": false, + "reason": "Qwen-3 NVFP4 runtime error and outdated RAG docs are unrelated." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The analysis is grounded in the packet and stays conservative on the soft edges. The accepted duplicate groups are plausible from the titles and descriptions (ZeRO-3 empty-index loading, AutoTokenizer token-ID regression, and the Qwen3_5Moe rotary-embedding/materialization bug), and the rejections avoid overstating weak similarity. No clear mergeability or duplicate claim looks unsupported." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 17, + "review_unit_index": 2, + "review_unit_count": 50, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-428", + "nodes": [ + "issue:36010", + "issue:39186", + "issue:39290", + "issue:40990", + "issue:41093", + "issue:41762", + "issue:43054", + "issue:43296", + "issue:43366", + "issue:43404", + "issue:43425", + "issue:43531", + "issue:43541", + "issue:43572", + "issue:43596", + "issue:43638", + "issue:43645", + "issue:43716", + "issue:43792", + "issue:43828", + "issue:43854", + "issue:43866", + "issue:43975", + "issue:43994", + "issue:44291", + "issue:44292", + "issue:44387", + "issue:44589", + "issue:44841", + "issue:44918", + "issue:44960", + "issue:45005", + "issue:45020", + "issue:45070", + "issue:45084", + "issue:45161", + "issue:45362", + "issue:45399", + "issue:45464", + "issue:45542", + "issue:45632", + "issue:45685", + "issue:45698", + "issue:45759", + "issue:45800", + "issue:45823", + "issue:45901", + "issue:45902" + ], + "soft_pairs": [ + "issue:39290|issue:43531", + "issue:43854|issue:43866", + "issue:43792|issue:45901", + "issue:41093|issue:43638", + "issue:43425|issue:45070", + "issue:43596|issue:45800", + "issue:43975|issue:45901", + "issue:43792|issue:45070", + "issue:45362|issue:45464", + "issue:44292|issue:45901", + "issue:39290|issue:45823", + "issue:45464|issue:45685", + "issue:43404|issue:43866", + "issue:43541|issue:45685", + "issue:43541|issue:43828", + "issue:44841|issue:45084", + "issue:36010|issue:39290", + "issue:45800|issue:45823", + "issue:41093|issue:41762", + "issue:43296|issue:43366", + "issue:43572|issue:45070", + "issue:43541|issue:45161", + "issue:44387|issue:45005", + "issue:43828|issue:45464", + "issue:44291|issue:45070", + "issue:43366|issue:45464", + "issue:45020|issue:45800", + "issue:43531|issue:45070", + "issue:43541|issue:44589", + "issue:41093|issue:43596", + "issue:43054|issue:43994", + "issue:45759|issue:45902", + "issue:45542|issue:45800", + "issue:39186|issue:44918", + "issue:40990|issue:44960", + "issue:43716|issue:44918", + "issue:45632|issue:45698", + "issue:43645|issue:45399", + "issue:43404|issue:45005" + ], + "prepared_review_unit_hash": "f6acee94451fa58c36b13d3ddd3e24daa5390fd4aa877ab3195593ffbb69a12f", + "component_budget": { + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 39, + "serialized_chars": 39126, + "estimated_input_tokens": 9782, + "estimated_eval_tokens": 19820 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly heterogeneous: many issues share broad themes like loading errors, dtype mismatches, zero-3, or model-specific regressions, but they do not look like true duplicates. The only strong duplicate-like pair is the SigLIP2 output/embedding regression.", + "confidence": 0.68, + "canonical_issue_reason": "issue:43994 is the clearest concrete report in the only likely duplicate pair: it directly describes the bad SigLIP2 behavior with AutoModel/pipeline and is more actionable than the comparison-style report in 43054.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43994 is the best representative of the cluster because it is the most specific, reproducible, and clearly matches the one accepted duplicate edge.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:39290", + "right": "issue:43531", + "accept": false, + "reason": "Both mention model config/windowing regressions, but they are different models and different failure modes; too broad to treat as the same bug." + }, + { + "left": "issue:43854", + "right": "issue:43866", + "accept": false, + "reason": "One is a unit-test model load failure, the other is a corrupted checkpoint report; not the same underlying issue." + }, + { + "left": "issue:43792", + "right": "issue:45901", + "accept": false, + "reason": "Whisper loading failure and table-question-answering crash are unrelated code paths and symptoms." + }, + { + "left": "issue:41093", + "right": "issue:43638", + "accept": false, + "reason": "Both involve zero-3/index errors, but the triggers and stack-level symptoms differ enough that this does not look like one concrete bug." + }, + { + "left": "issue:43425", + "right": "issue:45070", + "accept": false, + "reason": "Torch version incompatibility and a pydantic PretrainedConfig-field regression are unrelated." + }, + { + "left": "issue:43596", + "right": "issue:45800", + "accept": false, + "reason": "These are different issues in different areas: zero-3/Bert loading vs a torch-transformers version incompatibility." + }, + { + "left": "issue:43975", + "right": "issue:45901", + "accept": false, + "reason": "Incorrect detokenization and a table QA crash are not the same bug." + }, + { + "left": "issue:43792", + "right": "issue:45070", + "accept": false, + "reason": "Whisper cannot run and pydantic config handling are unrelated." + }, + { + "left": "issue:45362", + "right": "issue:45464", + "accept": false, + "reason": "Both mention Qwen3.5, but one is a generic chat crash and the other is a streaming inference API failure; not enough to merge as one bug." + }, + { + "left": "issue:44292", + "right": "issue:45901", + "accept": false, + "reason": "Qwen-3-8B-NVFP4 loading failure and table QA crash are unrelated." + }, + { + "left": "issue:39290", + "right": "issue:45823", + "accept": false, + "reason": "Both are model-specific regressions, but Gemma3 sliding-window behavior and Gemma4 device mismatch are different problems." + }, + { + "left": "issue:45464", + "right": "issue:45685", + "accept": false, + "reason": "Streaming chat failure and an MPS histogram/int issue are completely different code paths." + }, + { + "left": "issue:43404", + "right": "issue:43866", + "accept": false, + "reason": "Tied lm_head weights and a corrupted checkpoint are unrelated." + }, + { + "left": "issue:43541", + "right": "issue:45685", + "accept": false, + "reason": "Mixtral tracing/grouped_mm and MPS histogram failures are unrelated." + }, + { + "left": "issue:43541", + "right": "issue:43828", + "accept": false, + "reason": "Both are numeric/dtype-adjacent, but the specific model paths and failure mechanisms differ." + }, + { + "left": "issue:44841", + "right": "issue:45084", + "accept": false, + "reason": "Voxtral processor failure and a non-template-node compile error are unrelated." + }, + { + "left": "issue:36010", + "right": "issue:39290", + "accept": false, + "reason": "Importing GenerationMixin and Gemma3 sliding_window are unrelated issues." + }, + { + "left": "issue:45800", + "right": "issue:45823", + "accept": false, + "reason": "A torch/transformers compatibility complaint and a Gemma4 device-map bug are different failures." + }, + { + "left": "issue:41093", + "right": "issue:41762", + "accept": false, + "reason": "Both are zero-3-related index errors, but one is a mask-shape mismatch and the other is a model-loading out-of-bounds error; too specific to unify." + }, + { + "left": "issue:43296", + "right": "issue:43366", + "accept": false, + "reason": "PaddleOCR-VL loading in vLLM and GGUF gpt-oss support are unrelated." + }, + { + "left": "issue:43572", + "right": "issue:45070", + "accept": false, + "reason": "StableLm pad_token_idx regression and pydantic PretrainedConfig changes are not the same bug." + }, + { + "left": "issue:43541", + "right": "issue:45161", + "accept": false, + "reason": "Mixtral grouped_mm tracing and GPT-OSS MoE tensor-parallel support are different model/runtime problems." + }, + { + "left": "issue:44387", + "right": "issue:45005", + "accept": false, + "reason": "CUDA memory growth under int4 quantization and tied weights on translation models are unrelated." + }, + { + "left": "issue:43828", + "right": "issue:45464", + "accept": false, + "reason": "An autocast dtype mismatch and a Qwen3.5 streaming API crash are unrelated." + }, + { + "left": "issue:44291", + "right": "issue:45070", + "accept": false, + "reason": "init_empty_weights argument handling and a pydantic config-field regression do not match." + }, + { + "left": "issue:43366", + "right": "issue:45464", + "accept": false, + "reason": "GPT-OSS GGUF support and Qwen3.5 streaming chat failures are unrelated." + }, + { + "left": "issue:45020", + "right": "issue:45800", + "accept": false, + "reason": "Remote-code model breakage and a torch/transformers version incompatibility are too broad and not the same concrete bug." + }, + { + "left": "issue:43531", + "right": "issue:45070", + "accept": false, + "reason": "Qwen3-MoE sliding_window behavior and pydantic config handling are unrelated." + }, + { + "left": "issue:43541", + "right": "issue:44589", + "accept": false, + "reason": "Mixtral tracing errors and Float8 storage lookup failures are different runtime issues." + }, + { + "left": "issue:41093", + "right": "issue:43596", + "accept": false, + "reason": "Both mention zero-3 IndexErrors, but the exact failure surfaces and models differ enough to reject as duplicates." + }, + { + "left": "issue:43054", + "right": "issue:43994", + "accept": true, + "reason": "Same SigLIP2 regression: both reports describe bad/nonsensical SigLIP2 text embeddings or outputs compared with expected behavior." + }, + { + "left": "issue:45759", + "right": "issue:45902", + "accept": false, + "reason": "Composite Qwen config unwrapping and uninitialized rotary embedding memory are different bugs." + }, + { + "left": "issue:45542", + "right": "issue:45800", + "accept": false, + "reason": "TensorBoard/TensorFlow backend handling and torch/transformers version incompatibility are unrelated." + }, + { + "left": "issue:39186", + "right": "issue:44918", + "accept": false, + "reason": "FSDP 2-D weight error and Qwen3.5 SFT embedding unpacking are different problems." + }, + { + "left": "issue:40990", + "right": "issue:44960", + "accept": false, + "reason": "High perplexity on GPT-OSS and a GLM5 issue are unrelated." + }, + { + "left": "issue:43716", + "right": "issue:44918", + "accept": false, + "reason": "Mistral-3 image-preprocessor dtype mismatch and Qwen3.5 SFT embedding unpacking are unrelated." + }, + { + "left": "issue:45632", + "right": "issue:45698", + "accept": false, + "reason": "Both involve custom model loading, but one is a cache-path collision and the other is wrong module selection after save/load; not clearly the same concrete bug." + }, + { + "left": "issue:43645", + "right": "issue:45399", + "accept": false, + "reason": "Notebook custom-model initialization and flash-attn2 fallback gating are unrelated." + }, + { + "left": "issue:43404", + "right": "issue:45005", + "accept": false, + "reason": "Both discuss tied weights, but they affect different model families and likely different code paths; too broad to merge." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The writeup is grounded in the packet and stays conservative on the rejects. The one accepted soft pair, issue:43054 and issue:43994, is reasonably supported by the titles alone as the same SigLIP2 output-quality regression. No mergeability problems apply here." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 17, + "review_unit_index": 3, + "review_unit_count": 50, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-428", + "nodes": [ + "issue:36331", + "issue:38175", + "issue:39692", + "issue:41553", + "issue:41628", + "issue:41762", + "issue:42915", + "issue:42947", + "issue:43054", + "issue:43493", + "issue:43643", + "issue:43644", + "issue:43782", + "issue:43828", + "issue:43856", + "issue:43866", + "issue:43872", + "issue:43873", + "issue:43881", + "issue:43883", + "issue:43950", + "issue:43975", + "issue:44292", + "issue:44360", + "issue:44387", + "issue:44403", + "issue:44451", + "issue:44488", + "issue:44492", + "issue:44512", + "issue:44530", + "issue:44534", + "issue:44589", + "issue:44863", + "issue:44910", + "issue:44918", + "issue:44928", + "issue:44960", + "issue:45005", + "issue:45020", + "issue:45237", + "issue:45399", + "issue:45507", + "issue:45663", + "issue:45701", + "issue:45800", + "issue:45859", + "issue:45902" + ], + "soft_pairs": [ + "issue:42915|issue:45399", + "issue:44530|issue:45005", + "issue:43873|issue:45005", + "issue:44863|issue:45005", + "issue:44360|issue:44512", + "issue:44918|issue:45507", + "issue:44492|issue:44512", + "issue:43975|issue:44292", + "issue:44292|issue:45507", + "issue:43950|issue:44534", + "issue:43644|issue:43950", + "issue:39692|issue:43054", + "issue:43828|issue:44928", + "issue:36331|issue:41762", + "issue:41628|issue:45800", + "issue:43950|issue:45902", + "issue:43872|issue:44589", + "issue:43866|issue:44403", + "issue:43054|issue:45701", + "issue:41553|issue:43883", + "issue:38175|issue:43054", + "issue:44910|issue:45859", + "issue:45663|issue:45902", + "issue:41553|issue:43881", + "issue:42947|issue:44387", + "issue:43054|issue:43493", + "issue:44403|issue:44451", + "issue:43975|issue:45237", + "issue:43643|issue:45020", + "issue:44403|issue:44488", + "issue:42947|issue:43856", + "issue:41762|issue:42915", + "issue:43881|issue:43883", + "issue:43782|issue:44960" + ], + "prepared_review_unit_hash": "28ff74252fd5eb7c48632c059f5dc1887acb1b7e8491ed2c4e1aba2bd7c987fc", + "component_budget": { + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 34, + "serialized_chars": 37510, + "estimated_input_tokens": 9378, + "estimated_eval_tokens": 19012 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly heterogeneous, with one clear duplicate subgroup around the Transformers 5.x non-persistent-buffer corruption regression. The rest of the pairs are only broadly similar by subsystem or model family, not the same concrete bug.", + "confidence": 0.9, + "canonical_issue_reason": "issue:43644 is the earliest and broadest statement of the shared regression (\u201cfills non-persistent buffers with junk\u201d), making it the best anchor for the duplicate set; later reports like issue:43950 and issue:44534 look like narrower manifestations of the same loading bug.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43950 is the most actionable standalone report because it names the failure mode precisely (`from_pretrained()` corrupting `register_buffer(persistent=False)` state) and aligns cleanly with the duplicate reports.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:42915", + "right": "issue:45399", + "accept": false, + "reason": "Different problems: Qwen3Moe FP8 failure vs a flash-attn2 fallback gating issue." + }, + { + "left": "issue:44530", + "right": "issue:45005", + "accept": false, + "reason": "Different bugs: Qwen3.5 PagedAttention cache crash vs tied-weights handling in translation models." + }, + { + "left": "issue:43873", + "right": "issue:45005", + "accept": false, + "reason": "Quantization/offloading behavior is unrelated to tied-weights regressions." + }, + { + "left": "issue:44863", + "right": "issue:45005", + "accept": false, + "reason": "NemotronH checkpoint loading is a separate model-compatibility issue, not tied-weights handling." + }, + { + "left": "issue:44360", + "right": "issue:44512", + "accept": false, + "reason": "DSA indexer architecture discussion is unrelated to docs mentioning a removed CLI command." + }, + { + "left": "issue:44918", + "right": "issue:45507", + "accept": false, + "reason": "TRl SFT embedding unpacking and GraniteMoE invalid-method calls are different model/path failures." + }, + { + "left": "issue:44492", + "right": "issue:44512", + "accept": false, + "reason": "A typo in cache strategy docs is not the same as stale docs about the removed run command." + }, + { + "left": "issue:43975", + "right": "issue:44292", + "accept": false, + "reason": "Different model/output issues: detokenization regression vs NVFP4 runtime failure." + }, + { + "left": "issue:44292", + "right": "issue:45507", + "accept": false, + "reason": "Qwen NVFP4 runtime failure is unrelated to GraniteMoE calling an invalid method." + }, + { + "left": "issue:43950", + "right": "issue:44534", + "accept": true, + "reason": "Both report the same Transformers 5.x regression where non-persistent buffers are materialized with junk values." + }, + { + "left": "issue:43644", + "right": "issue:43950", + "accept": true, + "reason": "Same underlying bug: Transformers 5.x corrupts non-persistent buffers during loading/materialization." + }, + { + "left": "issue:39692", + "right": "issue:43054", + "accept": false, + "reason": "Both involve SigLIP2, but one is a docs/example mismatch and quantization failure while the other is degraded text embeddings." + }, + { + "left": "issue:43828", + "right": "issue:44928", + "accept": false, + "reason": "Both mention Qwen/positioning problems, but the concrete failures differ: dtype mismatch under autocast vs RLHF gradient explosion." + }, + { + "left": "issue:36331", + "right": "issue:41762", + "accept": false, + "reason": "CustomTrainer signature mismatch and Gemma3 ZeRO-3 loading error are unrelated runtime failures." + }, + { + "left": "issue:41628", + "right": "issue:45800", + "accept": false, + "reason": "AutoImageProcessor import/export breakage is not the same as a torch-version compatibility issue." + }, + { + "left": "issue:43950", + "right": "issue:45902", + "accept": false, + "reason": "Related buffer/materialization theme, but 45902 is a specific Qwen3.5 rotary-embedding uninitialized-memory bug, not the same generic from_pretrained corruption report." + }, + { + "left": "issue:43872", + "right": "issue:44589", + "accept": false, + "reason": "bitsandbytes parameter-construction incompatibility is distinct from missing Float8 storage support." + }, + { + "left": "issue:43866", + "right": "issue:44403", + "accept": false, + "reason": "Checkpoint corruption and noisy loading are different failures." + }, + { + "left": "issue:43054", + "right": "issue:45701", + "accept": false, + "reason": "Both mention tokenization/model output quality, but one is SigLIP2 text embedding degradation and the other is a general tokenization change regression." + }, + { + "left": "issue:41553", + "right": "issue:43883", + "accept": false, + "reason": "Bad AutoTokenizer error handling for Voxtral is unrelated to Molmo missing `all_tied_weights_keys`." + }, + { + "left": "issue:38175", + "right": "issue:43054", + "accept": false, + "reason": "Zero probabilities in siglip2-base and worse SigLIP2 text embeddings are different symptoms with different likely causes." + }, + { + "left": "issue:44910", + "right": "issue:45859", + "accept": false, + "reason": "Qwen3.5 flash-attention crash from 3D position_ids is not the same as CPU-offload incompatibility in rotary embedding." + }, + { + "left": "issue:45663", + "right": "issue:45902", + "accept": false, + "reason": "Gemma4 FSDP2 shared_kv_states bug is unrelated to Qwen3.5 rotary-embedding uninitialized memory." + }, + { + "left": "issue:41553", + "right": "issue:43881", + "accept": false, + "reason": "Bad AutoTokenizer error messaging for Voxtral and glm-4v-9b loading failure are separate model-loading issues." + }, + { + "left": "issue:42947", + "right": "issue:44387", + "accept": false, + "reason": "LoRA gradient-checkpointing ineffectiveness and int4 CUDA memory growth are different training/runtime regressions." + }, + { + "left": "issue:43054", + "right": "issue:43493", + "accept": false, + "reason": "Both are SigLIP2-related, but one is text embedding quality while the other is HF-vs-JAX implementation discrepancy." + }, + { + "left": "issue:44403", + "right": "issue:44451", + "accept": false, + "reason": "General loading noise and inability to load a specific model are not the same bug." + }, + { + "left": "issue:43975", + "right": "issue:45237", + "accept": false, + "reason": "Detokenization regression and AMD GPU runtime incompatibility are unrelated." + }, + { + "left": "issue:43643", + "right": "issue:45020", + "accept": false, + "reason": "Missing fields from AutoConfig with trust_remote_code is a different issue than models broadly breaking under recent transformers versions." + }, + { + "left": "issue:44403", + "right": "issue:44488", + "accept": false, + "reason": "Both are loading-related, but they target different models and failure modes." + }, + { + "left": "issue:42947", + "right": "issue:43856", + "accept": false, + "reason": "LoRA checkpointing behavior and Qwen3 MoE memory usage are different optimization/training issues." + }, + { + "left": "issue:41762", + "right": "issue:42915", + "accept": false, + "reason": "Gemma3 ZeRO-3 loading failure and Qwen3Moe FP8 failure are unrelated model-loading bugs." + }, + { + "left": "issue:43881", + "right": "issue:43883", + "accept": false, + "reason": "Different model load failures: glm-4v-9b loading vs Molmo tied-weights attribute error." + }, + { + "left": "issue:43782", + "right": "issue:44960", + "accept": false, + "reason": "Qwen3VL weight-only loading error and GLM5 are separate model-specific failures." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The analysis is grounded in the packet and stays conservative overall. The canonical cluster anchor around issue:43644 / issue:43950 / issue:44534 is well-supported by the titles, timestamps, and the shared non-persistent-buffer corruption wording. The remaining soft-edge rejections are cautious and consistent with the available evidence, and the accepted pairings are the strongest clear duplicate links in the set." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 17, + "review_unit_index": 4, + "review_unit_count": 50, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-428", + "nodes": [ + "issue:24643", + "issue:30064", + "issue:36331", + "issue:41093", + "issue:41553", + "issue:42491", + "issue:42915", + "issue:42947", + "issue:43257", + "issue:43366", + "issue:43404", + "issue:43638", + "issue:43742", + "issue:43792", + "issue:43824", + "issue:43827", + "issue:43828", + "issue:43854", + "issue:43866", + "issue:43872", + "issue:43940", + "issue:43950", + "issue:44220", + "issue:44291", + "issue:44368", + "issue:44479", + "issue:44488", + "issue:44509", + "issue:44530", + "issue:44560", + "issue:44661", + "issue:44805", + "issue:44863", + "issue:44928", + "issue:44991", + "issue:45020", + "issue:45081", + "issue:45305", + "issue:45313", + "issue:45325", + "issue:45356", + "issue:45399", + "issue:45538", + "issue:45588", + "issue:45685", + "issue:45823", + "issue:45902", + "issue:45923" + ], + "soft_pairs": [ + "issue:45685|issue:45823", + "issue:42491|issue:44368", + "issue:44991|issue:45399", + "issue:43824|issue:43872", + "issue:45020|issue:45823", + "issue:44661|issue:45020", + "issue:43828|issue:45823", + "issue:45081|issue:45399", + "issue:45020|issue:45356", + "issue:43742|issue:45399", + "issue:43827|issue:44509", + "issue:43404|issue:43950", + "issue:43638|issue:45823", + "issue:42915|issue:45020", + "issue:43792|issue:44220", + "issue:43257|issue:43866", + "issue:43872|issue:44291", + "issue:42915|issue:43872", + "issue:41553|issue:44991", + "issue:36331|issue:41093", + "issue:43257|issue:45305", + "issue:42947|issue:45923", + "issue:43940|issue:45313", + "issue:43854|issue:44863", + "issue:44928|issue:45588", + "issue:43872|issue:45538", + "issue:41093|issue:44805", + "issue:41553|issue:44488", + "issue:44479|issue:44560", + "issue:44530|issue:44863", + "issue:24643|issue:30064", + "issue:45325|issue:45902", + "issue:43366|issue:43828" + ], + "prepared_review_unit_hash": "faabc12376afaeba77f54a8cd36a21250ffeb99d184495018aa61abf6f726e8a", + "component_budget": { + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 33, + "serialized_chars": 38072, + "estimated_input_tokens": 9518, + "estimated_eval_tokens": 19292 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "Mostly a heterogeneous cluster of unrelated bugs. I only see a few true duplicate-style matches: two docs issues about stale v5 pipeline references, two mask-shape mismatch reports, and one `_is_hf_initialized` initialization regression. The rest are similar-at-a-glance but not the same underlying problem.", + "confidence": 0.74, + "canonical_issue_reason": "There is no single canonical issue for the whole cluster; 44509 is the cleanest anchor only for the docs/pipeline-removal duplicate pair because it is the broader, more explicit report.", + "canonical_pr_reason": null, + "best_issue_reason": "44509 is the strongest overall issue candidate among the confirmed duplicates: it clearly states the v5 docs regression and subsumes 43827.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45685", + "right": "issue:45823", + "accept": false, + "reason": "Different models and failures: MPS/int histogram error vs Gemma4 PLE device mismatch." + }, + { + "left": "issue:42491", + "right": "issue:44368", + "accept": false, + "reason": "Both involve Qwen3 LoRA/configuration, but one is a cross-version compatibility break and the other is a tie_word_embeddings warning; not the same bug." + }, + { + "left": "issue:44991", + "right": "issue:45399", + "accept": false, + "reason": "Tokenizer loading regression vs flash-attn fallback gating; different code paths and symptoms." + }, + { + "left": "issue:43824", + "right": "issue:43872", + "accept": false, + "reason": "Import error for a model class vs bitsandbytes `_is_hf_initialized` incompatibility; unrelated failures." + }, + { + "left": "issue:45020", + "right": "issue:45823", + "accept": false, + "reason": "Broad remote_code breakage vs a specific device-map forward mismatch; not the same underlying issue." + }, + { + "left": "issue:44661", + "right": "issue:45020", + "accept": false, + "reason": "A tooling failure in add-new-model-like vs a runtime remote_code regression; different problems." + }, + { + "left": "issue:43828", + "right": "issue:45823", + "accept": false, + "reason": "Autocast dtype mismatch on Phi-tiny-MoE is separate from Gemma4 device placement." + }, + { + "left": "issue:45081", + "right": "issue:45399", + "accept": false, + "reason": "Mistral tokenizer regex patch crash is unrelated to flash-attn fallback logic." + }, + { + "left": "issue:45020", + "right": "issue:45356", + "accept": false, + "reason": "Different tokenizer regressions with different failure modes; not one concrete bug." + }, + { + "left": "issue:43742", + "right": "issue:45399", + "accept": false, + "reason": "Generic model load failure vs flash-attn fallback checks; no shared code-path evidence." + }, + { + "left": "issue:43827", + "right": "issue:44509", + "accept": true, + "reason": "Same underlying docs regression: both report stale references to pipeline tasks removed in v5; 44509 is the broader formulation." + }, + { + "left": "issue:43404", + "right": "issue:43950", + "accept": false, + "reason": "lm_head tying bug vs non-persistent buffer corruption; different state-handling issues." + }, + { + "left": "issue:43638", + "right": "issue:45823", + "accept": false, + "reason": "DeepSpeed ZeRO-3 index error on a Bert model vs Gemma4 device mismatch; unrelated." + }, + { + "left": "issue:42915", + "right": "issue:45020", + "accept": false, + "reason": "Qwen3 MoE + FineGrainedFP8Config vs broad remote_code breakage; too different." + }, + { + "left": "issue:43792", + "right": "issue:44220", + "accept": false, + "reason": "Whisper failing to run and a fbank helper issue may be related, but the evidence is too thin to treat them as the same bug." + }, + { + "left": "issue:43257", + "right": "issue:43866", + "accept": false, + "reason": "Qwen3 MoE weight-conversion/loading bug vs corrupted Ovis2 checkpoint; different root causes." + }, + { + "left": "issue:43872", + "right": "issue:44291", + "accept": true, + "reason": "Same `_is_hf_initialized` initialization regression showing up in bitsandbytes/Int8Params and init_empty_weights loading paths." + }, + { + "left": "issue:42915", + "right": "issue:43872", + "accept": false, + "reason": "Different Qwen3/MoE loading and FP8-related problems; no shared concrete failure." + }, + { + "left": "issue:41553", + "right": "issue:44991", + "accept": false, + "reason": "Bad error message for Voxtral tokenizer loading vs est-roberta tokenizer load failure; different issues." + }, + { + "left": "issue:36331", + "right": "issue:41093", + "accept": false, + "reason": "Trainer API signature break vs tensor/mask shape mismatch; unrelated." + }, + { + "left": "issue:43257", + "right": "issue:45305", + "accept": false, + "reason": "Qwen3 MoE weight loading under accelerate+deepspeed vs gradient averaging with GAS; different training bugs." + }, + { + "left": "issue:42947", + "right": "issue:45923", + "accept": false, + "reason": "General ineffective gradient checkpointing under PEFT LoRA vs a specific missing supports_gradient_checkpointing flag; not enough evidence they are the same bug." + }, + { + "left": "issue:43940", + "right": "issue:45313", + "accept": false, + "reason": "Both are ZeRO-3 weight-loading failures on Qwen variants, but the affected model paths and symptoms differ enough that I would not merge them." + }, + { + "left": "issue:43854", + "right": "issue:44863", + "accept": false, + "reason": "Different model-loading failures for different architectures; no concrete match." + }, + { + "left": "issue:44928", + "right": "issue:45588", + "accept": false, + "reason": "RLHF NaN/SDPA fallback problem vs flash_attention `s_aux=None` crash; separate issues." + }, + { + "left": "issue:43872", + "right": "issue:45538", + "accept": false, + "reason": "Bitsandbytes `_is_hf_initialized` incompatibility vs CLIPTokenizer max length bug; unrelated." + }, + { + "left": "issue:41093", + "right": "issue:44805", + "accept": true, + "reason": "Same mask/tensor shape mismatch class with essentially the same IndexError, just reported with different concrete sizes." + }, + { + "left": "issue:41553", + "right": "issue:44488", + "accept": false, + "reason": "Bad Voxtral error message vs failure to load a different model; no shared bug." + }, + { + "left": "issue:44479", + "right": "issue:44560", + "accept": false, + "reason": "Separate video-input regressions affecting different model families and error paths." + }, + { + "left": "issue:44530", + "right": "issue:44863", + "accept": false, + "reason": "PagedAttentionCache/linear_attention crash vs checkpoint-loading failure; unrelated." + }, + { + "left": "issue:24643", + "right": "issue:30064", + "accept": false, + "reason": "DeepSpeed weight shape error vs void segmentation-map processing; completely different subsystems." + }, + { + "left": "issue:45325", + "right": "issue:45902", + "accept": false, + "reason": "Qwen2.5-VL rope scaling bug vs Qwen3.5 MoE rotary embedding memory/materialization bug; different code paths." + }, + { + "left": "issue:43366", + "right": "issue:43828", + "accept": false, + "reason": "GGUF gpt-oss support request vs Phi-tiny-MoE autocast dtype mismatch; not the same bug." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet and stays appropriately conservative. The three accepted duplicate-style pairs are well supported by the issue titles/bodies, and the rest are cautiously rejected without overclaiming shared root cause. No mergeability concerns apply here since these are issue pairs, not PR pairs." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 17, + "review_unit_index": 5, + "review_unit_count": 50, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-428", + "nodes": [ + "issue:36010", + "issue:38175", + "issue:39692", + "issue:41720", + "issue:42491", + "issue:42617", + "issue:42915", + "issue:43317", + "issue:43366", + "issue:43531", + "issue:43550", + "issue:43596", + "issue:43632", + "issue:43638", + "issue:43644", + "issue:43645", + "issue:43646", + "issue:43708", + "issue:43720", + "issue:43824", + "issue:43950", + "issue:43957", + "issue:43994", + "issue:44075", + "issue:44155", + "issue:44451", + "issue:44488", + "issue:44534", + "issue:44560", + "issue:44589", + "issue:44661", + "issue:44756", + "issue:44805", + "issue:44877", + "issue:44910", + "issue:44918", + "issue:44960", + "issue:44977", + "issue:45030", + "issue:45276", + "issue:45313", + "issue:45357", + "issue:45584", + "issue:45663", + "issue:45685", + "issue:45823", + "issue:45834" + ], + "soft_pairs": [ + "issue:45663|issue:45823", + "issue:44155|issue:45584", + "issue:43366|issue:43531", + "issue:43950|issue:45834", + "issue:43824|issue:43957", + "issue:44560|issue:45685", + "issue:43550|issue:44910", + "issue:43645|issue:43646", + "issue:42915|issue:43957", + "issue:43366|issue:44589", + "issue:44560|issue:44918", + "issue:42617|issue:43366", + "issue:36010|issue:42915", + "issue:44661|issue:45823", + "issue:44560|issue:44805", + "issue:43632|issue:45834", + "issue:38175|issue:43994", + "issue:43638|issue:44661", + "issue:43596|issue:44661", + "issue:44661|issue:44805", + "issue:44877|issue:45030", + "issue:43317|issue:44756", + "issue:43644|issue:44534", + "issue:38175|issue:39692", + "issue:42491|issue:43720", + "issue:41720|issue:42915", + "issue:44960|issue:44977", + "issue:44451|issue:44488", + "issue:45313|issue:45357", + "issue:42915|issue:44756", + "issue:42491|issue:45276", + "issue:43708|issue:44075" + ], + "prepared_review_unit_hash": "944e43036bcbde4e806bcc5895c616fc0b7dd3e2319d6e9543a46d4dfc552d99", + "component_budget": { + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 32, + "serialized_chars": 36619, + "estimated_input_tokens": 9155, + "estimated_eval_tokens": 18566 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly a mix of unrelated transformer bugs; only a few soft pairs look like true duplicates of the same concrete regression. The clearest duplicate subcluster is the Transformers v5 non-persistent-buffer corruption report.", + "confidence": 0.78, + "canonical_issue_reason": "No single canonical issue fits the whole cluster because most items describe different bugs. If one representative is needed, issue 43644 is the best anchor for the clear duplicate pair with 44534 about non-persistent buffers being filled with junk in Transformers v5.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 43644 is the strongest representative because it predates 44534 and describes the core regression most directly: Transformers v5 corrupting non-persistent buffers during load.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45663", + "right": "issue:45823", + "accept": false, + "reason": "Different Gemma4 failures: FSDP2 shared_kv_states KeyError vs device_map='auto' PLE device mismatch." + }, + { + "left": "issue:44155", + "right": "issue:45584", + "accept": false, + "reason": "Different models and failure modes: AudioFlamingo3 batched leakage vs Whisper empty-transcription generation." + }, + { + "left": "issue:43366", + "right": "issue:43531", + "accept": false, + "reason": "Unrelated Qwen3-MoE sliding_window bug vs GGUF gpt-oss architecture support request." + }, + { + "left": "issue:43950", + "right": "issue:45834", + "accept": false, + "reason": "Non-persistent buffer regression vs Kosmos2.5 long-ocr index error are different bugs." + }, + { + "left": "issue:43824", + "right": "issue:43957", + "accept": false, + "reason": "Different loading problems: missing Qwen2_5_VL export vs meta-device loading issues across models." + }, + { + "left": "issue:44560", + "right": "issue:45685", + "accept": false, + "reason": "Different subsystems and errors: Qwen3-vl StopIteration on video vs MPS histogram not implemented for Int." + }, + { + "left": "issue:43550", + "right": "issue:44910", + "accept": false, + "reason": "Different code paths: Bamba torch.compile/SDPA failure vs Qwen3.5 Flash Attention crash from 3D position_ids." + }, + { + "left": "issue:43645", + "right": "issue:43646", + "accept": true, + "reason": "Same underlying Transformers v5 custom-model initialization regression; one is a notebook manifestation of the same breakage." + }, + { + "left": "issue:42915", + "right": "issue:43957", + "accept": false, + "reason": "Qwen3-MoE FineGrainedFP8 failure is unrelated to meta-device loading regressions." + }, + { + "left": "issue:43366", + "right": "issue:44589", + "accept": false, + "reason": "GGUF gpt-oss support is unrelated to Float8 storage lookup failures." + }, + { + "left": "issue:44560", + "right": "issue:44918", + "accept": false, + "reason": "Different Qwen3-vl problems: video StopIteration vs embedding unpacking with TRL SFT trainer." + }, + { + "left": "issue:42617", + "right": "issue:43366", + "accept": false, + "reason": "3d_parallel.py execution failure is not the same bug as GGUF gpt-oss support." + }, + { + "left": "issue:36010", + "right": "issue:42915", + "accept": false, + "reason": "ImportError for GenerationMixin is unrelated to Qwen3Moe FineGrainedFP8 failure." + }, + { + "left": "issue:44661", + "right": "issue:45823", + "accept": false, + "reason": "TOKENIZER_MAPPING_NAMES add-new-model-like bug is unrelated to Gemma4 PLE device mismatch." + }, + { + "left": "issue:44560", + "right": "issue:44805", + "accept": false, + "reason": "Different runtime errors in different Qwen3-vl contexts: video StopIteration vs mask/tensor shape mismatch." + }, + { + "left": "issue:43632", + "right": "issue:45834", + "accept": false, + "reason": "_is_hf_initialized flag regression is unrelated to Kosmos2.5 OCR index errors." + }, + { + "left": "issue:38175", + "right": "issue:43994", + "accept": true, + "reason": "Both report the same SigLIP2 bad-output regression on google/siglip2-base-patch16-224, just with different symptoms." + }, + { + "left": "issue:43638", + "right": "issue:44661", + "accept": false, + "reason": "Zero-3 BertModel indexing error and add-new-model-like TOKENIZER_MAPPING_NAMES issue are distinct." + }, + { + "left": "issue:43596", + "right": "issue:44661", + "accept": false, + "reason": "DeepSpeed ZeRO-3 BertModel index error is unrelated to TOKENIZER_MAPPING_NAMES handling." + }, + { + "left": "issue:44661", + "right": "issue:44805", + "accept": false, + "reason": "TOKENIZER_MAPPING_NAMES bug and mask/tensor shape mismatch are unrelated." + }, + { + "left": "issue:44877", + "right": "issue:45030", + "accept": false, + "reason": "Strict config loading for granite_speech is not the same as glm4v config validation failure." + }, + { + "left": "issue:43317", + "right": "issue:44756", + "accept": false, + "reason": "device_map/offload loading failure is unrelated to the Strix Halo mmap OOM workaround request." + }, + { + "left": "issue:43644", + "right": "issue:44534", + "accept": true, + "reason": "Same regression: Transformers v5 fills non-persistent buffers with junk; 44534 is a later duplicate phrasing of the same bug." + }, + { + "left": "issue:38175", + "right": "issue:39692", + "accept": false, + "reason": "SigLIP2 wrong outputs are not the same as the documentation example's model/processor mismatch and quantization errors." + }, + { + "left": "issue:42491", + "right": "issue:43720", + "accept": false, + "reason": "Qwen3 MoE LoRA hf4.x/hf5.x incompatibility is unrelated to BitNet packed-weight loading in accelerate." + }, + { + "left": "issue:41720", + "right": "issue:42915", + "accept": false, + "reason": "Qwen3 auto device-mapping cudaErrorAssert and Qwen3Moe FineGrainedFP8 are different failures." + }, + { + "left": "issue:44960", + "right": "issue:44977", + "accept": false, + "reason": "GLM5 and Qwen3.5 flash-attention generation failures are unrelated bugs." + }, + { + "left": "issue:44451", + "right": "issue:44488", + "accept": false, + "reason": "Both are model-loading failures, but they concern different checkpoints and do not clearly share the same concrete bug." + }, + { + "left": "issue:45313", + "right": "issue:45357", + "accept": false, + "reason": "Qwen3.5 ZeRO-3 weight loading and save_pretrained visual-encoder key saving are different regressions." + }, + { + "left": "issue:42915", + "right": "issue:44756", + "accept": false, + "reason": "Qwen3Moe FP8 failure is unrelated to Strix Halo mmap OOM mitigation." + }, + { + "left": "issue:42491", + "right": "issue:45276", + "accept": false, + "reason": "Qwen3 MoE LoRA compatibility is unrelated to Gemma4 resize_token_embeddings not propagating." + }, + { + "left": "issue:43708", + "right": "issue:44075", + "accept": false, + "reason": "Trainer resume/max_steps calculation and SGD args not used are different Trainer bugs." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet and stays conservative. The accepted soft pairs are plausible duplicates based on the issue titles, and the rejected pairs are framed narrowly without overclaiming. The representative issue choice for the non-persistent-buffer regression is also supported by the packet." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 17, + "review_unit_index": 6, + "review_unit_count": 50, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-428", + "nodes": [ + "issue:36010", + "issue:39404", + "issue:39692", + "issue:41720", + "issue:42491", + "issue:42947", + "issue:43257", + "issue:43335", + "issue:43381", + "issue:43404", + "issue:43425", + "issue:43454", + "issue:43644", + "issue:43645", + "issue:43854", + "issue:43856", + "issue:43873", + "issue:43931", + "issue:43957", + "issue:43994", + "issue:44155", + "issue:44164", + "issue:44291", + "issue:44292", + "issue:44387", + "issue:44402", + "issue:44488", + "issue:44756", + "issue:44910", + "issue:44912", + "issue:44928", + "issue:44960", + "issue:44991", + "issue:45216", + "issue:45276", + "issue:45292", + "issue:45335", + "issue:45362", + "issue:45406", + "issue:45488", + "issue:45588", + "issue:45663", + "issue:45704", + "issue:45759", + "issue:45812", + "issue:45823", + "issue:45859", + "issue:46050" + ], + "soft_pairs": [ + "issue:45488|issue:45812", + "issue:44488|issue:44991", + "issue:39692|issue:43994", + "issue:43644|issue:43645", + "issue:43257|issue:43854", + "issue:43854|issue:43931", + "issue:43856|issue:43873", + "issue:42491|issue:43931", + "issue:45362|issue:45406", + "issue:43856|issue:44387", + "issue:44960|issue:46050", + "issue:42491|issue:45292", + "issue:44910|issue:45588", + "issue:44164|issue:44291", + "issue:45276|issue:45335", + "issue:36010|issue:39404", + "issue:41720|issue:44155", + "issue:43404|issue:43454", + "issue:39692|issue:44960", + "issue:43425|issue:43957", + "issue:44402|issue:44488", + "issue:42947|issue:45704", + "issue:43873|issue:44387", + "issue:45663|issue:45759", + "issue:44291|issue:44756", + "issue:45663|issue:45859", + "issue:44155|issue:45823", + "issue:44292|issue:44912", + "issue:44387|issue:44756", + "issue:43335|issue:45663", + "issue:43957|issue:44756", + "issue:42947|issue:43381", + "issue:45276|issue:45292", + "issue:44164|issue:44756", + "issue:43854|issue:45216", + "issue:43381|issue:44928" + ], + "prepared_review_unit_hash": "120350c1ca5b6d606835b0f2a1df39ef2dbaa6f41d79eed764191b09103f9e63", + "component_budget": { + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 36, + "serialized_chars": 38639, + "estimated_input_tokens": 9660, + "estimated_eval_tokens": 19576 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is still mostly unrelated, with two plausible same-area issue families: `resize_token_embeddings` regressions around 45276/45292/45335 and `lm_head` weight-tying reports around 43404/43454. However, the packet only shows title-level similarity, so I would not treat those soft edges as confirmed duplicates. No PRs are present.", + "confidence": 0.61, + "canonical_issue_reason": "45276 is the best anchor for the resize-token-embedding subgroup because it is the broadest concrete report in that area, but the related items should be treated as plausible regressions rather than confirmed duplicates.", + "canonical_pr_reason": null, + "best_issue_reason": "45276 is the strongest single issue to represent the cluster\u2019s main coherent subgroup, though the evidence is still too weak to collapse the nearby variants into one confirmed duplicate set.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45488", + "right": "issue:45812", + "accept": false, + "reason": "Both are tokenizer/model-output regressions, but they affect different model families and the packet does not show the same underlying bug." + }, + { + "left": "issue:44488", + "right": "issue:44991", + "accept": false, + "reason": "Both are loading failures, but for different models and different failure descriptions; title similarity is not enough to merge them." + }, + { + "left": "issue:39692", + "right": "issue:43994", + "accept": false, + "reason": "Same model family is mentioned, but one is a docs/example problem and the other is an inference quality issue." + }, + { + "left": "issue:43644", + "right": "issue:43645", + "accept": false, + "reason": "Both are Transformers 5.0.0 regressions, but they describe different code paths and symptoms." + }, + { + "left": "issue:43257", + "right": "issue:43854", + "accept": false, + "reason": "Different model-loading defects with different targets; no shared concrete bug is evident." + }, + { + "left": "issue:43854", + "right": "issue:43931", + "accept": false, + "reason": "Both concern loading, but the model families and reported failures are different." + }, + { + "left": "issue:43856", + "right": "issue:43873", + "accept": false, + "reason": "Both mention efficiency/memory, but one is training-memory behavior and the other is quantization/offloading behavior." + }, + { + "left": "issue:42491", + "right": "issue:43931", + "accept": false, + "reason": "Distinct issues: LoRA compatibility across branches versus a shape-mismatch load failure." + }, + { + "left": "issue:45362", + "right": "issue:45406", + "accept": false, + "reason": "Different components and failure modes; no clear common code-path bug from the packet." + }, + { + "left": "issue:43856", + "right": "issue:44387", + "accept": false, + "reason": "Both are memory-related, but they point to different root causes and different workflows." + }, + { + "left": "issue:44960", + "right": "issue:46050", + "accept": false, + "reason": "Unrelated model/version report versus a save/load corruption report." + }, + { + "left": "issue:42491", + "right": "issue:45292", + "accept": false, + "reason": "Different model/version compatibility problem versus a resize-token-embedding regression." + }, + { + "left": "issue:44910", + "right": "issue:45588", + "accept": false, + "reason": "Both involve attention code, but the failure modes are different and not mergeable as one fix." + }, + { + "left": "issue:44164", + "right": "issue:44291", + "accept": false, + "reason": "Save/load extra_state handling is unrelated to init_empty_weights argument handling." + }, + { + "left": "issue:45276", + "right": "issue:45335", + "accept": false, + "reason": "They look like the same family, but the packet only gives similar titles and not enough evidence to confirm the same underlying regression." + }, + { + "left": "issue:36010", + "right": "issue:39404", + "accept": false, + "reason": "ImportError for GenerationMixin is unrelated to Whisper pipeline `return_language` behavior." + }, + { + "left": "issue:41720", + "right": "issue:44155", + "accept": false, + "reason": "Different models and failure modes; they are not the same concrete bug." + }, + { + "left": "issue:43404", + "right": "issue:43454", + "accept": false, + "reason": "Both are lm_head weight-tying complaints, but the packet only shows title similarity across different model families." + }, + { + "left": "issue:39692", + "right": "issue:44960", + "accept": false, + "reason": "Completely different topics: SigLIP2 example errors versus a GLM5 issue." + }, + { + "left": "issue:43425", + "right": "issue:43957", + "accept": false, + "reason": "Unrelated compatibility/load problems with different failure surfaces." + }, + { + "left": "issue:44402", + "right": "issue:44488", + "accept": false, + "reason": "Both are tokenizer/vocab reports, but they concern different models and different loading symptoms." + }, + { + "left": "issue:42947", + "right": "issue:45704", + "accept": false, + "reason": "Gradient-checkpointing/PEFT behavior is unrelated to the RMSNorm memory leak report." + }, + { + "left": "issue:43873", + "right": "issue:44387", + "accept": false, + "reason": "Quantization/offloading behavior and int4 OOM are different defects." + }, + { + "left": "issue:45663", + "right": "issue:45759", + "accept": false, + "reason": "Different composite-model issues affecting different code paths." + }, + { + "left": "issue:44291", + "right": "issue:44756", + "accept": false, + "reason": "Different loading/init bug versus a platform-specific mmap/OOM issue." + }, + { + "left": "issue:45663", + "right": "issue:45859", + "accept": false, + "reason": "FSDP2 attention-state bug is not the same as CPU-offload incompatibility." + }, + { + "left": "issue:44155", + "right": "issue:45823", + "accept": false, + "reason": "Different model families and unrelated forward/device mismatch behavior." + }, + { + "left": "issue:44292", + "right": "issue:44912", + "accept": false, + "reason": "Both are quantization-related, but they affect different models and failure modes." + }, + { + "left": "issue:44387", + "right": "issue:44756", + "accept": false, + "reason": "Int4 quantization OOM is unrelated to disabling mmap on Strix Halo." + }, + { + "left": "issue:43335", + "right": "issue:45663", + "accept": false, + "reason": "Sparse-layer config bug is unrelated to the Gemma-4 FSDP2 attention-state issue." + }, + { + "left": "issue:43957", + "right": "issue:44756", + "accept": false, + "reason": "Meta-device loading failures and Strix Halo mmap/OOM are not the same bug." + }, + { + "left": "issue:42947", + "right": "issue:43381", + "accept": false, + "reason": "Both mention gradient checkpointing, but they describe different constraints and code paths." + }, + { + "left": "issue:45276", + "right": "issue:45292", + "accept": false, + "reason": "This is the same title family, but the packet does not provide enough detail to confirm the exact same underlying regression." + }, + { + "left": "issue:44164", + "right": "issue:44756", + "accept": false, + "reason": "Save/load extra_state handling is unrelated to mmap-induced OOM." + }, + { + "left": "issue:43854", + "right": "issue:45216", + "accept": false, + "reason": "Different model-loading regressions with different symptoms and targets." + }, + { + "left": "issue:43381", + "right": "issue:44928", + "accept": false, + "reason": "Eval-mode checkpointing restriction is unrelated to the RLHF gradient-explosion report." + } + ] + }, + "evaluator_result": { + "accept": false, + "feedback": "The summary is mostly grounded, but it overstates confidence on the duplicate subclusters. In particular, the analyst marks 45276\u219445335 and 43404\u219443454 (and 45276\u219445292) as true duplicates even though the packet only shows similar titles, not enough to justify \u2018same underlying bug\u2019 with certainty. Those are better treated as plausible same-area regressions rather than confirmed duplicates. The overall cluster being mostly unrelated is fine, and \u2018No PRs are present\u2019 is correct, but the soft-edge acceptances are not conservative enough." + }, + "evaluator_used": true, + "retried": true, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 17, + "review_unit_index": 7, + "review_unit_count": 50, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-428", + "nodes": [ + "issue:36032", + "issue:39404", + "issue:43257", + "issue:43299", + "issue:43425", + "issue:43452", + "issue:43596", + "issue:43611", + "issue:43645", + "issue:43646", + "issue:43708", + "issue:43716", + "issue:43828", + "issue:43856", + "issue:43901", + "issue:43906", + "issue:43931", + "issue:43957", + "issue:43994", + "issue:44079", + "issue:44164", + "issue:44265", + "issue:44402", + "issue:44451", + "issue:44479", + "issue:44509", + "issue:44530", + "issue:44589", + "issue:44671", + "issue:44683", + "issue:44756", + "issue:44805", + "issue:44863", + "issue:44877", + "issue:44898", + "issue:44912", + "issue:44918", + "issue:44991", + "issue:45084", + "issue:45292", + "issue:45335", + "issue:45588", + "issue:45663", + "issue:45684", + "issue:45698", + "issue:45701", + "issue:45759", + "issue:45823" + ], + "soft_pairs": [ + "issue:45684|issue:45698", + "issue:44479|issue:44877", + "issue:45588|issue:45663", + "issue:43646|issue:43708", + "issue:43645|issue:44164", + "issue:45292|issue:45335", + "issue:43828|issue:43957", + "issue:44402|issue:44451", + "issue:43425|issue:43828", + "issue:43257|issue:44863", + "issue:39404|issue:45701", + "issue:44589|issue:44756", + "issue:44683|issue:45823", + "issue:43425|issue:45823", + "issue:43611|issue:43645", + "issue:43425|issue:43716", + "issue:44530|issue:45084", + "issue:43906|issue:44530", + "issue:39404|issue:44912", + "issue:44265|issue:44898", + "issue:44479|issue:44805", + "issue:43901|issue:44509", + "issue:43994|issue:44079", + "issue:43425|issue:44918", + "issue:43299|issue:43931", + "issue:44589|issue:45084", + "issue:44164|issue:45759", + "issue:36032|issue:43452", + "issue:44671|issue:44991", + "issue:43596|issue:44683", + "issue:39404|issue:44671", + "issue:43856|issue:44756" + ], + "prepared_review_unit_hash": "cc646bd8f79efe8341865da36b0cefa834062491e38a8cd075722dace945986f", + "component_budget": { + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 32, + "serialized_chars": 37357, + "estimated_input_tokens": 9340, + "estimated_eval_tokens": 18936 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The cluster is mostly a loose collection of distinct Transformers v5 regressions and model-specific loading/compatibility bugs. Only one soft pair looks like a true duplicate: the two resize_token_embeddings reports.", + "confidence": 0.84, + "canonical_issue_reason": "issue:43611 is the broadest representative of the generic Transformers 5 model-loading regressions in this set; the other issues are narrower model-, pipeline-, or docs-specific failures.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43611 is the best single summary item for the cluster\u2019s main theme of v5 loading/init breakages, even though most other items are not duplicates of it.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45684", + "right": "issue:45698", + "accept": false, + "reason": "Both involve custom-model save/load flows, but one is about read-only source-file permissions and the other about loading the wrong module after save_pretrained; different bugs." + }, + { + "left": "issue:44479", + "right": "issue:44877", + "accept": false, + "reason": "Different failure modes and model families: video-input regression vs strict-config loading for granite_speech." + }, + { + "left": "issue:45588", + "right": "issue:45663", + "accept": false, + "reason": "Sink-less flash_attention crash vs Gemma4 FSDP2 shared_kv_states rebuild issue; unrelated code paths." + }, + { + "left": "issue:43646", + "right": "issue:43708", + "accept": false, + "reason": "Custom model initialization in notebooks vs Trainer resume_from_checkpoint max_steps calculation; not the same bug." + }, + { + "left": "issue:43645", + "right": "issue:44164", + "accept": false, + "reason": "Notebook custom-model init regression vs save/from_pretrained extra_state handling; different underlying problems." + }, + { + "left": "issue:45292", + "right": "issue:45335", + "accept": true, + "reason": "Same bug: resize_token_embeddings updates output embeddings incorrectly, including t5gemma decoder.embed_tokens." + }, + { + "left": "issue:43828", + "right": "issue:43957", + "accept": false, + "reason": "Both are dtype/loading regressions, but one is autocast on Phi-tiny-MoE and the other is torch.device('meta') loading; distinct issues." + }, + { + "left": "issue:44402", + "right": "issue:44451", + "accept": false, + "reason": "Both concern tokenizer/model loading, but the first is a vocab-size mismatch report and the second is a generic ScandiBERT load failure; not enough evidence of a single bug." + }, + { + "left": "issue:43425", + "right": "issue:43828", + "accept": false, + "reason": "Torch version incompatibility is too broad; the concrete symptoms and code paths differ." + }, + { + "left": "issue:43257", + "right": "issue:44863", + "accept": false, + "reason": "Different model-specific checkpoint loading problems: Qwen3 MoE conversion vs NemotronH checkpoint loading." + }, + { + "left": "issue:39404", + "right": "issue:45701", + "accept": false, + "reason": "Whisper pipeline return_language regression and version-dependent tokenization changes are unrelated." + }, + { + "left": "issue:44589", + "right": "issue:44756", + "accept": false, + "reason": "Float8 storage lookup error vs disabling mmap on Strix Halo to avoid OOM; unrelated symptoms and fixes." + }, + { + "left": "issue:44683", + "right": "issue:45823", + "accept": false, + "reason": "Compiled flex_attention on torch>=2.9 and Gemma4 PLE device mismatch are different failures." + }, + { + "left": "issue:43425", + "right": "issue:45823", + "accept": false, + "reason": "Generic torch incompatibility does not match the specific Gemma4 PLE device-mapping bug." + }, + { + "left": "issue:43611", + "right": "issue:43645", + "accept": false, + "reason": "Both are Transformers 5 loading/init regressions, but one is base_model_prefix loading and the other is Jupyter custom-model initialization; separate bugs." + }, + { + "left": "issue:43425", + "right": "issue:43716", + "accept": false, + "reason": "Torch incompatibility vs Mistral-3 image preprocessor/model dtype mismatch; not the same issue." + }, + { + "left": "issue:44530", + "right": "issue:45084", + "accept": false, + "reason": "PagedAttentionCache linear_attention crash and non-template node compilation error are different failures." + }, + { + "left": "issue:43906", + "right": "issue:44530", + "accept": false, + "reason": "An isolated reproduction of another issue vs a Qwen3.5 PagedAttentionCache crash; not duplicates." + }, + { + "left": "issue:39404", + "right": "issue:44912", + "accept": false, + "reason": "Whisper pipeline behavior and MXFP4 quantization fallback are unrelated." + }, + { + "left": "issue:44265", + "right": "issue:44898", + "accept": false, + "reason": "torch.export with torch_compilable_check and Perceiver image classification failure are different model/code-path issues." + }, + { + "left": "issue:44479", + "right": "issue:44805", + "accept": false, + "reason": "Video-input regression for Qwen VL models vs a generic mask/tensor shape mismatch; no clear duplicate evidence." + }, + { + "left": "issue:43901", + "right": "issue:44509", + "accept": false, + "reason": "Both are docs-related, but one is about return_all_scores docs drift and the other about removed pipeline tasks; separate documentation issues." + }, + { + "left": "issue:43994", + "right": "issue:44079", + "accept": false, + "reason": "Siglip2 nonsensical outputs and ModelOutput key assignment are different bugs." + }, + { + "left": "issue:43425", + "right": "issue:44918", + "accept": false, + "reason": "Generic torch incompatibility vs Qwen3.5 embedding unpacking with TRL SFT trainer; unrelated." + }, + { + "left": "issue:43299", + "right": "issue:43931", + "accept": false, + "reason": "Both are Qwen3-VL loading reports, but one is a MoE loading regression in transformers 5.0.0.dev0 and the other is a weight-shape mismatch for a specific checkpoint; not the same concrete bug." + }, + { + "left": "issue:44589", + "right": "issue:45084", + "accept": false, + "reason": "Different compile/runtime failures with no shared code path." + }, + { + "left": "issue:44164", + "right": "issue:45759", + "accept": false, + "reason": "extra_state save/load handling is unrelated to AutoModelForCausalLM.from_config text_config unwrapping." + }, + { + "left": "issue:36032", + "right": "issue:43452", + "accept": false, + "reason": "Tokenizer special-token attribute conflict vs gguf_file breaking AutoTokenizer/AutoModelForCausalLM.from_pretrained; distinct problems." + }, + { + "left": "issue:44671", + "right": "issue:44991", + "accept": false, + "reason": "CamemBERT masked-LM prediction regression and EMBEDDIA tokenizer loading failure are different bugs." + }, + { + "left": "issue:43596", + "right": "issue:44683", + "accept": false, + "reason": "Deepspeed ZeRO3/BertModel index error and compiled flex_attention on torch>=2.9 are unrelated." + }, + { + "left": "issue:39404", + "right": "issue:44671", + "accept": false, + "reason": "Whisper pipeline regression and CamemBERT masked-LM regression affect different models and code paths." + }, + { + "left": "issue:43856", + "right": "issue:44756", + "accept": false, + "reason": "Qwen3 MoE training memory usage and Strix Halo mmap OOM avoidance are different performance/memory issues, not duplicates." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet and stays conservative: it characterizes the cluster as mostly distinct v5 loading/init regressions, and it only treats the resize_token_embeddings pair (issue:45292 / issue:45335) as a true duplicate, which is well supported by the titles and linkage. The other soft-edge denials are phrased cautiously and do not overstate duplicate evidence." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 17, + "review_unit_index": 8, + "review_unit_count": 50, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-428", + "nodes": [ + "issue:32090", + "issue:34567", + "issue:36032", + "issue:39404", + "issue:40444", + "issue:40990", + "issue:42175", + "issue:42915", + "issue:43089", + "issue:43262", + "issue:43317", + "issue:43377", + "issue:43421", + "issue:43452", + "issue:43482", + "issue:43541", + "issue:43582", + "issue:43611", + "issue:43632", + "issue:43643", + "issue:43644", + "issue:43645", + "issue:43646", + "issue:43716", + "issue:43883", + "issue:44164", + "issue:44206", + "issue:44220", + "issue:44451", + "issue:44464", + "issue:44479", + "issue:44488", + "issue:44530", + "issue:44661", + "issue:44683", + "issue:44737", + "issue:44811", + "issue:44849", + "issue:44987", + "issue:44991", + "issue:45081", + "issue:45084", + "issue:45325", + "issue:45507", + "issue:45584", + "issue:45859", + "issue:45987", + "issue:46178" + ], + "soft_pairs": [ + "issue:44991|issue:45081", + "issue:32090|issue:39404", + "issue:44164|issue:44479", + "issue:44737|issue:45859", + "issue:43582|issue:43632", + "issue:44661|issue:44683", + "issue:44488|issue:45081", + "issue:44488|issue:44987", + "issue:43611|issue:43644", + "issue:40990|issue:43421", + "issue:43716|issue:45084", + "issue:44206|issue:44220", + "issue:42175|issue:46178", + "issue:36032|issue:42915", + "issue:43883|issue:44488", + "issue:43643|issue:44530", + "issue:34567|issue:40444", + "issue:44811|issue:45584", + "issue:42915|issue:44530", + "issue:44451|issue:44987", + "issue:43541|issue:45084", + "issue:43644|issue:43646", + "issue:44464|issue:45987", + "issue:43452|issue:43482", + "issue:36032|issue:43645", + "issue:44849|issue:45325", + "issue:43317|issue:43482", + "issue:44987|issue:44991", + "issue:43883|issue:44991", + "issue:45084|issue:45507", + "issue:43262|issue:43377", + "issue:43089|issue:45987" + ], + "prepared_review_unit_hash": "130286c9eb703ed8cbcffb2f777d229f885238245502ce7aa3682eec97473f6a", + "component_budget": { + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 32, + "serialized_chars": 37342, + "estimated_input_tokens": 9336, + "estimated_eval_tokens": 18928 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly a loose bag of unrelated issue reports. A few titles share broad themes (Transformers v5 regressions, Whisper/audio, model loading), but none of the soft-edge pairs look like the same concrete bug/change.", + "confidence": 0.92, + "canonical_issue_reason": "Issue 43646 is the broadest Transformers-v5 regression report in the set and the best umbrella representative if one must be chosen, but the cluster does not contain a true duplicate set.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 43646 is the best single issue to anchor the cluster because it is the most general and sits near several other v5 initialization/loading regressions; still, it should be treated as a loose thematic representative, not a duplicate canonical.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44991", + "right": "issue:45081", + "accept": false, + "reason": "Both are loading failures, but one is a tokenizer load regression for EMBEDDIA/est-roberta and the other is a Mistral regex patch crash; different code paths and symptoms." + }, + { + "left": "issue:32090", + "right": "issue:39404", + "accept": false, + "reason": "Trainer GPU broadcast NoneType error and Whisper return_language pipeline regression are unrelated subsystems and failure modes." + }, + { + "left": "issue:44164", + "right": "issue:44479", + "accept": false, + "reason": "save/from_pretrained extra_state handling and Qwen video-input regression are different bugs with no shared concrete code path." + }, + { + "left": "issue:44737", + "right": "issue:45859", + "accept": false, + "reason": "Both mention device/offload behavior, but XLNet CPU arange placement and Qwen3_5Moe CPU offload are distinct model-specific issues." + }, + { + "left": "issue:43582", + "right": "issue:43632", + "accept": false, + "reason": "Apple Silicon allocator warmup TypeError and _is_hf_initialized flag breakage are unrelated regression areas." + }, + { + "left": "issue:44661", + "right": "issue:44683", + "accept": false, + "reason": "Tokenizer mapping-name handling and compiled flex_attention on torch>=2.9 are separate features and failure paths." + }, + { + "left": "issue:44488", + "right": "issue:45081", + "accept": false, + "reason": "Both are tokenization/model-loading reports, but one is a model load failure for cjvt/sleng-bert and the other is a Mistral-specific regex patch crash." + }, + { + "left": "issue:44488", + "right": "issue:44987", + "accept": false, + "reason": "Both fail on model loading, but they target different models and likely different underlying regressions." + }, + { + "left": "issue:43611", + "right": "issue:43644", + "accept": false, + "reason": "Both are Transformers v5 regressions, but base_model_prefix loading and non-persistent buffer initialization are not the same bug." + }, + { + "left": "issue:40990", + "right": "issue:43421", + "accept": false, + "reason": "High perplexity on gpt-oss and TokenizersBackend post-processor updates are unrelated." + }, + { + "left": "issue:43716", + "right": "issue:45084", + "accept": false, + "reason": "Image preprocessor dtype mismatch and template-node compilation failure are unrelated code paths." + }, + { + "left": "issue:44206", + "right": "issue:44220", + "accept": false, + "reason": "Both are audio feature-extraction related, but one is an unsupported center argument in LasrFeatureExtractor and the other is a different _torch_extract_fbank_features issue." + }, + { + "left": "issue:42175", + "right": "issue:46178", + "accept": false, + "reason": "Backend packaging omissions for tensorflow vs tensorflow_text mapping errors are different backend-definition problems." + }, + { + "left": "issue:36032", + "right": "issue:42915", + "accept": false, + "reason": "T5 tokenizer add_special_tokens name conflict and Qwen3Moe FP8 config failure are unrelated." + }, + { + "left": "issue:43883", + "right": "issue:44488", + "accept": false, + "reason": "Molmo tied-weights attribute error and tokenizer loading failures for sleng-bert are unrelated." + }, + { + "left": "issue:43643", + "right": "issue:44530", + "accept": false, + "reason": "trust_remote_code config field loss and PagedAttentionCache linear_attention crash are different bugs." + }, + { + "left": "issue:34567", + "right": "issue:40444", + "accept": false, + "reason": "TrainerState token counting and multi-image IterableDataset finetuning failure are unrelated." + }, + { + "left": "issue:44811", + "right": "issue:45584", + "accept": false, + "reason": "Both involve Whisper special-token handling, but batch_decode skip_special_tokens and empty-transcription generation after align_special_tokens are different concrete failures." + }, + { + "left": "issue:42915", + "right": "issue:44530", + "accept": false, + "reason": "Qwen3Moe FineGrainedFP8Config failure and Qwen3.5 PagedAttentionCache linear_attention crash are distinct issues." + }, + { + "left": "issue:44451", + "right": "issue:44987", + "accept": false, + "reason": "Cannot load ScandiBERT and transformers>=5.1.0 fails loading physical-intelligence/fast are separate model-loading regressions." + }, + { + "left": "issue:43541", + "right": "issue:45084", + "accept": false, + "reason": "Mixtral grouped_mm torchdynamo tracing failure and template-node compilation error are unrelated." + }, + { + "left": "issue:43644", + "right": "issue:43646", + "accept": false, + "reason": "Both are broad Transformers v5 regressions, but one is about non-persistent buffers and the other about custom model initialization; too different to merge." + }, + { + "left": "issue:44464", + "right": "issue:45987", + "accept": false, + "reason": "Chunked generation inconsistency with compiled forward and StaticCache seq-length type violation are different generation/cache bugs." + }, + { + "left": "issue:43452", + "right": "issue:43482", + "accept": false, + "reason": "Both touch GGUF loading, but one is a general gguf_file breakage for tokenizer/model loaders and the other is a Qwen2.5-GGUF-specific v5 failure." + }, + { + "left": "issue:36032", + "right": "issue:43645", + "accept": false, + "reason": "Tokenizer special-token method conflict and Jupyter custom model initialization are unrelated." + }, + { + "left": "issue:44849", + "right": "issue:45325", + "accept": false, + "reason": "Qwen3.5 hidden-states bug and Qwen2.5-VL rope-index temporal scaling bug are different model-specific regressions." + }, + { + "left": "issue:43317", + "right": "issue:43482", + "accept": false, + "reason": "Dequantized model offload loading failure and Qwen2.5-GGUF loading failure are separate load paths and not the same bug." + }, + { + "left": "issue:44987", + "right": "issue:44991", + "accept": false, + "reason": "Both are version-gated loading failures, but they affect different models and no shared concrete defect is evident." + }, + { + "left": "issue:45084", + "right": "issue:45507", + "accept": false, + "reason": "Template compilation error and GraniteMoEHybrid invalid method calls are unrelated." + }, + { + "left": "issue:43262", + "right": "issue:43377", + "accept": false, + "reason": "Audio chat-template sampling-rate default and MIMI batched-vs-single padding mask issue are different audio bugs." + }, + { + "left": "issue:43089", + "right": "issue:45987", + "accept": false, + "reason": "Generation performance overhead and StaticCache seq-length return type are unrelated." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 17, + "review_unit_index": 9, + "review_unit_count": 50, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-428", + "nodes": [ + "issue:33666", + "issue:36032", + "issue:40444", + "issue:40990", + "issue:41553", + "issue:41720", + "issue:42915", + "issue:43296", + "issue:43322", + "issue:43377", + "issue:43404", + "issue:43425", + "issue:43531", + "issue:43540", + "issue:43582", + "issue:43611", + "issue:43632", + "issue:43645", + "issue:43646", + "issue:43742", + "issue:43792", + "issue:43819", + "issue:43994", + "issue:44062", + "issue:44079", + "issue:44220", + "issue:44265", + "issue:44291", + "issue:44403", + "issue:44451", + "issue:44466", + "issue:44479", + "issue:44534", + "issue:44589", + "issue:44610", + "issue:44683", + "issue:44756", + "issue:44811", + "issue:44877", + "issue:44987", + "issue:45020", + "issue:45081", + "issue:45084", + "issue:45127", + "issue:45216", + "issue:45325", + "issue:45357", + "issue:45663" + ], + "soft_pairs": [ + "issue:43645|issue:44534", + "issue:43646|issue:44534", + "issue:43322|issue:43540", + "issue:43404|issue:44466", + "issue:44403|issue:45081", + "issue:43994|issue:44610", + "issue:42915|issue:45084", + "issue:43377|issue:43819", + "issue:44265|issue:44610", + "issue:44451|issue:45081", + "issue:44987|issue:45081", + "issue:44220|issue:44683", + "issue:42915|issue:44291", + "issue:45216|issue:45357", + "issue:43632|issue:44756", + "issue:44062|issue:44987", + "issue:36032|issue:43742", + "issue:44479|issue:45325", + "issue:42915|issue:44589", + "issue:43404|issue:45127", + "issue:43296|issue:43531", + "issue:40444|issue:45325", + "issue:36032|issue:41720", + "issue:43611|issue:44534", + "issue:43531|issue:44877", + "issue:43425|issue:44403", + "issue:41553|issue:43582", + "issue:40990|issue:44811", + "issue:43632|issue:44534", + "issue:44079|issue:45663", + "issue:43792|issue:44589", + "issue:33666|issue:40444", + "issue:44683|issue:45020" + ], + "prepared_review_unit_hash": "1fe41e250d77384c381f5151ded2d51d50763e5ba279e284180c8c212da783ce", + "component_budget": { + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 33, + "serialized_chars": 37467, + "estimated_input_tokens": 9367, + "estimated_eval_tokens": 18990 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This set is mostly heterogeneous regression reports across different models and subsystems. I only see one plausible duplicate pair around Qwen3.5 save_pretrained checkpoint corruption; the rest are too far apart to safely merge.", + "confidence": 0.72, + "canonical_issue_reason": "issue:45216 is the broader, earlier report about Qwen3.5 save_pretrained producing an incorrect checkpoint; issue:45357 reads like a narrower follow-up focused on the visual encoder keys, so 45216 is the better canonical issue for that pair.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45216 is the strongest single issue to anchor the only likely duplicate cluster because it states the regression at the broader save_pretrained level and has the clearest canonical scope.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43645", + "right": "issue:44534", + "accept": false, + "reason": "Both mention Transformers v5 regressions, but one is about custom model init in notebooks and the other about non-persistent buffers being serialized incorrectly; different concrete bugs." + }, + { + "left": "issue:43646", + "right": "issue:44534", + "accept": false, + "reason": "Custom model initialization failure and non-persistent buffer junk are separate v5 issues with different failure modes and code paths." + }, + { + "left": "issue:43322", + "right": "issue:43540", + "accept": false, + "reason": "Both are multimodal loading/processing failures, but Llava-Next segfault and Qwen3OmniMoe video ValueError are not the same underlying bug." + }, + { + "left": "issue:43404", + "right": "issue:44466", + "accept": false, + "reason": "Both involve tied lm_head weights, but one is a model-specific tie failure and the other is a broader save/serialization inconsistency; not the same concrete bug." + }, + { + "left": "issue:44403", + "right": "issue:45081", + "accept": false, + "reason": "Unnecessary logging noise and a Mistral tokenizer crash are unrelated." + }, + { + "left": "issue:43994", + "right": "issue:44610", + "accept": false, + "reason": "Different models and different defects: bad SigLIP2 outputs versus an OmDet-Turbo processor size mismatch." + }, + { + "left": "issue:42915", + "right": "issue:45084", + "accept": false, + "reason": "Qwen3Moe FP8 config failure and a template-node compilation TypeError are unrelated." + }, + { + "left": "issue:43377", + "right": "issue:43819", + "accept": false, + "reason": "Both are audio-model correctness bugs, but MIMI padding-mask behavior and DAC missing-STE behavior are different code paths and fixes." + }, + { + "left": "issue:44265", + "right": "issue:44610", + "accept": false, + "reason": "torch.export failures from torch_compilable_check are unrelated to the OmDet-Turbo image-size regression." + }, + { + "left": "issue:44451", + "right": "issue:45081", + "accept": false, + "reason": "Loading ScandiBERT and the Mistral regex patch crash are separate issues." + }, + { + "left": "issue:44987", + "right": "issue:45081", + "accept": false, + "reason": "A failing remote-code model load and a tokenizer regex patch crash are not the same bug." + }, + { + "left": "issue:44220", + "right": "issue:44683", + "accept": false, + "reason": "Audio fbank extraction and compiled flex_attention on new torch versions are unrelated." + }, + { + "left": "issue:42915", + "right": "issue:44291", + "accept": false, + "reason": "Different failures: FP8 config support for Qwen3Moe versus init_empty_weights/_is_hf_initialized argument handling." + }, + { + "left": "issue:45216", + "right": "issue:45357", + "accept": true, + "reason": "Both describe the same Qwen3.5 save_pretrained regression; 45357 looks like a narrower symptom of incorrect visual-encoder key serialization within the broader checkpoint corruption reported in 45216." + }, + { + "left": "issue:43632", + "right": "issue:44756", + "accept": false, + "reason": "_is_hf_initialized breakage and Strix Halo mmap OOM are unrelated." + }, + { + "left": "issue:44062", + "right": "issue:44987", + "accept": false, + "reason": "AddedToken special-argument duplication and a model load failure for physical-intelligence/fast are different tokenizer/model issues." + }, + { + "left": "issue:36032", + "right": "issue:43742", + "accept": false, + "reason": "T5 tokenizer special-token conflict and MobileLLM key error are unrelated loading failures." + }, + { + "left": "issue:44479", + "right": "issue:45325", + "accept": false, + "reason": "Both are Qwen VL-family regressions, but one is a broad video-input regression and the other is a specific still-image rope-index scaling bug; not enough to treat as one issue." + }, + { + "left": "issue:42915", + "right": "issue:44589", + "accept": false, + "reason": "FP8-config failure for Qwen3Moe and missing Float8 storage support are different failures." + }, + { + "left": "issue:43404", + "right": "issue:45127", + "accept": false, + "reason": "Tie-weight serialization in Mistral3 and LoRA-merge collapse with extended vocab are separate tied-weights problems with different triggers." + }, + { + "left": "issue:43296", + "right": "issue:43531", + "accept": false, + "reason": "PaddleOCR-VL load failure and Qwen3-MoE sliding_window behavior are unrelated." + }, + { + "left": "issue:40444", + "right": "issue:45325", + "accept": false, + "reason": "Qwen2.5-VL finetuning with multiple images per prompt and still-image temporal scaling are not the same bug." + }, + { + "left": "issue:36032", + "right": "issue:41720", + "accept": false, + "reason": "Tokenizer load conflict and Qwen3 auto device-mapping cudaErrorAssert are unrelated." + }, + { + "left": "issue:43611", + "right": "issue:44534", + "accept": false, + "reason": "base_model_prefix loading breakage and non-persistent buffer junk are distinct v5 regressions." + }, + { + "left": "issue:43531", + "right": "issue:44877", + "accept": false, + "reason": "Qwen3-MoE sliding_window handling and granite_speech config strictness are unrelated." + }, + { + "left": "issue:43425", + "right": "issue:44403", + "accept": false, + "reason": "Torch 2.10 incompatibility and generic transformer-loading noise are unrelated." + }, + { + "left": "issue:41553", + "right": "issue:43582", + "accept": false, + "reason": "Voxtral tokenizer error messaging and AppleSilicon warmup TypeError are unrelated." + }, + { + "left": "issue:40990", + "right": "issue:44811", + "accept": false, + "reason": "High perplexity on gpt-oss-20b and Whisper batch_decode skip_special_tokens handling are unrelated." + }, + { + "left": "issue:43632", + "right": "issue:44534", + "accept": false, + "reason": "_is_hf_initialized regression and junk-filled non-persistent buffers are different v5 issues." + }, + { + "left": "issue:44079", + "right": "issue:45663", + "accept": false, + "reason": "ModelOutput key assignment and Gemma-4 FSDP2 shared_kv_states behavior are unrelated." + }, + { + "left": "issue:43792", + "right": "issue:44589", + "accept": false, + "reason": "Whisper load failure and missing Float8 storage object are unrelated." + }, + { + "left": "issue:33666", + "right": "issue:40444", + "accept": false, + "reason": "Both are Qwen VL training reports, but multi-GPU training and iterable-dataset multi-image finetuning are different problems." + }, + { + "left": "issue:44683", + "right": "issue:45020", + "accept": false, + "reason": "Compiled flex_attention on newer torch and remote_code breakage are unrelated." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet and stays conservative. The single accepted duplicate pair (issue:45216 / issue:45357) is well-supported by both titles, and the rest of the soft-edge rejections avoid overclaiming sameness across clearly different bugs and subsystems." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 17, + "review_unit_index": 10, + "review_unit_count": 50, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-428", + "nodes": [ + "issue:35707", + "issue:40444", + "issue:41720", + "issue:42175", + "issue:42915", + "issue:43296", + "issue:43317", + "issue:43377", + "issue:43381", + "issue:43421", + "issue:43454", + "issue:43482", + "issue:43525", + "issue:43531", + "issue:43611", + "issue:43632", + "issue:43644", + "issue:43646", + "issue:43749", + "issue:43792", + "issue:43818", + "issue:43856", + "issue:43957", + "issue:43994", + "issue:44079", + "issue:44265", + "issue:44291", + "issue:44403", + "issue:44462", + "issue:44466", + "issue:44534", + "issue:44568", + "issue:44589", + "issue:44829", + "issue:44898", + "issue:44991", + "issue:45072", + "issue:45081", + "issue:45325", + "issue:45357", + "issue:45381", + "issue:45479", + "issue:45491", + "issue:45538", + "issue:45542", + "issue:45812", + "issue:45834", + "issue:45923" + ], + "soft_pairs": [ + "issue:43421|issue:44568", + "issue:44403|issue:44991", + "issue:35707|issue:40444", + "issue:43381|issue:45923", + "issue:44462|issue:45812", + "issue:42175|issue:45542", + "issue:43317|issue:43856", + "issue:44829|issue:45479", + "issue:43482|issue:45538", + "issue:44291|issue:44589", + "issue:43482|issue:43525", + "issue:44079|issue:44534", + "issue:43818|issue:45357", + "issue:43377|issue:45491", + "issue:41720|issue:45081", + "issue:43531|issue:44291", + "issue:40444|issue:45381", + "issue:44265|issue:45072", + "issue:43644|issue:44079", + "issue:43749|issue:44079", + "issue:45325|issue:45381", + "issue:42175|issue:43296", + "issue:43632|issue:45072", + "issue:43611|issue:43646", + "issue:43792|issue:43957", + "issue:43994|issue:44898", + "issue:45325|issue:45834", + "issue:43454|issue:44466", + "issue:42915|issue:43482" + ], + "prepared_review_unit_hash": "712e408dc75365ed2629d8556d8ff1e2010840ab33c68351df634787951ae548", + "component_budget": { + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 29, + "serialized_chars": 36687, + "estimated_input_tokens": 9172, + "estimated_eval_tokens": 18600 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "Mostly a grab bag of unrelated Transformers issues. The only plausible duplicate pair is the v5 custom model initialization/loading regression around `base_model_prefix`; the rest differ by model, subsystem, or failure mode. No PRs are present.", + "confidence": 0.43, + "canonical_issue_reason": "issue:43646 is the best canonical issue: it is the broadest and most central v5 model-initialization regression, and it can naturally subsume the more specific `base_model_prefix` loading failure in issue:43611.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43646 has the widest scope among the candidates and best represents the underlying v5 custom-model init/load breakage.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43421", + "right": "issue:44568", + "accept": false, + "reason": "Both are tokenizer-related, but one is about runtime post-processor updates and the other about `add_special_tokens`/BOS-EOS behavior for a specific tokenizer; not the same concrete bug." + }, + { + "left": "issue:44403", + "right": "issue:44991", + "accept": false, + "reason": "One is generic loading noise, the other is a tokenizer load failure for a specific model; different symptoms and code paths." + }, + { + "left": "issue:35707", + "right": "issue:40444", + "accept": false, + "reason": "Different tasks entirely: generation with `inputs_embeds/past_key_values` versus Qwen2.5-VL finetuning with multiple images." + }, + { + "left": "issue:43381", + "right": "issue:45923", + "accept": false, + "reason": "Gradient-checkpointing-in-eval and missing `supports_gradient_checkpointing` on a trust-remote-code model are related only broadly, not the same bug." + }, + { + "left": "issue:44462", + "right": "issue:45812", + "accept": false, + "reason": "Both involve AutoTokenizer, but one ignores `tokenizer.json` while the other returns wrong token IDs for Granite models; different failures." + }, + { + "left": "issue:42175", + "right": "issue:45542", + "accept": false, + "reason": "Both mention TensorFlow backend packaging, but one is an install-extra/dependency issue and the other is a runtime undefined-backend error with tensorboard; not clearly the same fix." + }, + { + "left": "issue:43317", + "right": "issue:43856", + "accept": false, + "reason": "One is device-map/offload loading failure; the other is MoE training memory inefficiency. Different problem class." + }, + { + "left": "issue:44829", + "right": "issue:45479", + "accept": false, + "reason": "Different sequence-classification bugs: flash-attention-3 training degeneration versus incorrect loss for `num_labels=1`." + }, + { + "left": "issue:43482", + "right": "issue:45538", + "accept": false, + "reason": "Qwen2.5-GGUF loading on v5 versus CLIPTokenizer `model_max_length` constant; unrelated." + }, + { + "left": "issue:44291", + "right": "issue:44589", + "accept": false, + "reason": "Both are v5 initialization-related, but one is an unexpected `_is_hf_initialized` argument and the other is a missing Float8 storage type; different failures." + }, + { + "left": "issue:43482", + "right": "issue:43525", + "accept": false, + "reason": "Both are model-loading issues, but model-specific config/tokenization problems with different root causes." + }, + { + "left": "issue:44079", + "right": "issue:44534", + "accept": false, + "reason": "ModelOutput key assignment and non-persistent buffer junk are unrelated runtime/state bugs." + }, + { + "left": "issue:43818", + "right": "issue:45357", + "accept": false, + "reason": "Video-LLaVA tower/weight-sharing bug versus Qwen3.5 visual encoder save-pretrained regression; different models and code paths." + }, + { + "left": "issue:43377", + "right": "issue:45491", + "accept": false, + "reason": "Both are batching/padding issues, but they affect different models and different mechanisms." + }, + { + "left": "issue:41720", + "right": "issue:45081", + "accept": false, + "reason": "CUDA assert during Qwen3 auto device mapping is unrelated to the Mistral tokenizer regex patch crash." + }, + { + "left": "issue:43531", + "right": "issue:44291", + "accept": false, + "reason": "Sliding-window handling in Qwen3-MoE is a different bug from `_is_hf_initialized`/`init_empty_weights` initialization issues." + }, + { + "left": "issue:40444", + "right": "issue:45381", + "accept": false, + "reason": "Multimodal finetuning with multiple images is unrelated to Qwen2.5-VL vision position IDs for video input." + }, + { + "left": "issue:44265", + "right": "issue:45072", + "accept": false, + "reason": "`torch.export`/`torch_compilable_check` failure and dtype mismatch in inference are unrelated." + }, + { + "left": "issue:43644", + "right": "issue:44079", + "accept": false, + "reason": "Non-persistent buffer initialization bug versus ModelOutput key bookkeeping bug; different components." + }, + { + "left": "issue:43749", + "right": "issue:44079", + "accept": false, + "reason": "FSDP CPU RAM efficient loading failure is unrelated to ModelOutput key handling." + }, + { + "left": "issue:45325", + "right": "issue:45381", + "accept": false, + "reason": "Both are Qwen2.5-VL positional-id bugs, but one is still-image temporal scaling and the other is video-input vision positions; not enough to treat as the same concrete bug." + }, + { + "left": "issue:42175", + "right": "issue:43296", + "accept": false, + "reason": "TensorFlow backend packaging versus loading PaddleOCR-VL in vLLM 0.11.0; unrelated." + }, + { + "left": "issue:43632", + "right": "issue:45072", + "accept": false, + "reason": "`_is_hf_initialized` flag breakage is unrelated to SwitchTransformers/TimmWrapperModel bfloat16 dtype mismatches." + }, + { + "left": "issue:43611", + "right": "issue:43646", + "accept": true, + "reason": "Both point to the same Transformers v5 custom-model initialization/loading regression; `base_model_prefix` loading is a specific manifestation of the broader custom init breakage." + }, + { + "left": "issue:43792", + "right": "issue:43957", + "accept": false, + "reason": "Whisper run failure and meta-device loading failure are different model-loading problems." + }, + { + "left": "issue:43994", + "right": "issue:44898", + "accept": false, + "reason": "Different model families and different failure modes: SigLIP nonsensical outputs versus Perceiver resolution/interpolation failure." + }, + { + "left": "issue:45325", + "right": "issue:45834", + "accept": false, + "reason": "Both are vision-language issues, but one is a positional-index regression and the other is long OCR input index errors; not the same bug." + }, + { + "left": "issue:43454", + "right": "issue:44466", + "accept": false, + "reason": "AyaVision weight tying bug versus inconsistent lm_head serialization; related only by weights, not the same issue." + }, + { + "left": "issue:42915", + "right": "issue:43482", + "accept": false, + "reason": "Qwen3MoE FP8 configuration failure and Qwen2.5-GGUF loading failure are unrelated." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet and stays conservative overall. The only accepted soft pair, issue:43611 \u2194 issue:43646, is a plausible same v5 init/load regression by title, and the rest are reasonably rejected as different models, subsystems, or failure modes. No PRs are present." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 17, + "review_unit_index": 11, + "review_unit_count": 50, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-428", + "nodes": [ + "issue:34634", + "issue:35707", + "issue:39692", + "issue:40444", + "issue:40990", + "issue:41720", + "issue:42947", + "issue:43257", + "issue:43262", + "issue:43296", + "issue:43317", + "issue:43322", + "issue:43377", + "issue:43454", + "issue:43526", + "issue:43550", + "issue:43572", + "issue:43606", + "issue:43701", + "issue:43720", + "issue:43746", + "issue:43819", + "issue:43873", + "issue:43881", + "issue:43883", + "issue:44186", + "issue:44265", + "issue:44291", + "issue:44387", + "issue:44479", + "issue:44488", + "issue:44568", + "issue:44610", + "issue:44617", + "issue:44877", + "issue:44898", + "issue:44991", + "issue:45072", + "issue:45325", + "issue:45584", + "issue:45701", + "issue:45759", + "issue:45800", + "issue:45823", + "issue:45920", + "issue:45923", + "issue:46050" + ], + "soft_pairs": [ + "issue:43296|issue:45800", + "issue:44387|issue:44617", + "issue:43701|issue:45325", + "issue:44479|issue:45823", + "issue:43881|issue:44488", + "issue:44291|issue:44877", + "issue:43606|issue:44265", + "issue:43262|issue:43746", + "issue:43873|issue:45701", + "issue:34634|issue:35707", + "issue:40990|issue:45584", + "issue:44568|issue:45920", + "issue:35707|issue:41720", + "issue:44186|issue:44265", + "issue:43606|issue:43819", + "issue:41720|issue:45823", + "issue:43454|issue:43550", + "issue:39692|issue:40444", + "issue:43377|issue:44610", + "issue:43317|issue:43873", + "issue:43454|issue:45072", + "issue:43572|issue:45759", + "issue:43257|issue:45923", + "issue:43377|issue:44898", + "issue:40990|issue:42947", + "issue:43322|issue:46050", + "issue:43526|issue:43883", + "issue:41720|issue:43322", + "issue:43881|issue:44991", + "issue:43454|issue:43720" + ], + "prepared_review_unit_hash": "b52f508eec2c595d1794fd24b7abd2360fee7b5303182f014f9f8fbd0fedf037", + "component_budget": { + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 30, + "serialized_chars": 36268, + "estimated_input_tokens": 9067, + "estimated_eval_tokens": 18390 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The cluster is mostly a set of unrelated model-loading, tokenizer, quantization, and device-map bugs. None of the soft pairs look like true duplicates under a strict triage standard; the closest themes are offloading/quantization and tokenizer regressions, but they still appear to be distinct concrete failures. I\u2019d use the broad offloading/quantization report as the cluster representative.", + "confidence": 0.38, + "canonical_issue_reason": "issue:43873 is the broadest and most central issue in this set: it directly covers offloading behavior with quantization, has the most discussion/inbound references, and best represents the nearby offload/device-map/quantization failures without being overly model-specific.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43873 is the best single issue to anchor the cluster because it is open, high-signal, and spans the main recurring theme across the nearby reports: quantized/offloaded model loading and device placement.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43296", + "right": "issue:45800", + "accept": false, + "reason": "Different failures: a specific PaddleOCR-VL load issue vs a broad torch/transformers version incompatibility." + }, + { + "left": "issue:44387", + "right": "issue:44617", + "accept": false, + "reason": "Both mention OOM, but the models and root causes are unrelated; one is int4 quantization memory growth, the other is Sam3Video CUDA OOM." + }, + { + "left": "issue:43701", + "right": "issue:45325", + "accept": false, + "reason": "Checkpoint key mismatch is unrelated to the Qwen2.5-VL rope-index scaling bug." + }, + { + "left": "issue:44479", + "right": "issue:45823", + "accept": false, + "reason": "Distinct model families and symptoms: a video-input regression in Qwen variants vs a Gemma4 device-mismatch forward bug." + }, + { + "left": "issue:43881", + "right": "issue:44488", + "accept": false, + "reason": "Both are load failures, but they affect different models and likely different code paths." + }, + { + "left": "issue:44291", + "right": "issue:44877", + "accept": false, + "reason": "One is an init-empty-weights argument regression; the other is a strict config loading issue. Same area, different bug." + }, + { + "left": "issue:43606", + "right": "issue:44265", + "accept": false, + "reason": "Bark CPU offload device mismatch and torch.export torch_compilable_check failure are unrelated." + }, + { + "left": "issue:43262", + "right": "issue:43746", + "accept": false, + "reason": "Audio chat-template sampling-rate handling is unrelated to PEFT adapter checkpoint loading." + }, + { + "left": "issue:43873", + "right": "issue:45701", + "accept": false, + "reason": "Quantization/offloading bug vs tokenizer behavior change; not the same concrete issue." + }, + { + "left": "issue:34634", + "right": "issue:35707", + "accept": false, + "reason": "BarkProcessor voice_preset and progressive generation with past_key_values are different bugs." + }, + { + "left": "issue:40990", + "right": "issue:45584", + "accept": false, + "reason": "High perplexity on one model is unrelated to Whisper empty-transcription generation failure." + }, + { + "left": "issue:44568", + "right": "issue:45920", + "accept": false, + "reason": "Both are tokenizer regressions, but they describe different failure modes and model sets; too broad to treat as one bug." + }, + { + "left": "issue:35707", + "right": "issue:41720", + "accept": false, + "reason": "Progressive generation with inputs_embeds/past_key_values is unrelated to Qwen3 auto device-map CUDA asserts." + }, + { + "left": "issue:44186", + "right": "issue:44265", + "accept": false, + "reason": "Tokenizer crashes on NER/batched padding are unrelated to torch.export failures." + }, + { + "left": "issue:43606", + "right": "issue:43819", + "accept": false, + "reason": "Both involve audio/codecs, but CPU offload device mismatch and DAC STE mismatch are different problems." + }, + { + "left": "issue:41720", + "right": "issue:45823", + "accept": false, + "reason": "Both are device-map related, but they fail in different models and likely different code paths; not a duplicate." + }, + { + "left": "issue:43454", + "right": "issue:43550", + "accept": false, + "reason": "Weight-tying in AyaVision is unrelated to Bamba SDPA torch.compile failure." + }, + { + "left": "issue:39692", + "right": "issue:40444", + "accept": false, + "reason": "Doc-example/model-processor mismatch in SigLIP2 is unrelated to Qwen2.5-VL multi-image IterableDataset finetuning failure." + }, + { + "left": "issue:43377", + "right": "issue:44610", + "accept": false, + "reason": "Missing padding-mask support in MIMI encoder is unrelated to OmDet-Turbo image-size mismatch." + }, + { + "left": "issue:43317", + "right": "issue:43873", + "accept": false, + "reason": "These are the closest pair, but one is a dequantized-model load failure on gpu+cpu offload while the other is a broader quantization offloading complaint; not enough evidence they are the same concrete bug." + }, + { + "left": "issue:43454", + "right": "issue:45072", + "accept": false, + "reason": "Weight tying and bfloat16 dtype mismatch are different issues." + }, + { + "left": "issue:43572", + "right": "issue:45759", + "accept": false, + "reason": "StableLm config pad-token handling and composite Qwen config unwrapping are different config-path bugs." + }, + { + "left": "issue:43257", + "right": "issue:45923", + "accept": false, + "reason": "MoE weight conversion under accelerate+deepspeed is unrelated to missing supports_gradient_checkpointing on a trust_remote_code variant." + }, + { + "left": "issue:43377", + "right": "issue:44898", + "accept": false, + "reason": "Padding-mask correctness in MIMI encoder is unrelated to Perceiver non-default-resolution failure." + }, + { + "left": "issue:40990", + "right": "issue:42947", + "accept": false, + "reason": "Perplexity anomaly and gradient-checkpointing ineffectiveness are unrelated." + }, + { + "left": "issue:43322", + "right": "issue:46050", + "accept": false, + "reason": "Llava Next load crash and sinq quantized save/load breakage are different bugs." + }, + { + "left": "issue:43526", + "right": "issue:43883", + "accept": false, + "reason": "BeitImageProcessorFast label reduction and Molmo tied-weights attribute error are unrelated." + }, + { + "left": "issue:41720", + "right": "issue:43322", + "accept": false, + "reason": "Both are model-loading/runtime crashes, but the affected models and failure signatures are different." + }, + { + "left": "issue:43881", + "right": "issue:44991", + "accept": false, + "reason": "Model loading failure for glm-4v-9b vs tokenizer loading failure for EMBEDDIA/est-roberta are unrelated." + }, + { + "left": "issue:43454", + "right": "issue:43720", + "accept": false, + "reason": "AyaVision weight tying and BitNet packed-weight unpacking are distinct loading/model-structure bugs." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 17, + "review_unit_index": 12, + "review_unit_count": 50, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-428", + "nodes": [ + "issue:36010", + "issue:36331", + "issue:40990", + "issue:41720", + "issue:42617", + "issue:42915", + "issue:43295", + "issue:43296", + "issue:43317", + "issue:43404", + "issue:43450", + "issue:43526", + "issue:43550", + "issue:43577", + "issue:43606", + "issue:43708", + "issue:43720", + "issue:43749", + "issue:43873", + "issue:43881", + "issue:44008", + "issue:44186", + "issue:44387", + "issue:44423", + "issue:44466", + "issue:44479", + "issue:44534", + "issue:44610", + "issue:44671", + "issue:44743", + "issue:44756", + "issue:44877", + "issue:44898", + "issue:44918", + "issue:44977", + "issue:44991", + "issue:45072", + "issue:45081", + "issue:45161", + "issue:45357", + "issue:45406", + "issue:45632", + "issue:45684", + "issue:45701", + "issue:45704", + "issue:45812", + "issue:45823" + ], + "soft_pairs": [ + "issue:43749|issue:44534", + "issue:43296|issue:44877", + "issue:44466|issue:45357", + "issue:41720|issue:44977", + "issue:42617|issue:42915", + "issue:40990|issue:45701", + "issue:44008|issue:45823", + "issue:44387|issue:45704", + "issue:44186|issue:45072", + "issue:36331|issue:42617", + "issue:43577|issue:43708", + "issue:43404|issue:45357", + "issue:43873|issue:44756", + "issue:36331|issue:42915", + "issue:44877|issue:44991", + "issue:43550|issue:43720", + "issue:43450|issue:45701", + "issue:44423|issue:45406", + "issue:45632|issue:45684", + "issue:41720|issue:42617", + "issue:44479|issue:44918", + "issue:44610|issue:44898", + "issue:43550|issue:43606", + "issue:44186|issue:44610", + "issue:42915|issue:43317", + "issue:43295|issue:45701", + "issue:43526|issue:43881", + "issue:41720|issue:45161", + "issue:44671|issue:44743", + "issue:45081|issue:45812", + "issue:36010|issue:42617" + ], + "prepared_review_unit_hash": "f10fb66cea39a10eb12a9e6be0a5bba2f9576ad9de5d3ad2f286b39fd81a89e8", + "component_budget": { + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 31, + "serialized_chars": 36671, + "estimated_input_tokens": 9168, + "estimated_eval_tokens": 18592 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "Mostly a loose theme cluster around Transformers loading/device/offload/tokenization regressions, but the soft pairs describe different concrete bugs or models rather than true duplicates. I would not merge any of the candidate pairs.", + "confidence": 0.82, + "canonical_issue_reason": "Issue #43873 is the broadest representative of the cluster\u2019s shared offloading/loading pain points, but it is still a distinct bug from the others.", + "canonical_pr_reason": "No pull requests are present in this cluster.", + "best_issue_reason": "#43873 is the best single issue anchor because it is the most general and central of the loading/offload-related reports here.", + "best_pr_reason": "No pull requests are present in this cluster.", + "soft_edge_verdicts": [ + { + "left": "issue:43749", + "right": "issue:44534", + "accept": false, + "reason": "Both are memory/loading regressions, but one is FSDP CPU RAM efficient loading and the other is non-persistent buffer initialization; different failure modes and code paths." + }, + { + "left": "issue:43296", + "right": "issue:44877", + "accept": false, + "reason": "One is a model-load failure in vLLM, the other is strict config blocking granite_speech; unrelated root causes." + }, + { + "left": "issue:44466", + "right": "issue:45357", + "accept": false, + "reason": "Both involve saving weights, but one is tied lm_head serialization by device and the other is incorrect visual encoder keys for Qwen3.5." + }, + { + "left": "issue:41720", + "right": "issue:44977", + "accept": false, + "reason": "Different Qwen3 family problems: CUDA assert under auto device mapping versus flash-attention generation failure." + }, + { + "left": "issue:42617", + "right": "issue:42915", + "accept": false, + "reason": "Completely different contexts: 3d_parallel.py execution failure versus FineGrainedFP8Config failure on Qwen3Moe." + }, + { + "left": "issue:40990", + "right": "issue:45701", + "accept": false, + "reason": "Perplexity on a specific model/dataset is not the same bug as tokenizer-version-dependent tokenization changes." + }, + { + "left": "issue:44008", + "right": "issue:45823", + "accept": false, + "reason": "Gemma3n variable-name collision in forward is unrelated to Gemma4 PLE device mismatch." + }, + { + "left": "issue:44387", + "right": "issue:45704", + "accept": false, + "reason": "Both mention memory, but one is int4 quantization OOM in Transformers 5.x and the other is a T5/apex RMSNorm leak." + }, + { + "left": "issue:44186", + "right": "issue:45072", + "accept": false, + "reason": "Tokenizer crashes on batching/padding are unrelated to dtype mismatches in SwitchTransformers/TimmWrapperModel inference." + }, + { + "left": "issue:36331", + "right": "issue:42617", + "accept": false, + "reason": "Trainer compute_loss API break and 3d_parallel.py failure are different bugs with no shared code path." + }, + { + "left": "issue:43577", + "right": "issue:43708", + "accept": false, + "reason": "Blip2 dtype loading is separate from Trainer resume_from_checkpoint step calculation." + }, + { + "left": "issue:43404", + "right": "issue:45357", + "accept": false, + "reason": "Both touch weights/serialization, but one is missing lm_head tying in Mistral3 and the other is incorrect visual encoder keys in Qwen3.5." + }, + { + "left": "issue:43873", + "right": "issue:44756", + "accept": false, + "reason": "Both are offloading-related, but one is quantization/offload behavior and the other is mmap OOM on Strix Halo; not the same bug." + }, + { + "left": "issue:36331", + "right": "issue:42915", + "accept": false, + "reason": "A custom trainer signature regression is unrelated to a Qwen3Moe FP8 load failure." + }, + { + "left": "issue:44877", + "right": "issue:44991", + "accept": false, + "reason": "Strict config loading for granite_speech is unrelated to tokenizer loading failure for EMBEDDIA/est-roberta." + }, + { + "left": "issue:43550", + "right": "issue:43720", + "accept": false, + "reason": "Both are model-loading/runtime issues, but one is torch.compile SDPA on Bamba and the other is accelerate loading of BitNet packed weights." + }, + { + "left": "issue:43450", + "right": "issue:45701", + "accept": false, + "reason": "Incorrect video batch shapes are unrelated to tokenization changing across versions." + }, + { + "left": "issue:44423", + "right": "issue:45406", + "accept": false, + "reason": "Both affect serve paths, but one crashes on multimodal continuous batching and the other lacks Gemma4Processor._tokenizer." + }, + { + "left": "issue:45632", + "right": "issue:45684", + "accept": false, + "reason": "One is a cache-path collision for local models, the other is permission propagation in save_pretrained; distinct filesystem bugs." + }, + { + "left": "issue:41720", + "right": "issue:45161", + "accept": false, + "reason": "Both concern Qwen/GPT-OSS parallelism, but one is an auto device-map CUDA assert and the other is TP not working for GPT-OSS MoE." + }, + { + "left": "issue:44479", + "right": "issue:44918", + "accept": false, + "reason": "Qwen video-input regression and TRL SFT embedding unpacking failure are different code paths and symptoms." + }, + { + "left": "issue:44610", + "right": "issue:44898", + "accept": false, + "reason": "Different vision preprocessing bugs: OmDet-Turbo produces the wrong resolution, while Perceiver fails on non-default resolution despite interpolate_pos_encoding." + }, + { + "left": "issue:43550", + "right": "issue:43606", + "accept": false, + "reason": "Both involve device/compile behavior, but Bamba SDPA compile failure is not the same as suno/bark-small CPU-offload device mismatch." + }, + { + "left": "issue:44186", + "right": "issue:44610", + "accept": false, + "reason": "Tokenizer batching/truncation crash and image processor resolution mismatch are unrelated." + }, + { + "left": "issue:42915", + "right": "issue:43317", + "accept": false, + "reason": "Qwen3Moe FineGrainedFP8 loading and dequantized-model CPU/GPU offload failure are distinct bugs." + }, + { + "left": "issue:43295", + "right": "issue:45701", + "accept": false, + "reason": "Processor.tokenizer regression is not the same as broader tokenization changes across versions." + }, + { + "left": "issue:43526", + "right": "issue:43881", + "accept": false, + "reason": "A BeitImageProcessorFast label reduction bug is unrelated to glm-4v-9b loading failure." + }, + { + "left": "issue:44671", + "right": "issue:44743", + "accept": false, + "reason": "CamemBERT masked LM prediction errors and Qwen3.5 recurrent-state reset are different model-specific issues." + }, + { + "left": "issue:45081", + "right": "issue:45812", + "accept": false, + "reason": "Both are tokenizer-related, but one is a Mistral regex patch crash and the other is wrong token IDs for Granite models; different root causes." + }, + { + "left": "issue:36010", + "right": "issue:42617", + "accept": false, + "reason": "GenerationMixin import error is unrelated to running 3d_parallel.py." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 17, + "review_unit_index": 13, + "review_unit_count": 50, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-428", + "nodes": [ + "issue:29159", + "issue:30064", + "issue:33357", + "issue:34567", + "issue:34634", + "issue:35707", + "issue:36032", + "issue:41720", + "issue:42490", + "issue:42915", + "issue:43316", + "issue:43322", + "issue:43329", + "issue:43388", + "issue:43421", + "issue:43531", + "issue:43540", + "issue:43550", + "issue:43572", + "issue:43606", + "issue:43653", + "issue:43720", + "issue:43746", + "issue:43782", + "issue:43873", + "issue:43927", + "issue:44077", + "issue:44186", + "issue:44361", + "issue:44451", + "issue:44492", + "issue:44509", + "issue:44514", + "issue:44545", + "issue:44625", + "issue:44898", + "issue:44913", + "issue:44991", + "issue:45059", + "issue:45072", + "issue:45381", + "issue:45406", + "issue:45464", + "issue:45479", + "issue:45701", + "issue:45704", + "issue:45759", + "issue:45812" + ], + "soft_pairs": [ + "issue:43531|issue:43572", + "issue:43316|issue:44913", + "issue:43540|issue:43782", + "issue:42915|issue:43322", + "issue:43421|issue:45812", + "issue:44186|issue:44361", + "issue:43388|issue:45059", + "issue:29159|issue:36032", + "issue:45406|issue:45464", + "issue:33357|issue:35707", + "issue:30064|issue:34634", + "issue:41720|issue:43746", + "issue:43606|issue:43720", + "issue:43550|issue:43653", + "issue:43421|issue:45701", + "issue:43873|issue:45704", + "issue:44492|issue:44509", + "issue:43606|issue:43653", + "issue:34634|issue:41720", + "issue:43531|issue:45759", + "issue:42490|issue:43329", + "issue:34567|issue:35707", + "issue:44514|issue:44545", + "issue:44625|issue:45479", + "issue:43421|issue:43927", + "issue:44898|issue:45072", + "issue:43782|issue:45381", + "issue:44451|issue:44991", + "issue:44077|issue:44625" + ], + "prepared_review_unit_hash": "83016af477d902585c2d2c6c832137cd9acbae67bda6a240d62bb8ca0653b457", + "component_budget": { + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 29, + "serialized_chars": 36509, + "estimated_input_tokens": 9128, + "estimated_eval_tokens": 18512 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly heterogeneous issue reports, not a single duplicate set. The only clear duplicate pair is the Qwen2_5_VLProcessor.apply_chat_template padding=False crash; the rest only share broad subsystem/model-family similarity.", + "confidence": 0.97, + "canonical_issue_reason": "No single issue cleanly represents the whole cluster; the only true duplicate subcluster is issue:44514/issue:44545, and issue:44545 is the best anchor because it has the stronger discussion/inbound reference signal for the same exact bug.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:44545 is the strongest issue candidate because it exactly matches issue:44514's bug report and has slightly more triage signal, making it the best dedupe target for that subcluster.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43531", + "right": "issue:43572", + "accept": false, + "reason": "Both are model/config generation bugs, but they affect different models and different parameters; not the same underlying defect." + }, + { + "left": "issue:43316", + "right": "issue:44913", + "accept": false, + "reason": "Both concern config behavior, but one is a Gemma3TextConfig API mismatch and the other is rotary_pct persistence on reload; different bug paths." + }, + { + "left": "issue:43540", + "right": "issue:43782", + "accept": false, + "reason": "Different multimodal Qwen bugs: one is video-input processing, the other is from_pretrained weight_only loading." + }, + { + "left": "issue:42915", + "right": "issue:43322", + "accept": false, + "reason": "Unrelated failures in different models and code paths; one is FP8 config loading, the other a Llava Next segfault." + }, + { + "left": "issue:43421", + "right": "issue:45812", + "accept": false, + "reason": "Both are tokenizer-related, but one is runtime post-processor refresh and the other is a silent token-id regression for Granite models." + }, + { + "left": "issue:44186", + "right": "issue:44361", + "accept": false, + "reason": "Both are tokenizer task crashes, but on different tokenizer implementations and different failure modes." + }, + { + "left": "issue:43388", + "right": "issue:45059", + "accept": false, + "reason": "These are unrelated: metric gathering label truncation versus SAM3 text/bounding-box behavior." + }, + { + "left": "issue:29159", + "right": "issue:36032", + "accept": false, + "reason": "Both mention tokenizers, but one is slow/fast inconsistency and the other is a T5 loading AttributeError; different bugs." + }, + { + "left": "issue:45406", + "right": "issue:45464", + "accept": false, + "reason": "Serving crashes in different places: Gemma4Processor missing _tokenizer versus Qwen3.5 streaming inference failure." + }, + { + "left": "issue:33357", + "right": "issue:35707", + "accept": false, + "reason": "Different failures entirely: MacOS CLIP bus error versus generation with inputs_embeds/past_key_values." + }, + { + "left": "issue:30064", + "right": "issue:34634", + "accept": false, + "reason": "Different processors and symptoms; image segmentation-map handling is unrelated to Bark voice_preset handling." + }, + { + "left": "issue:41720", + "right": "issue:43746", + "accept": false, + "reason": "Different model/loading problems with no shared concrete code path." + }, + { + "left": "issue:43606", + "right": "issue:43720", + "accept": false, + "reason": "Both involve accelerate/CI loading, but one is CPU offload device mismatch and the other is packed-weight unpacking." + }, + { + "left": "issue:43550", + "right": "issue:43653", + "accept": false, + "reason": "Different subsystems: torch.compile/SDPA versus tokenizer special-token registration." + }, + { + "left": "issue:43421", + "right": "issue:45701", + "accept": false, + "reason": "Similar broad tokenizer theme, but 45701 is a generic version regression while 43421 is a runtime post-processor update bug." + }, + { + "left": "issue:43873", + "right": "issue:45704", + "accept": false, + "reason": "Unrelated issues: quantized offloading versus a T5 RMSNorm memory leak." + }, + { + "left": "issue:44492", + "right": "issue:44509", + "accept": false, + "reason": "Both are documentation-ish, but they concern different content and different fixes." + }, + { + "left": "issue:43606", + "right": "issue:43653", + "accept": false, + "reason": "Distinct bugs in different models/components; no shared underlying defect." + }, + { + "left": "issue:34634", + "right": "issue:41720", + "accept": false, + "reason": "Different model/runtime failures with no common code path." + }, + { + "left": "issue:43531", + "right": "issue:45759", + "accept": false, + "reason": "Both touch Qwen3 model config behavior, but sliding_window handling and text_config unwrapping are separate bugs." + }, + { + "left": "issue:42490", + "right": "issue:43329", + "accept": false, + "reason": "Completely different issues: trailing-slash local model import warning versus multimodal token counting in video branch." + }, + { + "left": "issue:34567", + "right": "issue:35707", + "accept": false, + "reason": "TrainerState token counting and progressive generation are unrelated." + }, + { + "left": "issue:44514", + "right": "issue:44545", + "accept": true, + "reason": "Exact same crash description and same code path in Qwen2_5_VLProcessor.apply_chat_template with batched input and padding=False." + }, + { + "left": "issue:44625", + "right": "issue:45479", + "accept": false, + "reason": "Config propagation bug versus a cross-model loss-degeneration problem; not the same bug." + }, + { + "left": "issue:43421", + "right": "issue:43927", + "accept": false, + "reason": "Both involve special-token/config persistence, but one is runtime post-processor syncing and the other is saved custom token IDs being lost." + }, + { + "left": "issue:44898", + "right": "issue:45072", + "accept": false, + "reason": "Different model failures: interpolation on Perceiver image classification versus bfloat16 dtype mismatches." + }, + { + "left": "issue:43782", + "right": "issue:45381", + "accept": false, + "reason": "Both are Qwen VL-related, but one is a load-time weight_only error and the other is wrong video vision_position_ids." + }, + { + "left": "issue:44451", + "right": "issue:44991", + "accept": false, + "reason": "Different tokenizer-loading regressions affecting different models; not the same underlying issue." + }, + { + "left": "issue:44077", + "right": "issue:44625", + "accept": false, + "reason": "PatchTSMixer post_init validation and Qwen3.5 num_labels propagation are unrelated." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet and stays conservative. The only clear duplicate pair in the evidence is issue:44514/issue:44545, and the choice of issue:44545 as the anchor is supported by its slightly higher inbound/discussion signal. The soft-edge rejections are also cautious and generally consistent with the titles and descriptions provided." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 17, + "review_unit_index": 14, + "review_unit_count": 50, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-428", + "nodes": [ + "issue:30064", + "issue:36010", + "issue:39692", + "issue:40444", + "issue:41720", + "issue:42886", + "issue:43381", + "issue:43421", + "issue:43531", + "issue:43540", + "issue:43596", + "issue:43618", + "issue:43645", + "issue:43653", + "issue:43697", + "issue:43720", + "issue:43746", + "issue:43819", + "issue:43844", + "issue:43873", + "issue:43994", + "issue:44008", + "issue:44062", + "issue:44186", + "issue:44220", + "issue:44368", + "issue:44451", + "issue:44589", + "issue:44610", + "issue:44661", + "issue:44671", + "issue:44843", + "issue:44918", + "issue:44936", + "issue:44987", + "issue:45042", + "issue:45081", + "issue:45356", + "issue:45357", + "issue:45588", + "issue:45663", + "issue:45704", + "issue:45759", + "issue:45859", + "issue:45907", + "issue:45923", + "issue:46050" + ], + "soft_pairs": [ + "issue:45759|issue:45859", + "issue:43531|issue:44589", + "issue:44368|issue:45357", + "issue:44987|issue:45042", + "issue:44062|issue:44451", + "issue:45588|issue:45759", + "issue:39692|issue:44610", + "issue:43746|issue:46050", + "issue:45704|issue:45923", + "issue:44918|issue:44936", + "issue:43421|issue:43653", + "issue:45663|issue:45923", + "issue:42886|issue:44843", + "issue:43381|issue:43844", + "issue:43421|issue:44186", + "issue:45704|issue:45859", + "issue:43994|issue:44671", + "issue:41720|issue:45704", + "issue:43531|issue:43645", + "issue:43746|issue:43873", + "issue:45081|issue:45356", + "issue:39692|issue:43873", + "issue:40444|issue:43746", + "issue:43596|issue:44220", + "issue:45356|issue:45907", + "issue:30064|issue:36010", + "issue:43720|issue:43819", + "issue:43653|issue:43720", + "issue:44220|issue:44661", + "issue:43540|issue:44008", + "issue:43618|issue:43697" + ], + "prepared_review_unit_hash": "9ff48fa0394695b65daf12c29ed7a5c0554e5f279194e3a83bbb3aa3f2c47593", + "component_budget": { + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 31, + "serialized_chars": 36808, + "estimated_input_tokens": 9202, + "estimated_eval_tokens": 18660 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly heterogeneous: the soft pairs share superficial model/tokenizer/offload wording, but they describe different bugs and code paths, so none should be deduplicated as the same underlying change or regression.", + "confidence": 0.94, + "canonical_issue_reason": "Issue 43873 is the closest representative of the cluster because it is the most discussed/high-signal item and sits in the broad quantization/offloading theme that recurs in a few nearby issues.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 43873 is the best single issue to anchor the cluster: it has the strongest engagement and is the most broadly framed, actionable bug among otherwise unrelated reports.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45759", + "right": "issue:45859", + "accept": false, + "reason": "Same Qwen family, but different failures: composite config unwrapping vs CPU-offload rotary embedding behavior." + }, + { + "left": "issue:43531", + "right": "issue:44589", + "accept": false, + "reason": "Qwen3-MoE sliding-window handling and missing Float8 storage are unrelated bugs." + }, + { + "left": "issue:44368", + "right": "issue:45357", + "accept": false, + "reason": "One is a tie_word_embeddings warning; the other is a save_pretrained key regression for visual encoder weights." + }, + { + "left": "issue:44987", + "right": "issue:45042", + "accept": false, + "reason": "Different subsystems: loading a model repo versus PIL image processors incorrectly requiring torchvision." + }, + { + "left": "issue:44062", + "right": "issue:44451", + "accept": false, + "reason": "AddedToken constructor duplication and inability to load ScandiBERT are not the same issue." + }, + { + "left": "issue:45588", + "right": "issue:45759", + "accept": false, + "reason": "flash_attention s_aux handling and Qwen multimodal text_config unwrapping are separate code paths." + }, + { + "left": "issue:39692", + "right": "issue:44610", + "accept": false, + "reason": "Both are processor/model issues, but one is a SigLIP2 docs/model mismatch and the other is an OmDet-Turbo image-size mismatch." + }, + { + "left": "issue:43746", + "right": "issue:46050", + "accept": false, + "reason": "Local checkpoint loading with PEFT adapters and quantized sinq save/load corruption are different persistence bugs." + }, + { + "left": "issue:45704", + "right": "issue:45923", + "accept": false, + "reason": "Apex RMSNorm memory leak and a missing supports_gradient_checkpointing flag are unrelated." + }, + { + "left": "issue:44918", + "right": "issue:44936", + "accept": false, + "reason": "Trainer embedding unpacking and evaluate-after-train failure do not share the same root cause." + }, + { + "left": "issue:43421", + "right": "issue:43653", + "accept": false, + "reason": "Both touch special tokens, but runtime post-processor refresh and BigBird mask-token registration are distinct problems." + }, + { + "left": "issue:45663", + "right": "issue:45923", + "accept": false, + "reason": "Gemma4 FSDP2 shared_kv_states KeyError is unrelated to a missing gradient-checkpointing capability flag." + }, + { + "left": "issue:42886", + "right": "issue:44843", + "accept": false, + "reason": "Both involve offline/tokenizer behavior, but one is generic cache loading and the other is a specific _patch_mistral_regex model_info call." + }, + { + "left": "issue:43381", + "right": "issue:43844", + "accept": false, + "reason": "Eval-mode gradient checkpointing and ZeRO-3 gradient growth are different training failures." + }, + { + "left": "issue:43421", + "right": "issue:44186", + "accept": false, + "reason": "Special-token post-processor updates and LayoutLMv2 NER/padding crashes are separate tokenizer bugs." + }, + { + "left": "issue:45704", + "right": "issue:45859", + "accept": false, + "reason": "T5 RMSNorm memory leak and Qwen offload incompatibility are unrelated." + }, + { + "left": "issue:43994", + "right": "issue:44671", + "accept": false, + "reason": "Different models and symptoms: SigLIP2 nonsensical outputs versus CamemBERT masked-LM regressions." + }, + { + "left": "issue:41720", + "right": "issue:45704", + "accept": false, + "reason": "Qwen3 device-mapping CUDA asserts and a T5 RMSNorm memory leak are not the same bug." + }, + { + "left": "issue:43531", + "right": "issue:43645", + "accept": false, + "reason": "Qwen3-MoE sliding-window behavior and Jupyter custom-model initialization are unrelated." + }, + { + "left": "issue:43746", + "right": "issue:43873", + "accept": false, + "reason": "Checkpoint loading with PEFT adapters and quantization/offloading behavior are different code paths." + }, + { + "left": "issue:45081", + "right": "issue:45356", + "accept": false, + "reason": "Both are tokenizer regressions, but one is _patch_mistral_regex crash behavior and the other is Kimi-K2.5 codec handling/warning issues." + }, + { + "left": "issue:39692", + "right": "issue:43873", + "accept": false, + "reason": "SigLIP2 example errors and quantization offloading problems are unrelated." + }, + { + "left": "issue:40444", + "right": "issue:43746", + "accept": false, + "reason": "Qwen2.5-VL multi-image iterable-dataset finetuning and local checkpoint loading with PEFT are different bugs." + }, + { + "left": "issue:43596", + "right": "issue:44220", + "accept": false, + "reason": "ZeRO-3/BertModel init failure and _torch_extract_fbank_features are unrelated." + }, + { + "left": "issue:45356", + "right": "issue:45907", + "accept": false, + "reason": "Kimi tokenizer regression and list_repo_files revision forwarding are not the same issue." + }, + { + "left": "issue:30064", + "right": "issue:36010", + "accept": false, + "reason": "Image processor void-segmentation-map failure and GenerationMixin import error are unrelated." + }, + { + "left": "issue:43720", + "right": "issue:43819", + "accept": false, + "reason": "BitNet packed-weight loading and DAC forward/from_latents mismatch are different model bugs." + }, + { + "left": "issue:43653", + "right": "issue:43720", + "accept": false, + "reason": "BigBird special-token registration and BitNet packed-weight loading are unrelated." + }, + { + "left": "issue:44220", + "right": "issue:44661", + "accept": false, + "reason": "Audio fbank extraction and add-new-model-like/tokenizer-mapping behavior are different problems." + }, + { + "left": "issue:43540", + "right": "issue:44008", + "accept": false, + "reason": "Video-input processing in Qwen3OmniMoe and Gemma3n audio_mel_mask variable collision are separate multimodal bugs." + }, + { + "left": "issue:43618", + "right": "issue:43697", + "accept": false, + "reason": "Missing CLIPOutput attentions and RTDetrV2 output drift in v5 are distinct regressions." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 17, + "review_unit_index": 15, + "review_unit_count": 50, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-428", + "nodes": [ + "issue:30064", + "issue:36331", + "issue:39692", + "issue:41720", + "issue:43295", + "issue:43322", + "issue:43335", + "issue:43388", + "issue:43421", + "issue:43531", + "issue:43540", + "issue:43550", + "issue:43577", + "issue:43643", + "issue:43644", + "issue:43653", + "issue:43716", + "issue:43749", + "issue:43756", + "issue:43782", + "issue:43818", + "issue:43819", + "issue:43950", + "issue:44008", + "issue:44062", + "issue:44186", + "issue:44442", + "issue:44448", + "issue:44488", + "issue:44671", + "issue:44811", + "issue:44821", + "issue:44843", + "issue:44877", + "issue:44898", + "issue:44977", + "issue:44991", + "issue:45072", + "issue:45081", + "issue:45084", + "issue:45183", + "issue:45381", + "issue:45479", + "issue:45584", + "issue:45685", + "issue:45834", + "issue:45987", + "issue:46032" + ], + "soft_pairs": [ + "issue:45183|issue:45479", + "issue:43550|issue:45072", + "issue:43388|issue:46032", + "issue:43295|issue:45834", + "issue:43550|issue:43819", + "issue:43716|issue:44008", + "issue:44062|issue:44488", + "issue:43322|issue:45685", + "issue:43335|issue:43756", + "issue:44186|issue:44898", + "issue:43653|issue:43819", + "issue:43950|issue:45987", + "issue:43531|issue:43643", + "issue:44843|issue:45081", + "issue:43643|issue:44877", + "issue:43540|issue:45381", + "issue:44448|issue:44671", + "issue:30064|issue:36331", + "issue:43421|issue:44811", + "issue:43577|issue:43818", + "issue:43782|issue:44977", + "issue:41720|issue:44991", + "issue:39692|issue:45834", + "issue:45584|issue:45834", + "issue:44821|issue:45084", + "issue:44186|issue:44442", + "issue:43644|issue:43749" + ], + "prepared_review_unit_hash": "00fc4dfb87215b40e11935b4ad638dadfed10602ef525b7e71134ad4a409b294", + "component_budget": { + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 27, + "serialized_chars": 36180, + "estimated_input_tokens": 9045, + "estimated_eval_tokens": 18346 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "All proposed soft edges are false positives: they span different models, subsystems, and failure modes rather than the same underlying bug or change. There is no clear canonical duplicate target in this set.", + "confidence": 0.92, + "canonical_issue_reason": "No single issue is a credible canonical representative; the items are a heterogeneous mix of unrelated tokenizer, multimodal, cache, loading, and dtype bugs.", + "canonical_pr_reason": null, + "best_issue_reason": "No issue stands out as a global duplicate target for the cluster, because the cluster does not collapse to one concrete defect.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45183", + "right": "issue:45479", + "accept": false, + "reason": "Both involve validation/classification behavior, but one is input-argument XOR checking across models and the other is a zero-loss bug for num_labels=1; different defects." + }, + { + "left": "issue:43550", + "right": "issue:45072", + "accept": false, + "reason": "Same general area of inference/runtime errors, but Bamba SDPA + torch.compile and dtype mismatches in SwitchTransformers/TimmWrapperModel are different code paths and failures." + }, + { + "left": "issue:43388", + "right": "issue:46032", + "accept": false, + "reason": "gather_for_metrics label truncation and Mamba2Mixer cache corruption are unrelated subsystems with different symptoms." + }, + { + "left": "issue:43295", + "right": "issue:45834", + "accept": false, + "reason": "Processor.tokenizer regression and Kosmos2.5 long-OCR index error are different multimodal loading/processing bugs." + }, + { + "left": "issue:43550", + "right": "issue:43819", + "accept": false, + "reason": "Bamba torch.compile/SDPA failure and DAC.from_latents mismatch are unrelated model-specific bugs." + }, + { + "left": "issue:43716", + "right": "issue:44008", + "accept": false, + "reason": "Mistral-3 image preprocessor dtype mismatch is not the same as Gemma3n's variable-name collision in forward()." + }, + { + "left": "issue:44062", + "right": "issue:44488", + "accept": false, + "reason": "AddedToken special-argument duplication and a tokenizer load failure for cjvt/sleng-bert are distinct tokenizer issues." + }, + { + "left": "issue:43322", + "right": "issue:45685", + "accept": false, + "reason": "Llava Next loading segfault and MPS histogram on Int in moe are different failure modes in different paths." + }, + { + "left": "issue:43335", + "right": "issue:43756", + "accept": false, + "reason": "SwitchTransformers sparse-layer construction and Smollm3 RoPE-layer dropping are different model/config bugs." + }, + { + "left": "issue:44186", + "right": "issue:44898", + "accept": false, + "reason": "LayoutLMv2Tokenizer NER/padding crash and Perceiver non-default-resolution failure are unrelated." + }, + { + "left": "issue:43653", + "right": "issue:43819", + "accept": false, + "reason": "BigBirdTokenizer special-token registration and DAC.from_latents/forward mismatch do not share a concrete code-path bug." + }, + { + "left": "issue:43950", + "right": "issue:45987", + "accept": false, + "reason": "Non-persistent buffer corruption on from_pretrained and StaticCache.get_seq_length returning a tensor are separate regression classes." + }, + { + "left": "issue:43531", + "right": "issue:43643", + "accept": false, + "reason": "Qwen3-MoE sliding_window behavior and trust_remote_code AutoConfig field loss are unrelated issues." + }, + { + "left": "issue:44843", + "right": "issue:45081", + "accept": false, + "reason": "Both mention Mistral regex patching, but one breaks offline mode via model_info() and the other crashes on a backend_tokenizer attribute assumption; not the same bug." + }, + { + "left": "issue:43643", + "right": "issue:44877", + "accept": false, + "reason": "Missing config fields with trust_remote_code and strict config rejection for granite_speech are different config-loading problems." + }, + { + "left": "issue:43540", + "right": "issue:45381", + "accept": false, + "reason": "Qwen3OmniMoe video processing ValueError and Qwen2.5-VL video vision_position_ids mismatch are different model-specific video bugs." + }, + { + "left": "issue:44448", + "right": "issue:44671", + "accept": false, + "reason": "Pegasus v4/v5 output differences and CamemBERT MLM prediction errors are unrelated model behavior regressions." + }, + { + "left": "issue:30064", + "right": "issue:36331", + "accept": false, + "reason": "Image processor void-segmentation handling and CustomTrainer.compute_loss keyword-arg failure are unrelated." + }, + { + "left": "issue:43421", + "right": "issue:44811", + "accept": false, + "reason": "Runtime special-token post-processor updates and Whisper batch_decode skip_special_tokens handling are different layers and symptoms." + }, + { + "left": "issue:43577", + "right": "issue:43818", + "accept": false, + "reason": "Blip2 dtype persistence and Video-LLaVA missing temporal attention/shared weights are unrelated defects." + }, + { + "left": "issue:43782", + "right": "issue:44977", + "accept": false, + "reason": "Qwen3VL weight_only loading failure and Qwen3.5 flash-attention generation bug are distinct model/runtime issues." + }, + { + "left": "issue:41720", + "right": "issue:44991", + "accept": false, + "reason": "A800 cudaErrorAssert under auto device mapping and tokenizer load failure for est-roberta have no shared underlying code path." + }, + { + "left": "issue:39692", + "right": "issue:45834", + "accept": false, + "reason": "SigLIP2 docs example errors and Kosmos2.5 long OCR index error are unrelated." + }, + { + "left": "issue:45584", + "right": "issue:45834", + "accept": false, + "reason": "Whisper empty-transcription generation failure and Kosmos2.5 OCR indexing are different modalities and bugs." + }, + { + "left": "issue:44821", + "right": "issue:45084", + "accept": false, + "reason": "AutoImageProcessor-from-URL loading and non-template-node compilation errors are unrelated." + }, + { + "left": "issue:44186", + "right": "issue:44442", + "accept": false, + "reason": "LayoutLMv2Tokenizer NER crash and AutoTokenizer loading FastSpeech2ConformerTokenizer are different tokenizer failures." + }, + { + "left": "issue:43644", + "right": "issue:43749", + "accept": false, + "reason": "Non-persistent buffer corruption and broken FSDP CPU RAM efficient loading are separate regressions." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 17, + "review_unit_index": 16, + "review_unit_count": 50, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-428", + "nodes": [ + "issue:28218", + "issue:29159", + "issue:34634", + "issue:36010", + "issue:36331", + "issue:42175", + "issue:43262", + "issue:43526", + "issue:43577", + "issue:43582", + "issue:43606", + "issue:43650", + "issue:43653", + "issue:43697", + "issue:43720", + "issue:43742", + "issue:43749", + "issue:43782", + "issue:43819", + "issue:43825", + "issue:43827", + "issue:43906", + "issue:43994", + "issue:44060", + "issue:44291", + "issue:44368", + "issue:44479", + "issue:44485", + "issue:44509", + "issue:44554", + "issue:44561", + "issue:44610", + "issue:44671", + "issue:44805", + "issue:44857", + "issue:44912", + "issue:44936", + "issue:45020", + "issue:45072", + "issue:45081", + "issue:45084", + "issue:45325", + "issue:45406", + "issue:45538", + "issue:45701", + "issue:45834", + "issue:45987", + "issue:46192" + ], + "soft_pairs": [ + "issue:44561|issue:45020", + "issue:43606|issue:45072", + "issue:44479|issue:45834", + "issue:43825|issue:43994", + "issue:34634|issue:36010", + "issue:43577|issue:43582", + "issue:45987|issue:46192", + "issue:44805|issue:44936", + "issue:44485|issue:44509", + "issue:43825|issue:43827", + "issue:44671|issue:45701", + "issue:28218|issue:29159", + "issue:44610|issue:45072", + "issue:43782|issue:45325", + "issue:43720|issue:45072", + "issue:42175|issue:45081", + "issue:43749|issue:44291", + "issue:45081|issue:45406", + "issue:43582|issue:44857", + "issue:43650|issue:43906", + "issue:43262|issue:43819", + "issue:43526|issue:43697", + "issue:44060|issue:44368", + "issue:34634|issue:36331", + "issue:43653|issue:44912", + "issue:43526|issue:44554", + "issue:43742|issue:45020", + "issue:45084|issue:45538" + ], + "prepared_review_unit_hash": "7b9d8ac8bfa88298c138081ecd3e815c3c6475ef469708c09451cb763070a4c7", + "component_budget": { + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 28, + "serialized_chars": 36305, + "estimated_input_tokens": 9077, + "estimated_eval_tokens": 18410 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This is a noisy grab-bag of unrelated Transformers issues across tokenization, processors, model loading, training, and docs. I don\u2019t see a single underlying bug/change that unifies the cluster, so there\u2019s no strong canonical item.", + "confidence": 0.96, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44561", + "right": "issue:45020", + "accept": false, + "reason": "Both mention remote_code breakage, but one is a specific v5 removal of is_torch_fx_available and the other is a broad report about recent versions; not clearly the same concrete bug." + }, + { + "left": "issue:43606", + "right": "issue:45072", + "accept": false, + "reason": "Both involve dtype/device mismatches, but they affect different models and code paths (CPU offload vs bfloat16 inference)." + }, + { + "left": "issue:44479", + "right": "issue:45834", + "accept": false, + "reason": "Different model families and failures: a video-input regression in Qwen VL models versus a long-OCR index error in Kosmos2.5." + }, + { + "left": "issue:43825", + "right": "issue:43994", + "accept": false, + "reason": "One is an error-message regression around removed translation tasks; the other is a bad AutoModel/pipeline result for SigLIP2. Different bugs." + }, + { + "left": "issue:34634", + "right": "issue:36010", + "accept": false, + "reason": "Completely different failures: BarkProcessor voice_preset handling versus GenerationMixin import errors." + }, + { + "left": "issue:43577", + "right": "issue:43582", + "accept": false, + "reason": "Unrelated issues: BLIP2 dtype propagation versus an Apple Silicon allocator warmup TypeError." + }, + { + "left": "issue:45987", + "right": "issue:46192", + "accept": false, + "reason": "Different API surfaces and symptoms; one is StaticCache seq-length return type, the other is text_kwargs/return_tensors handling." + }, + { + "left": "issue:44805", + "right": "issue:44936", + "accept": false, + "reason": "A tensor mask shape mismatch is not the same bug as trainer.evaluate() failing after train()." + }, + { + "left": "issue:44485", + "right": "issue:44509", + "accept": false, + "reason": "RoPE implementation discussion is unrelated to docs still referencing removed pipeline tasks." + }, + { + "left": "issue:43825", + "right": "issue:43827", + "accept": false, + "reason": "Both are about v5 pipeline removals, but one concerns runtime messaging and the other documentation; not the same concrete issue." + }, + { + "left": "issue:44671", + "right": "issue:45701", + "accept": false, + "reason": "CamemBERT masked-LM regressions and version-dependent tokenization changes are too broad and not clearly the same bug." + }, + { + "left": "issue:28218", + "right": "issue:29159", + "accept": false, + "reason": "Both are tokenizer complaints, but the titles don\u2019t establish the same root cause; one is added-token spacing, the other slow/fast inconsistency." + }, + { + "left": "issue:44610", + "right": "issue:45072", + "accept": false, + "reason": "Different models and failure modes: OmDet-Turbo processor size mismatch versus dtype mismatches in SwitchTransformers/TimmWrapperModel." + }, + { + "left": "issue:43782", + "right": "issue:45325", + "accept": false, + "reason": "One is a Qwen3VL weight_only load error, the other a Qwen2.5-VL rope-index scaling bug; different code paths and symptoms." + }, + { + "left": "issue:43720", + "right": "issue:45072", + "accept": false, + "reason": "BitNet packed-weight loading is unrelated to bfloat16 dtype mismatch issues in other models." + }, + { + "left": "issue:42175", + "right": "issue:45081", + "accept": false, + "reason": "Backend/package install configuration is unrelated to the Mistral tokenizer regex patch crash." + }, + { + "left": "issue:43749", + "right": "issue:44291", + "accept": false, + "reason": "Different loading mechanisms and failures: FSDP CPU RAM efficient loading versus init_empty_weights receiving an unexpected argument." + }, + { + "left": "issue:45081", + "right": "issue:45406", + "accept": false, + "reason": "Tokenizer regex patching and Gemma4Processor missing _tokenizer are different loading/processor bugs." + }, + { + "left": "issue:43582", + "right": "issue:44857", + "accept": false, + "reason": "AppleSilicon warmup TypeError is unrelated to a CUDA/AMP crash in LwDetrImageLoss." + }, + { + "left": "issue:43650", + "right": "issue:43906", + "accept": false, + "reason": "These titles are not about the same bug; one is nonspecific 'ADD THE DATA', the other is an isolated reproduction of another issue." + }, + { + "left": "issue:43262", + "right": "issue:43819", + "accept": false, + "reason": "Audio chat-template sampling-rate defaults and DAC from_latents/STE mismatch are unrelated." + }, + { + "left": "issue:43526", + "right": "issue:43697", + "accept": false, + "reason": "BeitImageProcessorFast reduce_labels behavior and RTDetrV2 output differences are different model-specific bugs." + }, + { + "left": "issue:44060", + "right": "issue:44368", + "accept": false, + "reason": "Both mention Qwen3/Qwen3.5 warnings, but one is a tied-weights warning bug and the other is a separate lora fine-tuning warning; not the same issue." + }, + { + "left": "issue:34634", + "right": "issue:36331", + "accept": false, + "reason": "BarkProcessor voice_preset and CustomTrainer.compute_loss signature mismatch are unrelated." + }, + { + "left": "issue:43653", + "right": "issue:44912", + "accept": false, + "reason": "BigBirdTokenizer special-token decode failure is unrelated to MXFP4 quantization fallback." + }, + { + "left": "issue:43526", + "right": "issue:44554", + "accept": false, + "reason": "A Beit image-processor label reduction bug is unrelated to an MPS attention correctness issue." + }, + { + "left": "issue:43742", + "right": "issue:45020", + "accept": false, + "reason": "Loading facebook/MobileLLM-125M key errors and remote_code breakages are different failure classes." + }, + { + "left": "issue:45084", + "right": "issue:45538", + "accept": false, + "reason": "Template compilation errors are unrelated to CLIPTokenizer model_max_length handling." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 17, + "review_unit_index": 17, + "review_unit_count": 50, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-428", + "nodes": [ + "issue:29159", + "issue:34567", + "issue:36010", + "issue:36331", + "issue:38175", + "issue:39692", + "issue:40990", + "issue:42490", + "issue:43257", + "issue:43262", + "issue:43329", + "issue:43450", + "issue:43452", + "issue:43454", + "issue:43531", + "issue:43550", + "issue:43577", + "issue:43606", + "issue:43638", + "issue:43643", + "issue:43645", + "issue:43696", + "issue:43720", + "issue:43746", + "issue:43749", + "issue:43819", + "issue:43854", + "issue:43927", + "issue:43931", + "issue:44479", + "issue:44568", + "issue:44589", + "issue:44610", + "issue:44756", + "issue:44811", + "issue:44849", + "issue:44912", + "issue:44936", + "issue:44977", + "issue:44991", + "issue:45042", + "issue:45356", + "issue:45459", + "issue:45600", + "issue:45701", + "issue:45759", + "issue:45803", + "issue:45859" + ], + "soft_pairs": [ + "issue:43329|issue:45600", + "issue:43746|issue:44811", + "issue:43452|issue:44991", + "issue:44589|issue:45042", + "issue:43638|issue:44936", + "issue:43550|issue:44912", + "issue:45459|issue:45803", + "issue:43819|issue:43927", + "issue:36010|issue:43645", + "issue:43257|issue:43329", + "issue:43854|issue:44977", + "issue:39692|issue:44912", + "issue:38175|issue:44610", + "issue:43720|issue:44610", + "issue:36010|issue:36331", + "issue:43257|issue:43696", + "issue:43454|issue:43577", + "issue:43262|issue:44811", + "issue:43531|issue:45042", + "issue:42490|issue:43257", + "issue:43329|issue:44479", + "issue:43643|issue:45759", + "issue:43749|issue:44756", + "issue:43329|issue:43450", + "issue:43606|issue:45859", + "issue:34567|issue:40990", + "issue:44568|issue:45356", + "issue:43257|issue:43931", + "issue:29159|issue:34567", + "issue:44849|issue:45701" + ], + "prepared_review_unit_hash": "cbe35caf3d26713b03d5195fcaf6eebe671272bba13e85f8bce8540dfde6dee1", + "component_budget": { + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 30, + "serialized_chars": 36943, + "estimated_input_tokens": 9236, + "estimated_eval_tokens": 18728 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The cluster is highly heterogeneous. The soft pairs mostly connect superficially similar Transformers issues across unrelated areas (tokenizers, loading, training, multimodal/video, offload, quantization), so none look like true duplicates.", + "confidence": 0.89, + "canonical_issue_reason": "issue:43329 is the most concrete, code-path-specific report in the densest soft-similarity pocket (multimodal/video token counting), so it is the best representative if a canonical must be chosen.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43329 is the clearest representative of the only semi-coherent subtheme, but the cluster overall is not a true duplicate set.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43329", + "right": "issue:45600", + "accept": false, + "reason": "Video token-counting undefined-variable bug vs removed config reference breaking CI; different subsystems and failure modes." + }, + { + "left": "issue:43746", + "right": "issue:44811", + "accept": false, + "reason": "PEFT adapter local-checkpoint loading vs Whisper batch_decode skip_special_tokens behavior; unrelated code paths." + }, + { + "left": "issue:43452", + "right": "issue:44991", + "accept": false, + "reason": "gguf_file loader breakage vs tokenizer codec handling regression for a specific model; not the same bug." + }, + { + "left": "issue:44589", + "right": "issue:45042", + "accept": false, + "reason": "Float8 storage lookup error vs PIL backend unexpectedly requiring torchvision; distinct import/runtime issues." + }, + { + "left": "issue:43638", + "right": "issue:44936", + "accept": false, + "reason": "Deepspeed ZeRO-3 init index error vs trainer.evaluate failing after train; different trainer lifecycle bugs." + }, + { + "left": "issue:43550", + "right": "issue:44912", + "accept": false, + "reason": "torch.compile + SDPA failure on Bamba vs MXFP4 quantization fallback on gpt-oss; not the same concrete path." + }, + { + "left": "issue:45459", + "right": "issue:45803", + "accept": false, + "reason": "Tokenizer protobuf import-error masking vs bare except in FuyuBatchFeature conversion; both are error-handling issues, but not the same defect." + }, + { + "left": "issue:43819", + "right": "issue:43927", + "accept": false, + "reason": "DAC latent reconstruction mismatch vs DiaConfig custom token IDs lost on save/load; different models and bug classes." + }, + { + "left": "issue:36010", + "right": "issue:43645", + "accept": false, + "reason": "GenerationMixin import regression vs custom-model initialization in Jupyter; unrelated initialization/import behavior." + }, + { + "left": "issue:43257", + "right": "issue:43329", + "accept": false, + "reason": "Qwen3-MoE loading/conversion with accelerate+deepspeed vs multimodal video token counting undefined variables; different failure points." + }, + { + "left": "issue:43854", + "right": "issue:44977", + "accept": false, + "reason": "Unit-test model loading for GLM-4.7 vs Qwen3.5 flash-attention generation bug; not the same underlying issue." + }, + { + "left": "issue:39692", + "right": "issue:44912", + "accept": false, + "reason": "SigLIP2 doc example/model-processor mismatch vs MXFP4 quantization fallback; one is documentation/examples, the other model loading/quantization." + }, + { + "left": "issue:38175", + "right": "issue:44610", + "accept": false, + "reason": "Unexpected zero probabilities in SigLIP2 vs OmDet-Turbo processor image-size mismatch; different models and symptoms." + }, + { + "left": "issue:43720", + "right": "issue:44610", + "accept": false, + "reason": "BitNet packed-weight unpacking during accelerate load vs processor output-size mismatch; not the same code path." + }, + { + "left": "issue:36010", + "right": "issue:36331", + "accept": false, + "reason": "GenerationMixin import failure vs CustomTrainer.compute_loss signature mismatch; unrelated trainer/import problems." + }, + { + "left": "issue:43257", + "right": "issue:43696", + "accept": false, + "reason": "Model weight conversion/load bug vs CUDA OOM on GPT-oss-20b; different root causes." + }, + { + "left": "issue:43454", + "right": "issue:43577", + "accept": false, + "reason": "lm_head weight tying bug vs dtype propagation issue in BLIP2; both model-loading bugs, but not the same defect." + }, + { + "left": "issue:43262", + "right": "issue:44811", + "accept": false, + "reason": "Audio chat template sample-rate default bug vs Whisper batch_decode skip_special_tokens handling; separate processor APIs." + }, + { + "left": "issue:43531", + "right": "issue:45042", + "accept": false, + "reason": "Qwen3-MoE sliding_window behavior vs PIL backend torchvision dependency regression; unrelated areas." + }, + { + "left": "issue:42490", + "right": "issue:43257", + "accept": false, + "reason": "Trailing-slash local-path import warning vs Qwen3-MoE accelerate/deepspeed load failure; not the same issue." + }, + { + "left": "issue:43329", + "right": "issue:44479", + "accept": false, + "reason": "Undefined multimodal video helper variables vs broad v5.3.0 video-input regression; overlapping topic, but not the same concrete bug." + }, + { + "left": "issue:43643", + "right": "issue:45759", + "accept": false, + "reason": "AutoConfig trust_remote_code missing fields vs AutoModelForCausalLM.from_config not unwrapping text_config; related config handling, but distinct." + }, + { + "left": "issue:43749", + "right": "issue:44756", + "accept": false, + "reason": "FSDP CPU RAM efficient loading regression vs Strix Halo mmap OOM mitigation; different memory-loading issues." + }, + { + "left": "issue:43329", + "right": "issue:43450", + "accept": false, + "reason": "Undefined video-token helper names vs batched video processor shape error; both multimodal/video, but separate bugs." + }, + { + "left": "issue:43606", + "right": "issue:45859", + "accept": false, + "reason": "Bark CPU-offload device mismatch vs Qwen3-MoE rotary embedding CPU-offload incompatibility; same general theme, different models/code paths." + }, + { + "left": "issue:34567", + "right": "issue:40990", + "accept": false, + "reason": "TrainerState token counter not updating vs high perplexity on GPT-oss-20b; no concrete overlap." + }, + { + "left": "issue:44568", + "right": "issue:45356", + "accept": false, + "reason": "Tokenizer add_special_tokens regression vs Kimi-K2.5 codec handling regression; both tokenizer issues, but different mechanics." + }, + { + "left": "issue:43257", + "right": "issue:43931", + "accept": false, + "reason": "Qwen3-MoE load/conversion issue vs Qwen3-VL-30B weight-shape mismatch; both model-loading problems, but not the same bug." + }, + { + "left": "issue:29159", + "right": "issue:34567", + "accept": false, + "reason": "Slow/fast tokenizer inconsistency vs TrainerState num_input_tokens_seen not updating; unrelated subsystems." + }, + { + "left": "issue:44849", + "right": "issue:45701", + "accept": false, + "reason": "Qwen3.5 hidden-states bug vs broad tokenization changes across versions; too generic and not the same defect." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 17, + "review_unit_index": 18, + "review_unit_count": 50, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-428", + "nodes": [ + "issue:24643", + "issue:28282", + "issue:29127", + "issue:29159", + "issue:33357", + "issue:34567", + "issue:34689", + "issue:38175", + "issue:41720", + "issue:42175", + "issue:42915", + "issue:43257", + "issue:43317", + "issue:43404", + "issue:43452", + "issue:43540", + "issue:43577", + "issue:43582", + "issue:43697", + "issue:43742", + "issue:43749", + "issue:43756", + "issue:43782", + "issue:43824", + "issue:43906", + "issue:43927", + "issue:44062", + "issue:44291", + "issue:44479", + "issue:44488", + "issue:44560", + "issue:44568", + "issue:44821", + "issue:44898", + "issue:44912", + "issue:44987", + "issue:44991", + "issue:45042", + "issue:45072", + "issue:45137", + "issue:45357", + "issue:45381", + "issue:45405", + "issue:45488", + "issue:45538", + "issue:45823", + "issue:45907", + "issue:45987" + ], + "soft_pairs": [ + "issue:44568|issue:45488", + "issue:42915|issue:43742", + "issue:44291|issue:45042", + "issue:43697|issue:45357", + "issue:42915|issue:44821", + "issue:41720|issue:43540", + "issue:44987|issue:45538", + "issue:44479|issue:45381", + "issue:43452|issue:43582", + "issue:43927|issue:44912", + "issue:42915|issue:45538", + "issue:33357|issue:34567", + "issue:43540|issue:44479", + "issue:43540|issue:44560", + "issue:42175|issue:43824", + "issue:43317|issue:45538", + "issue:44987|issue:45907", + "issue:44062|issue:44991", + "issue:43257|issue:43756", + "issue:24643|issue:29159", + "issue:45137|issue:45405", + "issue:28282|issue:34689", + "issue:43577|issue:45072", + "issue:42915|issue:43749", + "issue:43742|issue:44291", + "issue:43906|issue:45381", + "issue:42175|issue:44488", + "issue:28282|issue:29127", + "issue:38175|issue:44898", + "issue:43582|issue:43782", + "issue:43404|issue:45987", + "issue:43540|issue:45823" + ], + "prepared_review_unit_hash": "97a25fd0508cfd78976e09eccf492e73a38df24fdf91ed86433874c03a9df201", + "component_budget": { + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 32, + "serialized_chars": 37373, + "estimated_input_tokens": 9344, + "estimated_eval_tokens": 18944 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly heterogeneous issue reports spanning unrelated models and subsystems. I found one plausible duplicate pair around Qwen video-input regression; the rest look like separate bugs and should not be merged.", + "confidence": 0.78, + "canonical_issue_reason": "issue:44479 is the best canonical issue because it describes the broader video-input regression across multiple Qwen model variants, while the paired report is a narrower symptom/reproduction.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:44479 is the strongest representative of the only likely duplicate thread in the cluster and has the broadest, most actionable bug description.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44568", + "right": "issue:45488", + "accept": false, + "reason": "Both are tokenizer regressions, but they affect different tokenizer code paths and different models; not the same underlying bug." + }, + { + "left": "issue:42915", + "right": "issue:43742", + "accept": false, + "reason": "Different failure modes and model-loading paths: FineGrainedFP8Config vs a key error for MobileLLM." + }, + { + "left": "issue:44291", + "right": "issue:45042", + "accept": false, + "reason": "Unrelated subsystems: init_empty_weights loading bug vs PIL image processor torchvision dependency." + }, + { + "left": "issue:43697", + "right": "issue:45357", + "accept": false, + "reason": "Different bugs in different areas: RTDetrV2 output drift vs incorrect visual encoder keys on save_pretrained." + }, + { + "left": "issue:42915", + "right": "issue:44821", + "accept": false, + "reason": "These are unrelated model/config issues; no evidence they share the same code-path or root cause." + }, + { + "left": "issue:41720", + "right": "issue:43540", + "accept": false, + "reason": "Both mention Qwen and runtime failures, but one is auto device mapping CUDA assert and the other is video-input handling in Qwen3OmniMoe." + }, + { + "left": "issue:44987", + "right": "issue:45538", + "accept": false, + "reason": "Different problems: one is a model load failure, the other is tokenizer max_length behavior." + }, + { + "left": "issue:44479", + "right": "issue:45381", + "accept": true, + "reason": "Same underlying Qwen video-input regression; 45381 appears to be a narrower reproduction of the broader v5.3.0 video-position bug described in 44479." + }, + { + "left": "issue:43452", + "right": "issue:43582", + "accept": false, + "reason": "Unrelated: gguf_file tokenizer/model loading breakage vs an Apple Silicon allocator warmup TypeError." + }, + { + "left": "issue:43927", + "right": "issue:44912", + "accept": false, + "reason": "Different issues: custom token IDs lost on save/load vs MXFP4 quantization fallback during model loading." + }, + { + "left": "issue:42915", + "right": "issue:45538", + "accept": false, + "reason": "Completely different topics: FP8 config loading for Qwen3Moe vs CLIPTokenizer max_length handling." + }, + { + "left": "issue:33357", + "right": "issue:34567", + "accept": false, + "reason": "One is a MacOS bus error on CLIP model loading; the other is TrainerState token counting not updating." + }, + { + "left": "issue:43540", + "right": "issue:44479", + "accept": false, + "reason": "Both are video-related, but they involve different Qwen model families and different failure modes; not the same bug." + }, + { + "left": "issue:43540", + "right": "issue:44560", + "accept": false, + "reason": "Same broad model family and video inputs, but the reported errors and likely code paths differ." + }, + { + "left": "issue:42175", + "right": "issue:43824", + "accept": false, + "reason": "Both are import/backend availability problems, but one is TensorFlow packaging and the other is a missing Qwen class export." + }, + { + "left": "issue:43317", + "right": "issue:45538", + "accept": false, + "reason": "Different areas: device-map/offload dequantization failure vs tokenizer max_length behavior." + }, + { + "left": "issue:44987", + "right": "issue:45907", + "accept": false, + "reason": "Unrelated: model loading failure vs a helper function not forwarding the revision argument." + }, + { + "left": "issue:44062", + "right": "issue:44991", + "accept": false, + "reason": "Both concern tokenizers, but one is an AddedToken argument collision and the other is a specific model tokenizer load failure." + }, + { + "left": "issue:43257", + "right": "issue:43756", + "accept": false, + "reason": "Different Qwen3Moe loading/conversion issue vs a RoPE-layer configuration bug in Smollm3." + }, + { + "left": "issue:24643", + "right": "issue:29159", + "accept": false, + "reason": "No shared root cause: DeepSpeed training weight-shape error vs slow/fast tokenizer inconsistency." + }, + { + "left": "issue:45137", + "right": "issue:45405", + "accept": false, + "reason": "Totally unrelated: ZeRO3 deque underflow vs an unreleased PEFT version bump." + }, + { + "left": "issue:28282", + "right": "issue:34689", + "accept": false, + "reason": "Different import/load errors for different models and environment states; not the same issue." + }, + { + "left": "issue:43577", + "right": "issue:45072", + "accept": false, + "reason": "Different dtype-related bugs affecting different model classes and code paths." + }, + { + "left": "issue:42915", + "right": "issue:43749", + "accept": false, + "reason": "Different topics: Qwen3Moe FP8 config failure vs FSDP CPU RAM efficient loading regression." + }, + { + "left": "issue:43742", + "right": "issue:44291", + "accept": false, + "reason": "Different mechanisms and subsystems; no evidence of a common underlying bug." + }, + { + "left": "issue:43906", + "right": "issue:45381", + "accept": false, + "reason": "The second is a Qwen video-position bug; the first is only an isolated reproduction reference and does not establish the same root cause." + }, + { + "left": "issue:42175", + "right": "issue:44488", + "accept": false, + "reason": "Packaging/backend installation issue vs a specific model-loading regression; not duplicates." + }, + { + "left": "issue:28282", + "right": "issue:29127", + "accept": false, + "reason": "Different bugs: missing PyTorch import vs a LayoutLMv3 box-content error message clarity issue." + }, + { + "left": "issue:38175", + "right": "issue:44898", + "accept": false, + "reason": "Different models and behaviors: unexpected zero probabilities vs Perceiver non-default resolution failure." + }, + { + "left": "issue:43582", + "right": "issue:43782", + "accept": false, + "reason": "Apple Silicon allocator warmup TypeError is unrelated to Qwen3VL weight_only loading failure." + }, + { + "left": "issue:43404", + "right": "issue:45987", + "accept": false, + "reason": "Different model families and different bugs: lm_head weight tying vs StaticCache sequence length return type." + }, + { + "left": "issue:43540", + "right": "issue:45823", + "accept": false, + "reason": "Different device/runtime bugs in different models; not the same code-path problem." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet and stays conservative. The only plausible duplicate pair supported by the titles is 44479/45381, and the rest of the soft-edge rejections are appropriately cautious given the available evidence. No overstatement stands out." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 17, + "review_unit_index": 19, + "review_unit_count": 50, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-428", + "nodes": [ + "issue:24643", + "issue:29127", + "issue:34567", + "issue:36331", + "issue:41762", + "issue:43452", + "issue:43493", + "issue:43526", + "issue:43531", + "issue:43643", + "issue:43688", + "issue:43697", + "issue:43720", + "issue:43742", + "issue:43749", + "issue:43756", + "issue:43994", + "issue:44062", + "issue:44291", + "issue:44488", + "issue:44514", + "issue:44589", + "issue:44625", + "issue:44737", + "issue:44792", + "issue:44821", + "issue:44843", + "issue:44877", + "issue:44898", + "issue:44913", + "issue:44936", + "issue:44964", + "issue:45042", + "issue:45081", + "issue:45216", + "issue:45290", + "issue:45325", + "issue:45357", + "issue:45372", + "issue:45375", + "issue:45381", + "issue:45405", + "issue:45479", + "issue:45588", + "issue:45600", + "issue:45920", + "issue:46121", + "issue:46170" + ], + "soft_pairs": [ + "issue:44737|issue:45588", + "issue:43526|issue:45479", + "issue:43756|issue:45325", + "issue:43493|issue:43994", + "issue:34567|issue:36331", + "issue:44913|issue:46121", + "issue:44291|issue:44821", + "issue:44062|issue:45081", + "issue:44792|issue:44964", + "issue:45357|issue:45920", + "issue:45042|issue:45081", + "issue:43452|issue:44488", + "issue:43742|issue:44589", + "issue:43643|issue:45375", + "issue:45405|issue:45600", + "issue:43531|issue:44821", + "issue:44625|issue:44877", + "issue:43697|issue:45216", + "issue:43720|issue:44898", + "issue:24643|issue:29127", + "issue:44514|issue:45381", + "issue:45372|issue:46170", + "issue:41762|issue:44936", + "issue:43688|issue:43749", + "issue:43452|issue:44843", + "issue:44514|issue:45290" + ], + "prepared_review_unit_hash": "ecd87d52674b92ce5117d09baf5021e8c7696e2b05a3d3a1c511fdb900195d75", + "component_budget": { + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 26, + "serialized_chars": 36040, + "estimated_input_tokens": 9010, + "estimated_eval_tokens": 18276 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is highly heterogeneous: the soft pairs mostly share only broad themes like loading, configs, or trainer behavior, but they point to different models/code paths and should not be merged as duplicates.", + "confidence": 0.94, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44737", + "right": "issue:45588", + "accept": false, + "reason": "XLNet CPU placement bug vs. flash-attention sinkless-model AttributeError; unrelated subsystems and failure modes." + }, + { + "left": "issue:43526", + "right": "issue:45479", + "accept": false, + "reason": "BEiT fast processor label reduction bug vs. sequence-classification loss degeneracy; different APIs and code paths." + }, + { + "left": "issue:43756", + "right": "issue:45325", + "accept": false, + "reason": "Smollm3 RoPE-layer truncation vs. Qwen2.5-VL position-id scaling; both are RoPE-related but for different models and bugs." + }, + { + "left": "issue:43493", + "right": "issue:43994", + "accept": false, + "reason": "Both are SigLIP2-related, but one is an HF-vs-JAX discrepancy and the other is a checkpoint/pipeline output issue; not clearly the same concrete bug." + }, + { + "left": "issue:34567", + "right": "issue:36331", + "accept": false, + "reason": "TrainerState token counting update bug vs. custom compute_loss signature break; different trainer internals." + }, + { + "left": "issue:44913", + "right": "issue:46121", + "accept": false, + "reason": "Config reload default reset vs. TypeError converting JSON-loaded rope ignore keys; distinct config handling problems." + }, + { + "left": "issue:44291", + "right": "issue:44821", + "accept": false, + "reason": "init_empty_weights/_is_hf_initialized loading bug vs. loading AutoImageProcessor from URL; unrelated failures." + }, + { + "left": "issue:44062", + "right": "issue:45081", + "accept": false, + "reason": "AddedToken duplicate-special kwarg error vs. mistral regex/backend_tokenizer crash; both tokenizer-related but different code paths." + }, + { + "left": "issue:44792", + "right": "issue:44964", + "accept": false, + "reason": "Janus image-generation test failure vs. Phi-4 multimodal load failure; different models and symptoms." + }, + { + "left": "issue:45357", + "right": "issue:45920", + "accept": false, + "reason": "Qwen3.5 save_pretrained key mismatch vs. AutoTokenizer wrong token IDs across many models; different bugs and components." + }, + { + "left": "issue:45042", + "right": "issue:45081", + "accept": false, + "reason": "PIL backend torchvision dependency regression vs. mistral regex crash; no shared underlying fix." + }, + { + "left": "issue:43452", + "right": "issue:44488", + "accept": false, + "reason": "gguf_file loader regression vs. specific sleng-bert loading failure; too broad and not same bug." + }, + { + "left": "issue:43742", + "right": "issue:44589", + "accept": false, + "reason": "MobileLLM key lookup failure vs. Float8 storage lookup error; both are load-time issues but unrelated." + }, + { + "left": "issue:43643", + "right": "issue:45375", + "accept": false, + "reason": "AutoConfig trust_remote_code field loss vs. Qwen3_5MoeVisionConfig strict-field drop; different mechanisms and models." + }, + { + "left": "issue:45405", + "right": "issue:45600", + "accept": false, + "reason": "Version pin mismatch vs. removed config references in auto_mappings; unrelated maintenance issues." + }, + { + "left": "issue:43531", + "right": "issue:44821", + "accept": false, + "reason": "Qwen3-MoE sliding_window behavior vs. AutoImageProcessor URL loading; no shared code path." + }, + { + "left": "issue:44625", + "right": "issue:44877", + "accept": false, + "reason": "num_labels propagation bug in Qwen3.5 vs. strict config blocking granite_speech loading; different config surfaces." + }, + { + "left": "issue:43697", + "right": "issue:45216", + "accept": false, + "reason": "RTDetrV2 output regression vs. Qwen3.5 save_pretrained regression; different model families and behavior." + }, + { + "left": "issue:43720", + "right": "issue:44898", + "accept": false, + "reason": "BitNet packed-weight loading bug vs. Perceiver non-default-res interpolation failure; unrelated model-loading problems." + }, + { + "left": "issue:24643", + "right": "issue:29127", + "accept": false, + "reason": "DeepSpeed training weight-shape runtime error vs. LayoutLMv3 error-message clarity issue; one is a bug, the other a UX/reporting issue." + }, + { + "left": "issue:44514", + "right": "issue:45381", + "accept": false, + "reason": "Qwen2.5-VL chat-template batch crash vs. Qwen2.5-VL video position-id bug; same family but different inputs and code paths." + }, + { + "left": "issue:45372", + "right": "issue:46170", + "accept": false, + "reason": "Gemma 4 processor import failure vs. Windows/ROCm torch.distributed dependency issue; unrelated platform/library problems." + }, + { + "left": "issue:41762", + "right": "issue:44936", + "accept": false, + "reason": "Gemma3 ZeRO-3 loading IndexError vs. trainer.evaluate after train; different stack traces and lifecycle stages." + }, + { + "left": "issue:43688", + "right": "issue:43749", + "accept": false, + "reason": "Aux-loss normalization bug vs. FSDP CPU RAM-efficient loading regression; different model/training subsystems." + }, + { + "left": "issue:43452", + "right": "issue:44843", + "accept": false, + "reason": "General gguf loading breakage vs. mistral_regex offline-model_info call; both loading-related but not the same issue." + }, + { + "left": "issue:44514", + "right": "issue:45290", + "accept": false, + "reason": "apply_chat_template batched-padding crash vs. tool-call/no-content assistant crash; both chat-template issues but distinct triggers and fixes." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 17, + "review_unit_index": 20, + "review_unit_count": 50, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-428", + "nodes": [ + "issue:28282", + "issue:29159", + "issue:30064", + "issue:32090", + "issue:33357", + "issue:33666", + "issue:35141", + "issue:38175", + "issue:41720", + "issue:43295", + "issue:43452", + "issue:43531", + "issue:43550", + "issue:43577", + "issue:43646", + "issue:43653", + "issue:43688", + "issue:43697", + "issue:43742", + "issue:43749", + "issue:43819", + "issue:43906", + "issue:43927", + "issue:44361", + "issue:44442", + "issue:44464", + "issue:44466", + "issue:44484", + "issue:44493", + "issue:44568", + "issue:44704", + "issue:44743", + "issue:44811", + "issue:44843", + "issue:44936", + "issue:44977", + "issue:44991", + "issue:45005", + "issue:45356", + "issue:45440", + "issue:45491", + "issue:45542", + "issue:45701", + "issue:45704", + "issue:45823", + "issue:46002", + "issue:46050", + "issue:46178" + ], + "soft_pairs": [ + "issue:44466|issue:45005", + "issue:45542|issue:46178", + "issue:43452|issue:43577", + "issue:44493|issue:46050", + "issue:43819|issue:44811", + "issue:30064|issue:33666", + "issue:44936|issue:45823", + "issue:43295|issue:45356", + "issue:44484|issue:44704", + "issue:43653|issue:44361", + "issue:28282|issue:32090", + "issue:44361|issue:44442", + "issue:41720|issue:43906", + "issue:28282|issue:33357", + "issue:44464|issue:44743", + "issue:38175|issue:43295", + "issue:28282|issue:35141", + "issue:45491|issue:46002", + "issue:43653|issue:44568", + "issue:43577|issue:43697", + "issue:43688|issue:45704", + "issue:44843|issue:44991", + "issue:43688|issue:45440", + "issue:43646|issue:43749", + "issue:29159|issue:33357", + "issue:43688|issue:45701", + "issue:43653|issue:43927", + "issue:43531|issue:43742", + "issue:44464|issue:44977", + "issue:43550|issue:43927" + ], + "prepared_review_unit_hash": "81a19ac26b882e5ca85bd7b3a7bf754cd53b3116800a85c2ca28687eda58d247", + "component_budget": { + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 30, + "serialized_chars": 36719, + "estimated_input_tokens": 9180, + "estimated_eval_tokens": 18616 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This set is highly heterogeneous: it mixes unrelated tokenizer regressions, generation bugs, training errors, backend/config issues, and model-loading problems. I do not see any convincing duplicate pairs; most similarities are only at the subsystem level. Issue 45701 is the broadest umbrella-style issue in the set, but the cluster does not read like one single bug family.", + "confidence": 0.84, + "canonical_issue_reason": "Issue 45701 is the most umbrella-like and currently open item; it best matches the broad version-dependent tokenization regression theme that appears in a few nearby issues.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 45701 is the best anchor for triage because it is broad, active, and representative of the only recurring theme that shows up across multiple items (tokenization behavior changing across versions).", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44466", + "right": "issue:45005", + "accept": false, + "reason": "Both mention tied weights, but one is a concrete serialization/device-dependent lm_head.weight issue while the other is a broader translation-model v5 regression." + }, + { + "left": "issue:45542", + "right": "issue:46178", + "accept": false, + "reason": "Both are backend/TensorFlow related, but the failure modes differ: missing TensorFlow with tensorboard vs tensorflow_text not being in BACKENDS_MAPPING." + }, + { + "left": "issue:43452", + "right": "issue:43577", + "accept": false, + "reason": "Different subsystems: gguf tokenizer/model loading vs BLIP2 dtype propagation for qformer/model." + }, + { + "left": "issue:44493", + "right": "issue:46050", + "accept": false, + "reason": "Different bugs: unexpected position-id key handling versus quantized model save/load corruption after sinq." + }, + { + "left": "issue:43819", + "right": "issue:44811", + "accept": false, + "reason": "DAC latent conversion with missing STE is unrelated to Whisper batch_decode ignoring skip_special_tokens." + }, + { + "left": "issue:30064", + "right": "issue:33666", + "accept": false, + "reason": "Image segmentation map processing and Qwen2-VL multi-GPU training are unrelated." + }, + { + "left": "issue:44936", + "right": "issue:45823", + "accept": false, + "reason": "Trainer.evaluate() after train() is a training-loop bug, not the Gemma4 device-mismatch forward issue." + }, + { + "left": "issue:43295", + "right": "issue:45356", + "accept": false, + "reason": "Both are tokenizer/processor regressions, but one is processor.tokenizer/images passthrough and the other is Kimi codec/fix_mistral_regex behavior." + }, + { + "left": "issue:44484", + "right": "issue:44704", + "accept": false, + "reason": "Max shard size default sizing and AutoProcessor kwargs forwarding are unrelated." + }, + { + "left": "issue:43653", + "right": "issue:44361", + "accept": false, + "reason": "Both are tokenizer issues, but the concrete failures differ: BigBird special-token decode vs MLukeTokenizer AttributeError." + }, + { + "left": "issue:28282", + "right": "issue:32090", + "accept": false, + "reason": "Missing PyTorch import error and Trainer _gpu_broadcast_one NoneType error are unrelated." + }, + { + "left": "issue:44361", + "right": "issue:44442", + "accept": false, + "reason": "Both concern tokenizers, but they hit different tokenizer classes with different loading/runtime failures." + }, + { + "left": "issue:41720", + "right": "issue:43906", + "accept": false, + "reason": "Qwen3 auto device-map cudaErrorAssert is not the same as the separate isolated reproduction of issue 38071." + }, + { + "left": "issue:28282", + "right": "issue:33357", + "accept": false, + "reason": "PyTorch import failure and MacOS bus error with a CLIP model are unrelated." + }, + { + "left": "issue:44464", + "right": "issue:44743", + "accept": false, + "reason": "Both involve generation/caching, but chunked generation with compiled forward is a different code path from recurrent-state reset with seq_len>1." + }, + { + "left": "issue:38175", + "right": "issue:43295", + "accept": false, + "reason": "Siglip2 zero probabilities are unrelated to the processor.tokenizer regression." + }, + { + "left": "issue:28282", + "right": "issue:35141", + "accept": false, + "reason": "ImportError for missing PyTorch is unrelated to post_init reinitializing output embeddings after resize." + }, + { + "left": "issue:45491", + "right": "issue:46002", + "accept": false, + "reason": "Both mention NaNs, but one is mixed-length sliding-window attention on Gemma3 and the other is classification-head NaNs after checkout/init." + }, + { + "left": "issue:43653", + "right": "issue:44568", + "accept": false, + "reason": "Both are special-token/tokenizer issues, but they are different models and different behaviors (mask token registration vs BOS/EOS insertion)." + }, + { + "left": "issue:43577", + "right": "issue:43697", + "accept": false, + "reason": "Dtype staying float32 in BLIP2 is unrelated to RTDetrV2 output divergence in v5." + }, + { + "left": "issue:43688", + "right": "issue:45704", + "accept": false, + "reason": "Auxiliary-loss normalization in MoE models is unrelated to T5 silently using apex FusedRMSNorm and leaking memory." + }, + { + "left": "issue:44843", + "right": "issue:44991", + "accept": false, + "reason": "Both are tokenizer-loading regressions, but one is an offline model_info call bug and the other is a specific tokenizer load failure for est-roberta." + }, + { + "left": "issue:43688", + "right": "issue:45440", + "accept": false, + "reason": "Both touch MoE models, but auxiliary-loss normalization and DeepseekV3MoE divergence are not the same concrete bug." + }, + { + "left": "issue:43646", + "right": "issue:43749", + "accept": false, + "reason": "Custom model initialization breakage and FSDP CPU RAM efficient loading are different failure modes." + }, + { + "left": "issue:29159", + "right": "issue:33357", + "accept": false, + "reason": "Slow/fast tokenizer inconsistency is unrelated to the MacOS bus error report." + }, + { + "left": "issue:43688", + "right": "issue:45701", + "accept": false, + "reason": "MoE auxiliary-loss normalization is unrelated to version-dependent tokenization behavior." + }, + { + "left": "issue:43653", + "right": "issue:43927", + "accept": false, + "reason": "BigBird special-token decode and DiaConfig losing custom token IDs after save/load are different tokenizer/config bugs." + }, + { + "left": "issue:43531", + "right": "issue:43742", + "accept": false, + "reason": "Qwen3-MoE sliding_window behavior and MobileLLM-125M loading key error are unrelated." + }, + { + "left": "issue:44464", + "right": "issue:44977", + "accept": false, + "reason": "Both mention generation, but compiled-forward chunked generation and flash-attention failure on Qwen3.5 are different code-path problems." + }, + { + "left": "issue:43550", + "right": "issue:43927", + "accept": false, + "reason": "torch.compile+SDPA failure in Bamba and DiaConfig custom token ID loss are unrelated." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 17, + "review_unit_index": 21, + "review_unit_count": 50, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-428", + "nodes": [ + "issue:28218", + "issue:32090", + "issue:33357", + "issue:33666", + "issue:34567", + "issue:36010", + "issue:36331", + "issue:41720", + "issue:43122", + "issue:43262", + "issue:43278", + "issue:43295", + "issue:43452", + "issue:43493", + "issue:43504", + "issue:43525", + "issue:43550", + "issue:43572", + "issue:43577", + "issue:43606", + "issue:43643", + "issue:43688", + "issue:43720", + "issue:43742", + "issue:43756", + "issue:43825", + "issue:43906", + "issue:43927", + "issue:43994", + "issue:44361", + "issue:44451", + "issue:44464", + "issue:44466", + "issue:44479", + "issue:44554", + "issue:44561", + "issue:44610", + "issue:44625", + "issue:44821", + "issue:44849", + "issue:44913", + "issue:45357", + "issue:45588", + "issue:45823", + "issue:46002", + "issue:46050", + "issue:46192" + ], + "soft_pairs": [ + "issue:43643|issue:44561", + "issue:44479|issue:44625", + "issue:43550|issue:44361", + "issue:41720|issue:44464", + "issue:43122|issue:43906", + "issue:43577|issue:44554", + "issue:43577|issue:45588", + "issue:32090|issue:34567", + "issue:43504|issue:43550", + "issue:43504|issue:43720", + "issue:43504|issue:44610", + "issue:44466|issue:46050", + "issue:32090|issue:33666", + "issue:33666|issue:36010", + "issue:43742|issue:44821", + "issue:45823|issue:46050", + "issue:43452|issue:44451", + "issue:43720|issue:43927", + "issue:43606|issue:43927", + "issue:43278|issue:43688", + "issue:43994|issue:46192", + "issue:43262|issue:43295", + "issue:43525|issue:43572", + "issue:44625|issue:46002", + "issue:33666|issue:36331", + "issue:43720|issue:44361", + "issue:43756|issue:44913", + "issue:44849|issue:45357", + "issue:43493|issue:43825", + "issue:28218|issue:33357" + ], + "prepared_review_unit_hash": "a361b2d36e60a78c7ce317badf55be84e18c6bf988ca247fcac96a93d046b106", + "component_budget": { + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 30, + "serialized_chars": 36168, + "estimated_input_tokens": 9042, + "estimated_eval_tokens": 18340 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This set is mostly heterogeneous with many unrelated bug reports. The only credible duplicate-style overlap is the pair about v5 config losing pad-token fields on reload; the rest are distinct code paths and should stay separate.", + "confidence": 0.8, + "canonical_issue_reason": "issue:43525 is the clearest representative of the pad-token config regression, with the most generic title and a direct match to the shared underlying load/reload bug.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43525 best captures the recurring config-field-loss problem and is the strongest anchor for the one accepted duplicate pair.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43643", + "right": "issue:44561", + "accept": false, + "reason": "Both involve trust_remote_code/v5 breakage, but they describe different root causes: missing config fields vs removal of a helper used by remote-code models." + }, + { + "left": "issue:44479", + "right": "issue:44625", + "accept": false, + "reason": "Video-input regression and num_labels propagation are unrelated model bugs." + }, + { + "left": "issue:43550", + "right": "issue:44361", + "accept": false, + "reason": "Different subsystems and failures: SDPA/torch.compile on Bamba vs tokenizer AttributeError on tasks." + }, + { + "left": "issue:41720", + "right": "issue:44464", + "accept": false, + "reason": "Qwen3 auto device mapping cuda assert and chunked generation compiled-forward inconsistency are separate issues." + }, + { + "left": "issue:43122", + "right": "issue:43906", + "accept": false, + "reason": "Tokenization-version change and an isolated reproduction of an older issue are not enough to show the same underlying bug." + }, + { + "left": "issue:43577", + "right": "issue:44554", + "accept": false, + "reason": "Different problems: dtype handling in BLIP2 loading vs an MPS attention correctness issue." + }, + { + "left": "issue:43577", + "right": "issue:45588", + "accept": false, + "reason": "BLIP2 dtype mismatch and flash_attention s_aux=None crash are unrelated." + }, + { + "left": "issue:32090", + "right": "issue:34567", + "accept": false, + "reason": "Trainer broadcast NoneType error and num_input_tokens_seen not updating are distinct trainer bugs." + }, + { + "left": "issue:43504", + "right": "issue:43550", + "accept": false, + "reason": "Legacy-field model loading failure vs torch.compile SDPA failure are different code paths." + }, + { + "left": "issue:43504", + "right": "issue:43720", + "accept": false, + "reason": "Both are loading-related, but one is a legacy preset field issue and the other is about unpacking packed weights during accelerate loading." + }, + { + "left": "issue:43504", + "right": "issue:44610", + "accept": false, + "reason": "Different model families and failures: legacy preset loading vs processor size mismatch." + }, + { + "left": "issue:44466", + "right": "issue:46050", + "accept": false, + "reason": "Both mention save/load behavior, but one is lm_head serialization inconsistency and the other is quantized sinq model restore breakage." + }, + { + "left": "issue:32090", + "right": "issue:33666", + "accept": false, + "reason": "Trainer GPU broadcast failure and Qwen2-VL multi-GPU training are not the same bug." + }, + { + "left": "issue:33666", + "right": "issue:36010", + "accept": false, + "reason": "Multi-GPU training and GenerationMixin import failure are unrelated." + }, + { + "left": "issue:43742", + "right": "issue:44821", + "accept": false, + "reason": "KeyError loading one model and inability to load AutoImageProcessor from URL are different loading issues." + }, + { + "left": "issue:45823", + "right": "issue:46050", + "accept": false, + "reason": "Device-map forward mismatch and sinq save/load breakage are unrelated." + }, + { + "left": "issue:43452", + "right": "issue:44451", + "accept": false, + "reason": "Generic tokenizer/model loading failure and inability to load a specific model are too broad to treat as the same bug." + }, + { + "left": "issue:43720", + "right": "issue:43927", + "accept": false, + "reason": "Packed-weight loading and DiaConfig token-id loss are different serialization problems." + }, + { + "left": "issue:43606", + "right": "issue:43927", + "accept": false, + "reason": "CPU offload device mismatch and config token-id loss are unrelated." + }, + { + "left": "issue:43278", + "right": "issue:43688", + "accept": false, + "reason": "Embedding dtype drift in evaluate and auxiliary-loss normalization are separate issues." + }, + { + "left": "issue:43994", + "right": "issue:46192", + "accept": false, + "reason": "SigLIP2 output problems and return_tensors being ignored are unrelated API bugs." + }, + { + "left": "issue:43262", + "right": "issue:43295", + "accept": false, + "reason": "Audio processor sampling-rate defaulting and processor/tokenizer image-handling regression are different behaviors." + }, + { + "left": "issue:43525", + "right": "issue:43572", + "accept": true, + "reason": "Both report v5 config reload losing pad-token fields; they appear to be the same underlying config-propagation regression with different field names/model classes." + }, + { + "left": "issue:44625", + "right": "issue:46002", + "accept": false, + "reason": "num_labels propagation and NaNs in classification heads are not the same defect." + }, + { + "left": "issue:33666", + "right": "issue:36331", + "accept": false, + "reason": "Multi-GPU training and custom compute_loss signature mismatch are unrelated trainer issues." + }, + { + "left": "issue:43720", + "right": "issue:44361", + "accept": false, + "reason": "Packed-weight loading and MLukeTokenizer task AttributeError are different failures." + }, + { + "left": "issue:43756", + "right": "issue:44913", + "accept": false, + "reason": "Smollm3 RoPE-layer count mismatch and GPTNeoX rotary_pct reload regression are only loosely related to rotary settings, not the same bug." + }, + { + "left": "issue:44849", + "right": "issue:45357", + "accept": false, + "reason": "Qwen3.5 hidden-states bug and incorrect visual encoder key serialization are different issues." + }, + { + "left": "issue:43493", + "right": "issue:43825", + "accept": false, + "reason": "SigLIP2 implementation discrepancy and pipeline translation-message regression are unrelated." + }, + { + "left": "issue:28218", + "right": "issue:33357", + "accept": false, + "reason": "Tokenizer extra-space behavior and a MacOS bus error on a CLIP model are unrelated." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is well grounded in the packet, and the only accepted pair is the one soft edge that plausibly represents the same v5 config reload regression. The other verdicts remain conservative and do not overclaim duplicate status." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 17, + "review_unit_index": 22, + "review_unit_count": 50, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-428", + "nodes": [ + "issue:33357", + "issue:33666", + "issue:39692", + "issue:43257", + "issue:43278", + "issue:43295", + "issue:43299", + "issue:43334", + "issue:43493", + "issue:43504", + "issue:43540", + "issue:43572", + "issue:43575", + "issue:43606", + "issue:43646", + "issue:43653", + "issue:43701", + "issue:43756", + "issue:43824", + "issue:43844", + "issue:43927", + "issue:43994", + "issue:44112", + "issue:44291", + "issue:44361", + "issue:44423", + "issue:44464", + "issue:44466", + "issue:44488", + "issue:44493", + "issue:44568", + "issue:44589", + "issue:44734", + "issue:44792", + "issue:44843", + "issue:44964", + "issue:44977", + "issue:45083", + "issue:45200", + "issue:45229", + "issue:45405", + "issue:45406", + "issue:45520", + "issue:45584", + "issue:45759", + "issue:45812", + "issue:45823", + "issue:46121" + ], + "soft_pairs": [ + "issue:44568|issue:45812", + "issue:39692|issue:43493", + "issue:44964|issue:45405", + "issue:43540|issue:44977", + "issue:45200|issue:45229", + "issue:44488|issue:44843", + "issue:43334|issue:43572", + "issue:43493|issue:43844", + "issue:43278|issue:43844", + "issue:43257|issue:43299", + "issue:43334|issue:45759", + "issue:43756|issue:45083", + "issue:39692|issue:43295", + "issue:45229|issue:45823", + "issue:43824|issue:44589", + "issue:43646|issue:44466", + "issue:43504|issue:43606", + "issue:33357|issue:33666", + "issue:43927|issue:44361", + "issue:44589|issue:45520", + "issue:43572|issue:44291", + "issue:44464|issue:45584", + "issue:43504|issue:43653", + "issue:43606|issue:44112", + "issue:43278|issue:43701", + "issue:43575|issue:45405", + "issue:44792|issue:45405", + "issue:43994|issue:44493", + "issue:44423|issue:44734", + "issue:45406|issue:46121" + ], + "prepared_review_unit_hash": "d33fef169d33194cc24c0a32053a8914aae2cbfd5f5b92be6bb2799ecd217644", + "component_budget": { + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 30, + "serialized_chars": 36957, + "estimated_input_tokens": 9240, + "estimated_eval_tokens": 18736 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is highly heterogeneous: it mixes unrelated tokenizer, multimodal, serving, training, config, and CI regressions. None of the soft pairs appears to describe the same concrete bug or change, so all soft edges should be rejected.", + "confidence": 0.97, + "canonical_issue_reason": "No single canonical issue stands out; the items span distinct models and failure modes, so any duplicate anchor would be arbitrary.", + "canonical_pr_reason": null, + "best_issue_reason": "No globally best issue candidate is evident because the cluster is not centered on one underlying defect.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44568", + "right": "issue:45812", + "accept": false, + "reason": "Different tokenizer regressions: missing BOS/EOS on mdeberta-v3 vs wrong token IDs on Granite." + }, + { + "left": "issue:39692", + "right": "issue:43493", + "accept": false, + "reason": "Docs/example errors in SigLIP2 vs an implementation discrepancy with the original JAX model." + }, + { + "left": "issue:44964", + "right": "issue:45405", + "accept": false, + "reason": "Loading Phi-4 multimodal breaks vs a PEFT version bump; unrelated failure modes." + }, + { + "left": "issue:43540", + "right": "issue:44977", + "accept": false, + "reason": "Video-input processing error in Qwen3OmniMoe vs flash-attention generation issues in Qwen3.5." + }, + { + "left": "issue:45200", + "right": "issue:45229", + "accept": false, + "reason": "Text-only mm_token_type_ids behavior in Gemma 4 vs multi-GPU CUDA OOM in Gemma4 31B." + }, + { + "left": "issue:44488", + "right": "issue:44843", + "accept": false, + "reason": "Model loading failure for sleng-bert vs offline-mode breakage from an unconditional hub call." + }, + { + "left": "issue:43334", + "right": "issue:43572", + "accept": false, + "reason": "Missing pad_token_id in Qwen3-VL config vs missing pad_token_idx in StableLmConfig." + }, + { + "left": "issue:43493", + "right": "issue:43844", + "accept": false, + "reason": "SigLIP2 implementation mismatch vs exploding gradients under ZeRO-3; different bugs." + }, + { + "left": "issue:43278", + "right": "issue:43844", + "accept": false, + "reason": "Dtype drift between train/eval vs gradient instability under DeepSpeed ZeRO-3." + }, + { + "left": "issue:43257", + "right": "issue:43299", + "accept": false, + "reason": "Accelerate/DeepSpeed weight conversion issue vs dev-version load failure for Qwen3VL MoE models." + }, + { + "left": "issue:43334", + "right": "issue:45759", + "accept": false, + "reason": "Pad-token config bug in Qwen3-VL vs composite config unwrapping for Qwen 3.5/3.6." + }, + { + "left": "issue:43756", + "right": "issue:45083", + "accept": false, + "reason": "RoPE layer dropout in Smollm3 vs a feat-extract helper bug in qwen3_omni_moe." + }, + { + "left": "issue:39692", + "right": "issue:43295", + "accept": false, + "reason": "SigLIP2 doc example problems vs processor.tokenizer/image-tokenizer regression." + }, + { + "left": "issue:45229", + "right": "issue:45823", + "accept": false, + "reason": "OOM during Gemma4 inference vs device mismatch with device_map='auto'." + }, + { + "left": "issue:43824", + "right": "issue:44589", + "accept": false, + "reason": "Import error for Qwen2_5_VLForConditionalGeneration vs Float8 storage lookup failure." + }, + { + "left": "issue:43646", + "right": "issue:44466", + "accept": false, + "reason": "Custom model init breakage in v5 vs device-dependent lm_head serialization inconsistency." + }, + { + "left": "issue:43504", + "right": "issue:43606", + "accept": false, + "reason": "Legacy field prevents BEiT segmentation load vs Bark-small CPU offload device mismatch." + }, + { + "left": "issue:33357", + "right": "issue:33666", + "accept": false, + "reason": "MacOS bus error on CLIP model load vs Qwen2-VL multi-GPU training; unrelated." + }, + { + "left": "issue:43927", + "right": "issue:44361", + "accept": false, + "reason": "DiaConfig save/load token-id loss vs MLukeTokenizer AttributeError on tasks." + }, + { + "left": "issue:44589", + "right": "issue:45520", + "accept": false, + "reason": "Float8 storage type lookup failure vs Python 3.13 flash_attn import key error." + }, + { + "left": "issue:43572", + "right": "issue:44291", + "accept": false, + "reason": "Missing StableLmConfig field vs init_empty_weights signature mismatch; distinct initialization bugs." + }, + { + "left": "issue:44464", + "right": "issue:45584", + "accept": false, + "reason": "Compiled-forward chunked generation inconsistency vs Whisper empty-transcription failure." + }, + { + "left": "issue:43504", + "right": "issue:43653", + "accept": false, + "reason": "BEiT legacy config load failure vs BigBirdTokenizer special-token registration issue." + }, + { + "left": "issue:43606", + "right": "issue:44112", + "accept": false, + "reason": "Runtime device mismatch in Bark-small vs a stale GraniteSpeech CI test." + }, + { + "left": "issue:43278", + "right": "issue:43701", + "accept": false, + "reason": "Embedding dtype regression vs resume_from_checkpoint key mismatch." + }, + { + "left": "issue:43575", + "right": "issue:45405", + "accept": false, + "reason": "Tensor-parallel OOM on Qwen2-57B vs unreleased PEFT version bump." + }, + { + "left": "issue:44792", + "right": "issue:45405", + "accept": false, + "reason": "Janus image-generation test failure vs PEFT dependency versioning; unrelated." + }, + { + "left": "issue:43994", + "right": "issue:44493", + "accept": false, + "reason": "SigLIP2 nonsense outputs vs unexpected position-id key warnings across models." + }, + { + "left": "issue:44423", + "right": "issue:44734", + "accept": false, + "reason": "Two different serve crashes: string .to() error vs wrong KV-cache indexing." + }, + { + "left": "issue:45406", + "right": "issue:46121", + "accept": false, + "reason": "Gemma4Processor missing _tokenizer in serve vs rope-parameter validation TypeError." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 17, + "review_unit_index": 23, + "review_unit_count": 50, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-428", + "nodes": [ + "issue:15354", + "issue:28218", + "issue:29159", + "issue:29942", + "issue:30333", + "issue:32090", + "issue:34567", + "issue:37428", + "issue:38175", + "issue:39692", + "issue:42490", + "issue:43232", + "issue:43295", + "issue:43450", + "issue:43493", + "issue:43540", + "issue:43572", + "issue:43582", + "issue:43696", + "issue:43701", + "issue:43824", + "issue:43844", + "issue:43873", + "issue:43957", + "issue:44186", + "issue:44315", + "issue:44521", + "issue:44554", + "issue:44610", + "issue:44849", + "issue:44964", + "issue:44977", + "issue:44987", + "issue:45072", + "issue:45084", + "issue:45198", + "issue:45200", + "issue:45290", + "issue:45325", + "issue:45356", + "issue:45446", + "issue:45520", + "issue:45588", + "issue:45823", + "issue:46050", + "issue:46178", + "issue:46192" + ], + "soft_pairs": [ + "issue:43873|issue:46050", + "issue:29942|issue:30333", + "issue:45588|issue:46050", + "issue:42490|issue:43696", + "issue:44521|issue:45290", + "issue:43844|issue:44964", + "issue:28218|issue:34567", + "issue:29159|issue:32090", + "issue:43582|issue:44977", + "issue:44554|issue:45446", + "issue:45084|issue:45520", + "issue:43540|issue:45200", + "issue:15354|issue:38175", + "issue:45200|issue:45823", + "issue:45072|issue:45198", + "issue:37428|issue:43957", + "issue:43232|issue:44849", + "issue:44186|issue:46192", + "issue:43493|issue:43701", + "issue:45823|issue:46192", + "issue:44315|issue:44849", + "issue:43295|issue:44610", + "issue:44554|issue:45356", + "issue:43295|issue:43450", + "issue:43572|issue:44987", + "issue:45200|issue:45325", + "issue:39692|issue:46192", + "issue:43572|issue:43957", + "issue:37428|issue:43824", + "issue:45520|issue:46178" + ], + "prepared_review_unit_hash": "b3e926f781aa5bf3c86469b60c31927d08d046b5e8c15ebe820e9fefe1f81be3", + "component_budget": { + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 30, + "serialized_chars": 36155, + "estimated_input_tokens": 9039, + "estimated_eval_tokens": 18334 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is highly heterogeneous: most items are unrelated tokenizer, multimodal processor, flash-attention, loading, training, or quantization bugs. The soft edges are mostly embedding-level similarities rather than true duplicates, so none should be merged.", + "confidence": 0.62, + "canonical_issue_reason": "No single canonical issue is safe here because the cluster is over-broad. If one representative must be chosen, issue:43295 is the closest concrete, user-facing regression among the few multimodal/processor-related items, but it does not represent the whole cluster.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue:43295 is the best representative of the densest coherent subtheme in this cluster: a regression around processor/tokenizer behavior in multimodal code paths. It is more concrete than the broader, noisier entries and has the strongest structural similarity to the few nearby processor-related issues.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43873", + "right": "issue:46050", + "accept": false, + "reason": "Different bugs: quantization/offloading behavior versus save/load correctness for a quantized model. Same broad area, not same underlying defect." + }, + { + "left": "issue:29942", + "right": "issue:30333", + "accept": false, + "reason": "Completely unrelated subsystems: Flash Attention tests versus MLFlow reporting on failed jobs." + }, + { + "left": "issue:45588", + "right": "issue:46050", + "accept": false, + "reason": "Different code paths: flash_attention sink-less crash versus sinq quantized model serialization/deserialization." + }, + { + "left": "issue:42490", + "right": "issue:43696", + "accept": false, + "reason": "Local-path import warning is unrelated to CUDA OOM during GPT-oss loading." + }, + { + "left": "issue:44521", + "right": "issue:45290", + "accept": false, + "reason": "Both involve apply_chat_template, but one is multimodal assistant masks and the other is tool-call tokenization crash; these are distinct failures." + }, + { + "left": "issue:43844", + "right": "issue:44964", + "accept": false, + "reason": "Unrelated: ZeRO-3 gradient anomaly versus loading failure for a specific multimodal model." + }, + { + "left": "issue:28218", + "right": "issue:34567", + "accept": false, + "reason": "Tokenizer spacing bug and TrainerState token counter update are unrelated." + }, + { + "left": "issue:29159", + "right": "issue:32090", + "accept": false, + "reason": "Slow/fast tokenizer inconsistency is unrelated to a Trainer GPU broadcast NoneType error." + }, + { + "left": "issue:43582", + "right": "issue:44977", + "accept": false, + "reason": "Apple Silicon allocator warmup TypeError and Qwen3.5 flash-attention generation bug are different problems." + }, + { + "left": "issue:44554", + "right": "issue:45446", + "accept": false, + "reason": "Both touch attention/flex_attention, but one is an MPS correctness issue and the other is a version-check import bug." + }, + { + "left": "issue:45084", + "right": "issue:45520", + "accept": false, + "reason": "Compiler/template-node error is unrelated to the Python 3.13 flash_attn import KeyError." + }, + { + "left": "issue:43540", + "right": "issue:45200", + "accept": false, + "reason": "Both are multimodal, but one is video-input processing in Qwen3OmniMoe and the other is a Gemma4 token-type-id defaulting issue." + }, + { + "left": "issue:15354", + "right": "issue:38175", + "accept": false, + "reason": "TorchScript GeneratorExp export failure and SigLIP2 zero probabilities are unrelated model behaviors." + }, + { + "left": "issue:45200", + "right": "issue:45823", + "accept": false, + "reason": "Same model family (Gemma4) but different bugs: required mm_token_type_ids versus device-map forward mismatch." + }, + { + "left": "issue:45072", + "right": "issue:45198", + "accept": false, + "reason": "Different failures in different models: dtype mismatch in inference versus save_pretrained/tokenization breakage." + }, + { + "left": "issue:37428", + "right": "issue:43957", + "accept": false, + "reason": "Flash-attention import symbol error and meta-device model-loading breakage are distinct issues." + }, + { + "left": "issue:43232", + "right": "issue:44849", + "accept": false, + "reason": "Generation kwargs sync issue and Qwen3.5 hidden-states bug are separate generation-path defects." + }, + { + "left": "issue:44186", + "right": "issue:46192", + "accept": false, + "reason": "NER/batched-padding tokenizer crash is not the same as silently ignored return_tensors when text_kwargs is passed." + }, + { + "left": "issue:43493", + "right": "issue:43701", + "accept": false, + "reason": "SigLIP2 implementation discrepancy and checkpoint resume key mismatch are unrelated." + }, + { + "left": "issue:45823", + "right": "issue:46192", + "accept": false, + "reason": "Device-map mismatch in Gemma4 and ignored return_tensors in multimodal kwargs are different bugs." + }, + { + "left": "issue:44315", + "right": "issue:44849", + "accept": false, + "reason": "Liger Kernel application with model_init is unrelated to Qwen3.5 output_hidden_states behavior." + }, + { + "left": "issue:43295", + "right": "issue:44610", + "accept": false, + "reason": "Both involve multimodal processing, but the first is a processor.tokenizer regression and the second is a video processor output-shape bug; not the same underlying defect." + }, + { + "left": "issue:44554", + "right": "issue:45356", + "accept": false, + "reason": "MPS attention correctness and Kimi-K2.5 tokenizer codec regression are unrelated." + }, + { + "left": "issue:43295", + "right": "issue:43450", + "accept": false, + "reason": "Processor.tokenizer regression and batched video processor shape bug are different code paths and would not be merged into one fix." + }, + { + "left": "issue:43572", + "right": "issue:44987", + "accept": false, + "reason": "StableLmConfig pad_token_idx regression is unrelated to physical-intelligence/fast loading failure." + }, + { + "left": "issue:45200", + "right": "issue:45325", + "accept": false, + "reason": "Both are multimodal model issues, but one is missing/default mm_token_type_ids and the other is a Qwen2.5-VL rope-index scaling bug." + }, + { + "left": "issue:39692", + "right": "issue:46192", + "accept": false, + "reason": "SigLIP2 documentation errors and ignored return_tensors are unrelated." + }, + { + "left": "issue:43572", + "right": "issue:43957", + "accept": false, + "reason": "Pad token config regression and meta-device model-loading failure are not the same bug." + }, + { + "left": "issue:37428", + "right": "issue:43824", + "accept": false, + "reason": "Flash-attention utility import error and Qwen2.5-VL class import error are separate import problems with different missing symbols." + }, + { + "left": "issue:45520", + "right": "issue:46178", + "accept": false, + "reason": "Both are import/configuration errors, but flash_attn KeyError on Python 3.13 and tensorflow_text backend mapping are unrelated." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 17, + "review_unit_index": 24, + "review_unit_count": 50, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-428", + "nodes": [ + "issue:28218", + "issue:35141", + "issue:37428", + "issue:38175", + "issue:42490", + "issue:42915", + "issue:43262", + "issue:43278", + "issue:43295", + "issue:43334", + "issue:43335", + "issue:43388", + "issue:43425", + "issue:43450", + "issue:43493", + "issue:43525", + "issue:43526", + "issue:43550", + "issue:43575", + "issue:43606", + "issue:43653", + "issue:43819", + "issue:43824", + "issue:43827", + "issue:43844", + "issue:43856", + "issue:43873", + "issue:43901", + "issue:43975", + "issue:44112", + "issue:44206", + "issue:44442", + "issue:44466", + "issue:44514", + "issue:44521", + "issue:44610", + "issue:44792", + "issue:44857", + "issue:44898", + "issue:45083", + "issue:45127", + "issue:45198", + "issue:45412", + "issue:45520", + "issue:45563", + "issue:45584", + "issue:46170", + "issue:46192" + ], + "soft_pairs": [ + "issue:44898|issue:45412", + "issue:44466|issue:45127", + "issue:43295|issue:43873", + "issue:43550|issue:43844", + "issue:43334|issue:43525", + "issue:44792|issue:45563", + "issue:43824|issue:43975", + "issue:42915|issue:43824", + "issue:43388|issue:43526", + "issue:43450|issue:43575", + "issue:43278|issue:43856", + "issue:42490|issue:43493", + "issue:43975|issue:45520", + "issue:28218|issue:35141", + "issue:43653|issue:44112", + "issue:38175|issue:43493", + "issue:43526|issue:46192", + "issue:44442|issue:45198", + "issue:44514|issue:44521", + "issue:45520|issue:46170", + "issue:38175|issue:45412", + "issue:44206|issue:45584", + "issue:43827|issue:43901", + "issue:43262|issue:43450", + "issue:43819|issue:44112", + "issue:44610|issue:44857", + "issue:43819|issue:45083", + "issue:37428|issue:43425", + "issue:43335|issue:43550", + "issue:43606|issue:45083" + ], + "prepared_review_unit_hash": "391218718bb8c247740b4efd9963e0a30109b5e7e3a07cf0cce28d249aca013a", + "component_budget": { + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 30, + "serialized_chars": 36907, + "estimated_input_tokens": 9227, + "estimated_eval_tokens": 18710 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly heterogeneous: the soft pairs are largely same-subsystem but not same-bug matches, so I reject them all. There is no strong duplicate canonical issue; if an anchor is needed, issue 44521 is the broadest and most active issue in the closest thematic area (multimodal apply_chat_template).", + "confidence": 0.86, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": "Issue 44521 has the broadest scope and the most traction among the only plausible subcluster (multimodal apply_chat_template regressions), so it is the best anchor even though the cluster is not a true duplicate set.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44898", + "right": "issue:45412", + "accept": false, + "reason": "Different bugs: image-resolution handling vs memory not being released on deletion." + }, + { + "left": "issue:44466", + "right": "issue:45127", + "accept": false, + "reason": "One is tied-weight serialization/device-dependent behavior; the other is LoRA merge collapse with extended vocab. Related theme, different failure." + }, + { + "left": "issue:43295", + "right": "issue:43873", + "accept": false, + "reason": "Processor API regression and quantization/offload behavior are unrelated code paths." + }, + { + "left": "issue:43550", + "right": "issue:43844", + "accept": false, + "reason": "Torch.compile/SDPA failure is not the same as a ZeRO-3 gradient growth issue." + }, + { + "left": "issue:43334", + "right": "issue:43525", + "accept": false, + "reason": "Both mention missing pad_token_id, but they are model-specific config failures and not clearly one shared bug." + }, + { + "left": "issue:44792", + "right": "issue:45563", + "accept": false, + "reason": "Janus image-generation test failure and a paged-generate warning are distinct issues." + }, + { + "left": "issue:43824", + "right": "issue:43975", + "accept": false, + "reason": "ImportError for a missing class and a detokenization regression are different failures." + }, + { + "left": "issue:42915", + "right": "issue:43824", + "accept": false, + "reason": "FineGrainedFP8Config failure is unrelated to a missing Qwen2.5-VL import." + }, + { + "left": "issue:43388", + "right": "issue:43526", + "accept": false, + "reason": "Different layers: metric gathering vs image-processor label reduction." + }, + { + "left": "issue:43450", + "right": "issue:43575", + "accept": false, + "reason": "Batched video shape bug and tensor-parallel OOM are not the same underlying problem." + }, + { + "left": "issue:43278", + "right": "issue:43856", + "accept": false, + "reason": "Training/eval dtype mismatch is unrelated to MoE memory usage." + }, + { + "left": "issue:42490", + "right": "issue:43493", + "accept": false, + "reason": "Local-path import warning and SigLIP2 implementation discrepancy are unrelated." + }, + { + "left": "issue:43975", + "right": "issue:45520", + "accept": false, + "reason": "Detokenization regression and flash_attn import-time KeyError are unrelated." + }, + { + "left": "issue:28218", + "right": "issue:35141", + "accept": false, + "reason": "Tokenizer spacing bug and embedding reinit after resize are different code paths." + }, + { + "left": "issue:43653", + "right": "issue:44112", + "accept": false, + "reason": "Special-token registration bug and a stale CI test are not duplicates." + }, + { + "left": "issue:38175", + "right": "issue:43493", + "accept": false, + "reason": "Both involve SigLIP2, but zero probabilities vs JAX discrepancy are not clearly the same concrete bug." + }, + { + "left": "issue:43526", + "right": "issue:46192", + "accept": false, + "reason": "Label reduction behavior and text_kwargs/return_tensors handling are unrelated." + }, + { + "left": "issue:44442", + "right": "issue:45198", + "accept": false, + "reason": "Different tokenizer/processor failures in different model families." + }, + { + "left": "issue:44514", + "right": "issue:44521", + "accept": false, + "reason": "Same API area, but one is a batched padding=False crash and the other is all-zero assistant masks; not the same concrete failure." + }, + { + "left": "issue:45520", + "right": "issue:46170", + "accept": false, + "reason": "Different import/platform dependency problems on different stacks." + }, + { + "left": "issue:38175", + "right": "issue:45412", + "accept": false, + "reason": "SigLIP2 probability issue and RT-DETR memory leak are unrelated." + }, + { + "left": "issue:44206", + "right": "issue:45584", + "accept": false, + "reason": "Unsupported center argument in feature extraction is unrelated to empty-transcription failure after align_special_tokens." + }, + { + "left": "issue:43827", + "right": "issue:43901", + "accept": false, + "reason": "Both are docs issues, but they cover different APIs and deprecations." + }, + { + "left": "issue:43262", + "right": "issue:43450", + "accept": false, + "reason": "Audio sampling-rate default bug and video batched-shape bug are unrelated." + }, + { + "left": "issue:43819", + "right": "issue:44112", + "accept": false, + "reason": "DAC latent reconstruction mismatch and a stale CI test are unrelated." + }, + { + "left": "issue:44610", + "right": "issue:44857", + "accept": false, + "reason": "Processor input-size mismatch and AMP/CUDA loss crash are different problems." + }, + { + "left": "issue:43819", + "right": "issue:45083", + "accept": false, + "reason": "Latent decoding mismatch and feature-length helper behavior are not the same bug." + }, + { + "left": "issue:37428", + "right": "issue:43425", + "accept": false, + "reason": "Missing import for a model class and Torch version incompatibility are unrelated." + }, + { + "left": "issue:43335", + "right": "issue:43550", + "accept": false, + "reason": "Sparse-layer config bug and torch.compile/SDPA failure are different issues." + }, + { + "left": "issue:43606", + "right": "issue:45083", + "accept": false, + "reason": "CPU-offload device mismatch and feature-length helper behavior are unrelated." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 17, + "review_unit_index": 25, + "review_unit_count": 50, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-428", + "nodes": [ + "issue:29127", + "issue:29942", + "issue:32090", + "issue:37428", + "issue:38175", + "issue:39692", + "issue:42490", + "issue:43278", + "issue:43335", + "issue:43504", + "issue:43526", + "issue:43550", + "issue:43606", + "issue:43611", + "issue:43653", + "issue:43720", + "issue:43819", + "issue:43825", + "issue:43854", + "issue:43931", + "issue:44112", + "issue:44206", + "issue:44351", + "issue:44360", + "issue:44448", + "issue:44493", + "issue:44561", + "issue:44743", + "issue:44779", + "issue:44855", + "issue:44857", + "issue:44977", + "issue:45072", + "issue:45083", + "issue:45127", + "issue:45200", + "issue:45229", + "issue:45245", + "issue:45290", + "issue:45310", + "issue:45381", + "issue:45478", + "issue:45588", + "issue:45593", + "issue:45823", + "issue:45859", + "issue:45920", + "issue:46192" + ], + "soft_pairs": [ + "issue:44857|issue:45593", + "issue:45290|issue:45381", + "issue:44448|issue:44743", + "issue:45127|issue:45245", + "issue:29942|issue:32090", + "issue:45200|issue:45588", + "issue:43278|issue:43825", + "issue:37428|issue:44351", + "issue:44206|issue:45823", + "issue:43335|issue:43653", + "issue:43611|issue:43931", + "issue:43550|issue:44112", + "issue:43819|issue:44493", + "issue:44855|issue:45083", + "issue:43335|issue:43504", + "issue:45588|issue:45859", + "issue:29127|issue:32090", + "issue:43335|issue:45083", + "issue:39692|issue:43526", + "issue:45310|issue:45478", + "issue:39692|issue:44493", + "issue:43720|issue:44112", + "issue:43335|issue:45072", + "issue:45920|issue:46192", + "issue:43335|issue:44112", + "issue:44360|issue:44561", + "issue:44448|issue:44779", + "issue:42490|issue:43854", + "issue:44977|issue:45229", + "issue:38175|issue:44493", + "issue:43606|issue:44857" + ], + "prepared_review_unit_hash": "85e61ef1745c23f4b1e4d9e5803f09b992e03a86c7403686d7e135e120e0de52", + "component_budget": { + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 31, + "serialized_chars": 37145, + "estimated_input_tokens": 9287, + "estimated_eval_tokens": 18830 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "Mostly unrelated singleton issues. There is one very similar pair (45310/45478) with matching titles, but the evidence is still soft, so the cluster should not be treated as a confirmed duplicate set.", + "confidence": 0.68, + "canonical_issue_reason": "Issue 45310 is the best representative for the shared Qwen3.5 Moe from_pretrained theme, but the evidence only supports a soft similarity, not a confirmed duplicate cluster.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 45310 is the strongest issue representative because it is earlier and matches 45478 most closely, but it should be treated as a likely topic match rather than a proven duplicate.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44857", + "right": "issue:45593", + "accept": false, + "reason": "Different failures: AMP/CUDA loss crash vs missing auxiliary losses." + }, + { + "left": "issue:45290", + "right": "issue:45381", + "accept": false, + "reason": "Tool-call chat-template crash and video vision_position_ids bug are unrelated." + }, + { + "left": "issue:44448", + "right": "issue:44743", + "accept": false, + "reason": "Pegasus output drift and Qwen3 recurrent-state reset affect different code paths." + }, + { + "left": "issue:45127", + "right": "issue:45245", + "accept": false, + "reason": "LoRA/vocab collapse and category-count limit are distinct bugs." + }, + { + "left": "issue:29942", + "right": "issue:32090", + "accept": false, + "reason": "Flash Attention test failures and NoneType GPU broadcast error are different problems." + }, + { + "left": "issue:45200", + "right": "issue:45588", + "accept": false, + "reason": "Gemma4 token-type handling and flash_attention s_aux=None crash are unrelated." + }, + { + "left": "issue:43278", + "right": "issue:43825", + "accept": false, + "reason": "BF16/FP32 dtype regression and pipeline task-support messaging do not share a code path." + }, + { + "left": "issue:37428", + "right": "issue:44351", + "accept": false, + "reason": "Both are import errors, but they involve different missing symbols and likely different fixes." + }, + { + "left": "issue:44206", + "right": "issue:45823", + "accept": false, + "reason": "Unsupported center argument in feature extractor and Gemma4 device mismatch are unrelated." + }, + { + "left": "issue:43335", + "right": "issue:43653", + "accept": false, + "reason": "Sparse-layer config bug and tokenizer special-token registration bug affect different components." + }, + { + "left": "issue:43611", + "right": "issue:43931", + "accept": false, + "reason": "base_model_prefix loading breakage and Qwen3-VL shape mismatch are not the same defect." + }, + { + "left": "issue:43550", + "right": "issue:44112", + "accept": false, + "reason": "torch.compile SDPA failure and CI device override failure are different issues." + }, + { + "left": "issue:43819", + "right": "issue:44493", + "accept": false, + "reason": "DAC from_latents mismatch and unexpected position-id warnings are unrelated model bugs." + }, + { + "left": "issue:44855", + "right": "issue:45083", + "accept": false, + "reason": "Python 3.13 import parsing failure and qwen3_omni_moe helper behavior are different problems." + }, + { + "left": "issue:43335", + "right": "issue:43504", + "accept": false, + "reason": "Sparse layer creation bug and BEiT legacy-field loading bug are unrelated." + }, + { + "left": "issue:45588", + "right": "issue:45859", + "accept": false, + "reason": "Both mention offload/flash attention, but the concrete failures differ." + }, + { + "left": "issue:29127", + "right": "issue:32090", + "accept": false, + "reason": "LayoutLMv3 error-message clarity and Trainer broadcast NoneType error are unrelated." + }, + { + "left": "issue:43335", + "right": "issue:45083", + "accept": false, + "reason": "SwitchTransformers config bug and qwen3_omni_moe helper bug are unrelated." + }, + { + "left": "issue:39692", + "right": "issue:43526", + "accept": false, + "reason": "SigLIP2 docs/example errors and BeitImageProcessorFast reduce_labels bug are different problems." + }, + { + "left": "issue:45310", + "right": "issue:45478", + "accept": false, + "reason": "Titles are near-identical, but the packet only provides soft similarity and no strong duplicate evidence." + }, + { + "left": "issue:39692", + "right": "issue:44493", + "accept": false, + "reason": "SigLIP2 documentation/model mismatch and unexpected position-id keys are unrelated." + }, + { + "left": "issue:43720", + "right": "issue:44112", + "accept": false, + "reason": "BitNet packed-weight loading and GraniteSpeech device-override CI issues are different." + }, + { + "left": "issue:43335", + "right": "issue:45072", + "accept": false, + "reason": "Sparse-layer construction and bfloat16 dtype mismatch are unrelated." + }, + { + "left": "issue:45920", + "right": "issue:46192", + "accept": false, + "reason": "Tokenizer wrong-token-IDs regression and return_tensors handling bug are different API issues." + }, + { + "left": "issue:44360", + "right": "issue:44561", + "accept": false, + "reason": "DSA indexer discussion and removal of is_torch_fx_available are unrelated." + }, + { + "left": "issue:44448", + "right": "issue:44779", + "accept": false, + "reason": "Pegasus output differences and DeepSeek tokenizer regression are different subsystems." + }, + { + "left": "issue:42490", + "right": "issue:43854", + "accept": false, + "reason": "Local-path import warning and GLM-4.7-Flash load failure are unrelated." + }, + { + "left": "issue:44977", + "right": "issue:45229", + "accept": false, + "reason": "Qwen3.5 flash-attention generation bug and Gemma4 CUDA OOM are unrelated." + }, + { + "left": "issue:38175", + "right": "issue:44493", + "accept": false, + "reason": "SigLIP2 zero-probabilities and position-id key warnings are unrelated." + }, + { + "left": "issue:43606", + "right": "issue:44857", + "accept": false, + "reason": "CPU offload device mismatch in bark-small and AMP crash in LwDetrImageLoss are unrelated." + } + ] + }, + "evaluator_result": { + "accept": false, + "feedback": "The summary is too strong for the packet. Although issue 45310 and 45478 have the highest similarity and matching titles, the packet marks that pair as deterministic_accept=false, so it does not support calling it a clear/true duplicate pair. There are no explicit links or stronger evidence establishing any duplicate cluster, so the claim of \"the only clear duplicate pair\" and the 0.97 confidence overstate the evidence. A more conservative read would be that the cluster is mostly singleton-like with one highly similar issue pair, not a confirmed duplicate." + }, + "evaluator_used": true, + "retried": true, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 17, + "review_unit_index": 26, + "review_unit_count": 50, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-428", + "nodes": [ + "issue:10000", + "issue:15354", + "issue:28218", + "issue:29942", + "issue:32090", + "issue:33357", + "issue:34689", + "issue:35141", + "issue:37428", + "issue:40444", + "issue:41720", + "issue:42490", + "issue:43278", + "issue:43329", + "issue:43381", + "issue:43412", + "issue:43526", + "issue:43550", + "issue:43575", + "issue:43701", + "issue:43723", + "issue:43873", + "issue:43927", + "issue:43976", + "issue:44448", + "issue:44462", + "issue:44554", + "issue:44561", + "issue:44568", + "issue:44779", + "issue:44857", + "issue:44977", + "issue:44987", + "issue:45072", + "issue:45083", + "issue:45198", + "issue:45229", + "issue:45372", + "issue:45406", + "issue:45440", + "issue:45459", + "issue:45491", + "issue:45563", + "issue:45588", + "issue:45698", + "issue:45712", + "issue:45803", + "issue:45823" + ], + "soft_pairs": [ + "issue:34689|issue:35141", + "issue:44857|issue:45072", + "issue:44448|issue:44977", + "issue:40444|issue:43526", + "issue:28218|issue:32090", + "issue:43701|issue:45083", + "issue:43723|issue:44554", + "issue:29942|issue:35141", + "issue:43550|issue:45083", + "issue:32090|issue:33357", + "issue:43873|issue:45229", + "issue:44561|issue:44779", + "issue:43927|issue:45698", + "issue:44462|issue:44561", + "issue:41720|issue:45229", + "issue:37428|issue:44987", + "issue:44977|issue:45563", + "issue:43278|issue:43329", + "issue:44568|issue:45491", + "issue:42490|issue:43575", + "issue:45372|issue:45588", + "issue:10000|issue:15354", + "issue:45198|issue:45803", + "issue:43412|issue:45588", + "issue:45372|issue:45406", + "issue:44857|issue:45588", + "issue:43976|issue:45440", + "issue:45459|issue:45712", + "issue:45588|issue:45823", + "issue:43278|issue:43381" + ], + "prepared_review_unit_hash": "dc37b8bed06cac499fef760da2a986dddc02b2f31e1755541a2fcf30e6442f0e", + "component_budget": { + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 30, + "serialized_chars": 36869, + "estimated_input_tokens": 9218, + "estimated_eval_tokens": 18692 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is highly heterogeneous: the pairs share broad themes like loading, tokenization, attention, or dtype issues, but they do not appear to be the same underlying bug/change. I would not merge any of the soft edges.", + "confidence": 0.94, + "canonical_issue_reason": "If a single representative is needed, issue:45698 is the broadest and most centrally discussed loading-related bug in the set (custom module resolution after save/load).", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45698 is the best single issue candidate because it has the strongest discussion/inbound-reference signal and a general save_pretrained/from_pretrained failure mode that is easier to use as an umbrella reference than the other, more specialized reports.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:34689", + "right": "issue:35141", + "accept": false, + "reason": "Different bugs: model loading for Llama 3.2 Vision vs embedding reinitialization after resizing tokens." + }, + { + "left": "issue:44857", + "right": "issue:45072", + "accept": false, + "reason": "Both involve dtype/AMP, but they hit different models and failure modes: LwDetr float16 CUDA crash vs bfloat16 inference mismatch." + }, + { + "left": "issue:44448", + "right": "issue:44977", + "accept": false, + "reason": "Different models and code paths: Pegasus v4/v5 output drift vs Qwen3.5 flash-attention generation failure." + }, + { + "left": "issue:40444", + "right": "issue:43526", + "accept": false, + "reason": "Different layers: Qwen2.5-VL iterable-dataset multi-image failure vs BeitImageProcessorFast reduce_labels bug." + }, + { + "left": "issue:28218", + "right": "issue:32090", + "accept": false, + "reason": "Tokenizer added-space behavior is unrelated to the Trainer _gpu_broadcast_one NoneType error." + }, + { + "left": "issue:43701", + "right": "issue:45083", + "accept": false, + "reason": "Checkpoint resume key mismatch is unrelated to the qwen3_omni_moe helper-length bug." + }, + { + "left": "issue:43723", + "right": "issue:44554", + "accept": false, + "reason": "AutoTokenizer v5 loading regression is unrelated to the MPS attention head-dimension correctness issue." + }, + { + "left": "issue:29942", + "right": "issue:35141", + "accept": false, + "reason": "Flash Attention 2 test failures are unrelated to post_init reinitializing resized embeddings." + }, + { + "left": "issue:43550", + "right": "issue:45083", + "accept": false, + "reason": "torch.compile/SDPA failure on Bamba-9B-v2 is not the same as a qwen3_omni_moe feature-length helper bug." + }, + { + "left": "issue:32090", + "right": "issue:33357", + "accept": false, + "reason": "Trainer broadcast type error and MacOS bus error on a CLIP model are different failures." + }, + { + "left": "issue:43873", + "right": "issue:45229", + "accept": false, + "reason": "Quantization/offloading behavior and Gemma4 multi-GPU CUDA OOM are distinct problems." + }, + { + "left": "issue:44561", + "right": "issue:44779", + "accept": false, + "reason": "Removal of is_torch_fx_available breaking trust_remote_code is unrelated to DeepSeek tokenizer correctness." + }, + { + "left": "issue:43927", + "right": "issue:45698", + "accept": false, + "reason": "Both involve save/load behavior, but one is custom token IDs in config and the other is wrong custom module resolution; different root causes." + }, + { + "left": "issue:44462", + "right": "issue:44561", + "accept": false, + "reason": "AutoTokenizer ignoring tokenizer.json is unrelated to the v5 removal of is_torch_fx_available." + }, + { + "left": "issue:41720", + "right": "issue:45229", + "accept": false, + "reason": "Qwen3 auto device-mapping cuda assert is not the same bug as Gemma4 multi-GPU OOM." + }, + { + "left": "issue:37428", + "right": "issue:44987", + "accept": false, + "reason": "Missing _flash_supports_window_size import and loading physical-intelligence/fast on 5.1.0 are not clearly the same concrete failure." + }, + { + "left": "issue:44977", + "right": "issue:45563", + "accept": false, + "reason": "Flash-attention generation corruption and a stale num_return_sequences warning are unrelated." + }, + { + "left": "issue:43278", + "right": "issue:43329", + "accept": false, + "reason": "Embedding dtype drift in eval mode is unrelated to the undefined helper variables in the video token-count path." + }, + { + "left": "issue:44568", + "right": "issue:45491", + "accept": false, + "reason": "Tokenizer add_special_tokens regression is unrelated to Gemma3 NaN embeddings from sliding-window attention." + }, + { + "left": "issue:42490", + "right": "issue:43575", + "accept": false, + "reason": "Import warning for trailing-slash local paths and tp-related OOM on Qwen2-57B are different issues." + }, + { + "left": "issue:45372", + "right": "issue:45588", + "accept": false, + "reason": "Gemma4 processor import failure from mistral_common is unrelated to flash_attention crashing on s_aux=None." + }, + { + "left": "issue:10000", + "right": "issue:15354", + "accept": false, + "reason": "German DistilBertModel issue and GeneratorExp/jit.script export failure are unrelated." + }, + { + "left": "issue:45198", + "right": "issue:45803", + "accept": false, + "reason": "Wav2Vec2 save_pretrained/tokenization failure is unrelated to the bare except swallowing errors in FuyuBatchFeature.convert_to_tensors()." + }, + { + "left": "issue:43412", + "right": "issue:45588", + "accept": false, + "reason": "Gemma3n executorch export problems are unrelated to the flash_attention s_aux=None crash." + }, + { + "left": "issue:45372", + "right": "issue:45406", + "accept": false, + "reason": "Both are Gemma4-related, but one is a processor import error and the other is a missing _tokenizer attribute in serve; different concrete bugs." + }, + { + "left": "issue:44857", + "right": "issue:45588", + "accept": false, + "reason": "LwDetr float16 CUDA crash is unrelated to the flash_attention s_aux=None AttributeError." + }, + { + "left": "issue:43976", + "right": "issue:45440", + "accept": false, + "reason": "Python-version compatibility breakage is unrelated to DeepseekV3MoE diverging from remote implementation." + }, + { + "left": "issue:45459", + "right": "issue:45712", + "accept": false, + "reason": "Hidden tokenizer errors from protobuf handling are unrelated to leftover dummy classes in dummy_pt_objects.py." + }, + { + "left": "issue:45588", + "right": "issue:45823", + "accept": false, + "reason": "flash_attention sink-less-model crash and Gemma4 PLE device mismatch under auto device map are different code paths." + }, + { + "left": "issue:43278", + "right": "issue:43381", + "accept": false, + "reason": "Embedding dtype regression in eval mode is unrelated to gradient checkpointing being disallowed in eval mode." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 17, + "review_unit_index": 27, + "review_unit_count": 50, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-428", + "nodes": [ + "issue:25251", + "issue:29127", + "issue:29942", + "issue:30064", + "issue:32090", + "issue:33290", + "issue:33357", + "issue:34567", + "issue:39692", + "issue:42490", + "issue:42915", + "issue:42994", + "issue:43065", + "issue:43066", + "issue:43257", + "issue:43278", + "issue:43381", + "issue:43421", + "issue:43611", + "issue:43653", + "issue:43701", + "issue:43756", + "issue:43824", + "issue:43873", + "issue:43994", + "issue:44186", + "issue:44206", + "issue:44265", + "issue:44279", + "issue:44361", + "issue:44442", + "issue:44448", + "issue:44462", + "issue:44466", + "issue:44479", + "issue:44610", + "issue:44743", + "issue:44871", + "issue:44977", + "issue:45198", + "issue:45245", + "issue:45372", + "issue:45412", + "issue:45440", + "issue:45459", + "issue:45491", + "issue:45563", + "issue:46121" + ], + "soft_pairs": [ + "issue:44448|issue:44466", + "issue:45245|issue:45563", + "issue:43653|issue:44743", + "issue:25251|issue:32090", + "issue:43065|issue:43994", + "issue:44206|issue:44479", + "issue:45372|issue:45459", + "issue:44186|issue:45198", + "issue:43065|issue:45412", + "issue:43611|issue:46121", + "issue:44871|issue:44977", + "issue:44361|issue:45198", + "issue:42915|issue:42994", + "issue:42490|issue:43873", + "issue:32090|issue:33290", + "issue:33357|issue:39692", + "issue:29942|issue:30064", + "issue:29127|issue:34567", + "issue:43994|issue:45412", + "issue:44265|issue:45198", + "issue:43066|issue:44462", + "issue:43756|issue:45440", + "issue:43257|issue:43278", + "issue:43421|issue:45491", + "issue:43653|issue:44442", + "issue:43381|issue:43701", + "issue:44610|issue:45412", + "issue:43824|issue:44279" + ], + "prepared_review_unit_hash": "25107c1da88ee4c6c20c70e063f7659e4b7fb1f3a40fbc575cfdced30ecfe85b", + "component_budget": { + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 28, + "serialized_chars": 36089, + "estimated_input_tokens": 9023, + "estimated_eval_tokens": 18302 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The cluster is a grab bag of unrelated Transformers issues; none of the soft pairs look like true duplicates or the same fix. I\u2019d use the core tokenizer-loading bug as the representative issue, but with low confidence because the set is highly heterogeneous.", + "confidence": 0.44, + "canonical_issue_reason": "issue:44462 is the clearest, most generally useful concrete bug report here: a core AutoTokenizer loading problem with a specific failure mode and broad relevance across models.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:44462 is the best representative issue because it is specific, reproducible, and centered on a core library code path rather than a model-specific edge case or an ancillary regression.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44448", + "right": "issue:44466", + "accept": false, + "reason": "Different problems: output divergence for a model checkpoint vs inconsistent lm_head.weight serialization/device-dependent saving." + }, + { + "left": "issue:45245", + "right": "issue:45563", + "accept": false, + "reason": "Unrelated failures: category cardinality error in preprocessing vs a stale warning in paged generate()." + }, + { + "left": "issue:43653", + "right": "issue:44743", + "accept": false, + "reason": "Different subsystems: BigBirdTokenizer special-token registration vs Qwen3.5 recurrent-state caching behavior." + }, + { + "left": "issue:25251", + "right": "issue:32090", + "accept": false, + "reason": "Top-k pipeline nesting bug is unrelated to Trainer broadcast errors involving NoneType." + }, + { + "left": "issue:43065", + "right": "issue:43994", + "accept": false, + "reason": "Sam3PixelDecoder dummy Conv2d issue is not the same as SigLIP2 bad outputs from AutoModel/pipeline loading." + }, + { + "left": "issue:44206", + "right": "issue:44479", + "accept": false, + "reason": "Feature extractor center-arg crash and Qwen VL video regression are different code paths and failure modes." + }, + { + "left": "issue:45372", + "right": "issue:45459", + "accept": false, + "reason": "Gemma 4 loading import error from mistral_common is distinct from tokenizer error masking when protobuf is missing." + }, + { + "left": "issue:44186", + "right": "issue:45198", + "accept": false, + "reason": "LayoutLMv2Tokenizer NER/padding crash is unrelated to Wav2Vec2 save_pretrained/tokenization failure." + }, + { + "left": "issue:43065", + "right": "issue:45412", + "accept": false, + "reason": "A dummy layer in Sam3PixelDecoder is not the same as RT-DETR models leaking memory after deletion." + }, + { + "left": "issue:43611", + "right": "issue:46121", + "accept": false, + "reason": "Base-model-prefix loading breakage in v5 is unrelated to convert_rope_params_to_dict rejecting a JSON-loaded list." + }, + { + "left": "issue:44871", + "right": "issue:44977", + "accept": false, + "reason": "Gemma eos_token_id mismatch and Qwen3.5 flash-attention generation issues are separate model/config problems." + }, + { + "left": "issue:44361", + "right": "issue:45198", + "accept": false, + "reason": "MLukeTokenizer AttributeError on tasks is not the same as Wav2Vec2 save/tokenization failure." + }, + { + "left": "issue:42915", + "right": "issue:42994", + "accept": false, + "reason": "FineGrainedFP8Config conversion failure and quantized model saving are different quantization workflows." + }, + { + "left": "issue:42490", + "right": "issue:43873", + "accept": false, + "reason": "Trailing-slash local model import warning is unrelated to offloading behavior with quantization." + }, + { + "left": "issue:32090", + "right": "issue:33290", + "accept": false, + "reason": "NoneType broadcast error in Trainer is unrelated to Adafactor/deepspeed OOM." + }, + { + "left": "issue:33357", + "right": "issue:39692", + "accept": false, + "reason": "MacOS bus error for a pretrained CLIP model is not the same as a documentation example with model/processor mismatch and quantization issues." + }, + { + "left": "issue:29942", + "right": "issue:30064", + "accept": false, + "reason": "Flash Attention 2 test failures and void segmentation-map processing are different failures in different areas." + }, + { + "left": "issue:29127", + "right": "issue:34567", + "accept": false, + "reason": "LayoutLMv3 error-message clarity is unrelated to TrainerState num_input_tokens_seen not updating." + }, + { + "left": "issue:43994", + "right": "issue:45412", + "accept": false, + "reason": "SigLIP2 nonsense outputs and RT-DETR memory retention are unrelated bugs." + }, + { + "left": "issue:44265", + "right": "issue:45198", + "accept": false, + "reason": "torch.export exportability with torch_compilable_check is not the same as Wav2Vec2 tokenization/save_pretrained failure." + }, + { + "left": "issue:43066", + "right": "issue:44462", + "accept": false, + "reason": "Wrong tokenizer decoder type in v5 and ignoring tokenizer.json from the repo are both tokenizer-related but not the same underlying defect." + }, + { + "left": "issue:43756", + "right": "issue:45440", + "accept": false, + "reason": "Smollm3 RoPE-layer dropping and DeepseekV3MoE divergence are separate architecture/implementation mismatches." + }, + { + "left": "issue:43257", + "right": "issue:43278", + "accept": false, + "reason": "Qwen3 MOE weight conversion with accelerate/deepspeed is unrelated to embedding dtype changing in evaluation." + }, + { + "left": "issue:43421", + "right": "issue:45491", + "accept": false, + "reason": "Runtime special-token post-processor staleness and Gemma3 NaN embeddings in mixed-length batches are different bugs." + }, + { + "left": "issue:43653", + "right": "issue:44442", + "accept": false, + "reason": "BigBirdTokenizer decode-output bug and AutoTokenizer failing to load FastSpeech2ConformerTokenizer are distinct tokenizer issues." + }, + { + "left": "issue:43381", + "right": "issue:43701", + "accept": false, + "reason": "Gradient checkpointing in eval mode and resume_from_checkpoint key mismatch do not share the same failure mechanism." + }, + { + "left": "issue:44610", + "right": "issue:45412", + "accept": false, + "reason": "Processor/model image-size mismatch is unrelated to RT-DETR memory not being released." + }, + { + "left": "issue:43824", + "right": "issue:44279", + "accept": false, + "reason": "Missing Qwen2_5_VLForConditionalGeneration import is a specific API availability bug, not the generic dependency issue described by the other report." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 17, + "review_unit_index": 28, + "review_unit_count": 50, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-428", + "nodes": [ + "issue:29127", + "issue:29942", + "issue:33357", + "issue:35141", + "issue:36010", + "issue:42490", + "issue:42915", + "issue:43262", + "issue:43334", + "issue:43381", + "issue:43441", + "issue:43526", + "issue:43531", + "issue:43575", + "issue:43646", + "issue:43701", + "issue:43723", + "issue:43761", + "issue:43931", + "issue:44062", + "issue:44077", + "issue:44079", + "issue:44186", + "issue:44206", + "issue:44387", + "issue:44442", + "issue:44464", + "issue:44479", + "issue:44625", + "issue:44743", + "issue:44811", + "issue:44857", + "issue:44977", + "issue:45003", + "issue:45072", + "issue:45103", + "issue:45245", + "issue:45290", + "issue:45372", + "issue:45440", + "issue:45479", + "issue:45491", + "issue:45584", + "issue:45588", + "issue:45593", + "issue:45834", + "issue:45859", + "issue:45907" + ], + "soft_pairs": [ + "issue:45440|issue:45593", + "issue:29127|issue:33357", + "issue:29942|issue:36010", + "issue:43761|issue:45859", + "issue:43526|issue:45834", + "issue:43526|issue:44186", + "issue:45103|issue:45372", + "issue:44625|issue:45584", + "issue:29127|issue:35141", + "issue:44442|issue:45072", + "issue:43334|issue:43531", + "issue:44977|issue:45245", + "issue:42915|issue:43931", + "issue:43526|issue:44857", + "issue:44464|issue:45491", + "issue:42490|issue:43381", + "issue:43575|issue:43701", + "issue:43262|issue:44206", + "issue:43575|issue:44387", + "issue:44743|issue:44811", + "issue:43646|issue:44077", + "issue:43441|issue:45588", + "issue:44062|issue:45907", + "issue:43761|issue:44079", + "issue:45003|issue:45440", + "issue:43723|issue:45479", + "issue:44479|issue:45290" + ], + "prepared_review_unit_hash": "37613528d34c6cf8bf92e37cf50bff71ee9a8a0dbd169f9dcfc33492c5d57937", + "component_budget": { + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 27, + "serialized_chars": 35861, + "estimated_input_tokens": 8966, + "estimated_eval_tokens": 18188 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "All proposed soft edges look like false positives: the issues span unrelated models, tokenizers, processors, generation paths, and utility bugs rather than the same underlying defect.", + "confidence": 0.96, + "canonical_issue_reason": "No canonical issue stands out because this is not a true duplicate cluster; the items cover many unrelated subsystems and symptoms.", + "canonical_pr_reason": null, + "best_issue_reason": "If forced to pick one representative issue, issue 43761 is a clear, well-scoped regression report, but it is not representative of the rest of the heterogeneous set.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45440", + "right": "issue:45593", + "accept": false, + "reason": "Both are MoE-related, but one is a model-implementation divergence and the other is an auxiliary-loss behavior change; different bugs." + }, + { + "left": "issue:29127", + "right": "issue:33357", + "accept": false, + "reason": "LayoutLMv3 error messaging vs. a CLIP Mac bus error are unrelated issues in different models and failure modes." + }, + { + "left": "issue:29942", + "right": "issue:36010", + "accept": false, + "reason": "FlashAttention 2 test failures and a GenerationMixin import error are different code paths and symptoms." + }, + { + "left": "issue:43761", + "right": "issue:45859", + "accept": false, + "reason": "Both are regressions, but CLIPVision hidden_states handling and Qwen3_5Moe CPU-offload rotary embedding are separate bugs." + }, + { + "left": "issue:43526", + "right": "issue:45834", + "accept": false, + "reason": "BEiT reduce_labels behavior and Kosmos2.5 OCR index errors affect different processors and failure mechanisms." + }, + { + "left": "issue:43526", + "right": "issue:44186", + "accept": false, + "reason": "These are different image/text processor bugs; they do not point to the same underlying defect." + }, + { + "left": "issue:45103", + "right": "issue:45372", + "accept": false, + "reason": "Auto-docstring crashes with future annotations is unrelated to the Gemma 4/ReasoningEffort import failure." + }, + { + "left": "issue:44625", + "right": "issue:45584", + "accept": false, + "reason": "Qwen3.5 num_labels propagation and Whisper empty-transcription generation are unrelated bugs." + }, + { + "left": "issue:29127", + "right": "issue:35141", + "accept": false, + "reason": "One is about clearer validation errors in LayoutLMv3; the other is output embedding reinitialization after resize." + }, + { + "left": "issue:44442", + "right": "issue:45072", + "accept": false, + "reason": "Tokenizer loading failure and bfloat16 dtype mismatch are different issues with no shared concrete code path." + }, + { + "left": "issue:43334", + "right": "issue:43531", + "accept": false, + "reason": "Both mention Qwen models, but one is a pad_token_id/config loading bug and the other is a sliding_window runtime issue." + }, + { + "left": "issue:44977", + "right": "issue:45245", + "accept": false, + "reason": "Flash-attention generation regression and the 2^24 category limit error are unrelated." + }, + { + "left": "issue:42915", + "right": "issue:43931", + "accept": false, + "reason": "FineGrainedFP8Config failure and Qwen3-VL weight-shape mismatch are separate loading problems." + }, + { + "left": "issue:43526", + "right": "issue:44857", + "accept": false, + "reason": "BEiT label reduction and LwDetrImageLoss AMP crash are unrelated vision bugs." + }, + { + "left": "issue:44464", + "right": "issue:45491", + "accept": false, + "reason": "Compiled-forward chunked generation inconsistency and Gemma3 NaNs with sliding-window attention are different problems." + }, + { + "left": "issue:42490", + "right": "issue:43381", + "accept": false, + "reason": "Local-path import warnings and gradient checkpointing being disallowed in eval mode are unrelated." + }, + { + "left": "issue:43575", + "right": "issue:43701", + "accept": false, + "reason": "Tensor-parallel OOM and resume_from_checkpoint key mismatch are unrelated training/loading bugs." + }, + { + "left": "issue:43262", + "right": "issue:44206", + "accept": false, + "reason": "Audio chat-template sample-rate defaulting and an unsupported feature-extractor arg are different APIs and bugs." + }, + { + "left": "issue:43575", + "right": "issue:44387", + "accept": false, + "reason": "Both involve memory, but one is load-time OOM with tensor parallelism and the other is int4 CUDA reserved-memory growth." + }, + { + "left": "issue:44743", + "right": "issue:44811", + "accept": false, + "reason": "Qwen3.5 recurrent-state reset and Whisper batch_decode skip_special_tokens handling are unrelated." + }, + { + "left": "issue:43646", + "right": "issue:44077", + "accept": false, + "reason": "Custom model initialization breakage and patchtsmixer post_init policy are not the same issue." + }, + { + "left": "issue:43441", + "right": "issue:45588", + "accept": false, + "reason": "Ministral-3 FlashAttention failure and sink-less-model s_aux=None crash are different flash-attention bugs." + }, + { + "left": "issue:44062", + "right": "issue:45907", + "accept": false, + "reason": "AddedToken special-argument duplication and list_repo_files revision forwarding are unrelated utility issues." + }, + { + "left": "issue:43761", + "right": "issue:44079", + "accept": false, + "reason": "CLIPVision hidden_states regression and ModelOutput key assignment are different output-handling bugs." + }, + { + "left": "issue:45003", + "right": "issue:45440", + "accept": false, + "reason": "Unsafe sys.modules access in modeling_utils and DeepseekV3MoE divergence are unrelated." + }, + { + "left": "issue:43723", + "right": "issue:45479", + "accept": false, + "reason": "AutoTokenizer loading failure and zero-loss single-label classification are different bug classes." + }, + { + "left": "issue:44479", + "right": "issue:45290", + "accept": false, + "reason": "Qwen video-input regression and apply_chat_template tool-call crashes are separate code paths." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 17, + "review_unit_index": 29, + "review_unit_count": 50, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-428", + "nodes": [ + "issue:25251", + "issue:28282", + "issue:33290", + "issue:34567", + "issue:34689", + "issue:35141", + "issue:36010", + "issue:36331", + "issue:42490", + "issue:42907", + "issue:43065", + "issue:43232", + "issue:43295", + "issue:43316", + "issue:43381", + "issue:43388", + "issue:43493", + "issue:43525", + "issue:43540", + "issue:43611", + "issue:43643", + "issue:43720", + "issue:43749", + "issue:43761", + "issue:43825", + "issue:43827", + "issue:43873", + "issue:43931", + "issue:44077", + "issue:44188", + "issue:44190", + "issue:44442", + "issue:44479", + "issue:44496", + "issue:44514", + "issue:44617", + "issue:44743", + "issue:44855", + "issue:44871", + "issue:44898", + "issue:44913", + "issue:44987", + "issue:45137", + "issue:45216", + "issue:45698", + "issue:45759", + "issue:45803", + "issue:46178" + ], + "soft_pairs": [ + "issue:43232|issue:43388", + "issue:44442|issue:44898", + "issue:33290|issue:34567", + "issue:33290|issue:34689", + "issue:43316|issue:43525", + "issue:33290|issue:35141", + "issue:42490|issue:44743", + "issue:43381|issue:43873", + "issue:43761|issue:44188", + "issue:44442|issue:45803", + "issue:34689|issue:43611", + "issue:44987|issue:46178", + "issue:43720|issue:44442", + "issue:44077|issue:44479", + "issue:43931|issue:45216", + "issue:42907|issue:45698", + "issue:43493|issue:43827", + "issue:34689|issue:36010", + "issue:43316|issue:45759", + "issue:33290|issue:36331", + "issue:43749|issue:43931", + "issue:43873|issue:44871", + "issue:43643|issue:44913", + "issue:44190|issue:44496", + "issue:44855|issue:45137", + "issue:43540|issue:44514", + "issue:25251|issue:28282", + "issue:43295|issue:43825", + "issue:43065|issue:44617" + ], + "prepared_review_unit_hash": "759edfabeb50972a369751d5e60860d6b63ae084478b931481be73a1703c8538", + "component_budget": { + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 29, + "serialized_chars": 36622, + "estimated_input_tokens": 9156, + "estimated_eval_tokens": 18568 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is highly heterogeneous: the items span unrelated bugs in loading, generation, training, quantization, vision/video, and docs. None of the soft pairs looks like the same concrete underlying issue, so I would not merge any of them.", + "confidence": 0.98, + "canonical_issue_reason": "issue:43611 is the broadest representative of the cluster\u2019s dominant loading/regression theme, but even it is not a true duplicate of the others; the rest are mostly separate bugs across different subsystems.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43611 is the best single representative because it targets a general model-loading regression with wider applicability than the more model-specific or peripheral reports.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43232", + "right": "issue:43388", + "accept": false, + "reason": "Generation kwargs after sync_gpus is unrelated to metric gathering dropping tuple labels." + }, + { + "left": "issue:44442", + "right": "issue:44898", + "accept": false, + "reason": "Tokenizer loading failure and Perceiver interpolation failure are different code paths." + }, + { + "left": "issue:33290", + "right": "issue:34567", + "accept": false, + "reason": "Deepspeed/Adafactor OOM is not the same as TrainerState token-count bookkeeping." + }, + { + "left": "issue:33290", + "right": "issue:34689", + "accept": false, + "reason": "Optimizer OOM and model-loading regression are different failure modes." + }, + { + "left": "issue:43316", + "right": "issue:43525", + "accept": false, + "reason": "Gemma3 config API mismatch and missing Llama4 pad_token_id are distinct config bugs." + }, + { + "left": "issue:33290", + "right": "issue:35141", + "accept": false, + "reason": "Training OOM does not match embedding reinitialization in post_init." + }, + { + "left": "issue:42490", + "right": "issue:44743", + "accept": false, + "reason": "Local-path import warning is unrelated to Qwen3.5 recurrent-state reset behavior." + }, + { + "left": "issue:43381", + "right": "issue:43873", + "accept": false, + "reason": "Gradient checkpointing in eval mode is separate from quantization offloading behavior." + }, + { + "left": "issue:43761", + "right": "issue:44188", + "accept": false, + "reason": "CLIP hidden_states regression and attention-kernel divergence are different issues." + }, + { + "left": "issue:44442", + "right": "issue:45803", + "accept": false, + "reason": "Tokenizer loading failure is not the same as a bare-except bug in tensor conversion." + }, + { + "left": "issue:34689", + "right": "issue:43611", + "accept": false, + "reason": "Both are loading-related, but they concern different regressions and different triggering conditions/models." + }, + { + "left": "issue:44987", + "right": "issue:46178", + "accept": false, + "reason": "Model loading failure for a specific repo is unrelated to backend mapping missing tensorflow_text." + }, + { + "left": "issue:43720", + "right": "issue:44442", + "accept": false, + "reason": "Packed-weight unpacking in BitNet is unrelated to AutoTokenizer failing for FastSpeech2ConformerTokenizer." + }, + { + "left": "issue:44077", + "right": "issue:44479", + "accept": false, + "reason": "Optional post_init enforcement and video-input regression are separate changes." + }, + { + "left": "issue:43931", + "right": "issue:45216", + "accept": false, + "reason": "Weight-shape mismatch at load time is different from save_pretrained producing a bad checkpoint." + }, + { + "left": "issue:42907", + "right": "issue:45698", + "accept": false, + "reason": "Saving dequantized models and loading the wrong custom module after save_pretrained are distinct bugs." + }, + { + "left": "issue:43493", + "right": "issue:43827", + "accept": false, + "reason": "Implementation discrepancy in SigLIP2 is unrelated to stale docs referencing pipeline()." + }, + { + "left": "issue:34689", + "right": "issue:36010", + "accept": false, + "reason": "Model-loading regression and GenerationMixin import error are different API breakages." + }, + { + "left": "issue:43316", + "right": "issue:45759", + "accept": false, + "reason": "Config API discrepancy is not the same as failing to unwrap text_config in from_config." + }, + { + "left": "issue:33290", + "right": "issue:36331", + "accept": false, + "reason": "OOM in DeepSpeed/Adafactor does not match a compute_loss signature break." + }, + { + "left": "issue:43749", + "right": "issue:43931", + "accept": false, + "reason": "FSDP CPU-efficient loading and Qwen3-VL weight mismatch are both loading-related but not the same bug." + }, + { + "left": "issue:43873", + "right": "issue:44871", + "accept": false, + "reason": "Quantization offloading is unrelated to eos_token_id configuration inconsistency." + }, + { + "left": "issue:43643", + "right": "issue:44913", + "accept": false, + "reason": "trust_remote_code missing fields and GPTNeoX rotary_pct reload behavior are separate config issues." + }, + { + "left": "issue:44190", + "right": "issue:44496", + "accept": false, + "reason": "Dataset loading in an example script is unrelated to an unrecognized model config." + }, + { + "left": "issue:44855", + "right": "issue:45137", + "accept": false, + "reason": "Python 3.13 TorchScript parsing and ZeRO3 deque underflow are different runtime failures." + }, + { + "left": "issue:43540", + "right": "issue:44514", + "accept": false, + "reason": "Video-input processing in Qwen3OmniMoe is unrelated to batched chat-template padding crash." + }, + { + "left": "issue:25251", + "right": "issue:28282", + "accept": false, + "reason": "Pipeline top_k nesting bug and missing PyTorch import are entirely unrelated." + }, + { + "left": "issue:43295", + "right": "issue:43825", + "accept": false, + "reason": "processor.tokenizer regression is not the same as an error-message/doc wording bug about translation tasks." + }, + { + "left": "issue:43065", + "right": "issue:44617", + "accept": false, + "reason": "Dummy Conv2d in Sam3PixelDecoder and Sam3Video OOM are different defects." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 17, + "review_unit_index": 30, + "review_unit_count": 50, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-428", + "nodes": [ + "issue:16998", + "issue:22355", + "issue:32090", + "issue:35141", + "issue:41950", + "issue:42617", + "issue:43295", + "issue:43316", + "issue:43388", + "issue:43404", + "issue:43441", + "issue:43452", + "issue:43475", + "issue:43504", + "issue:43550", + "issue:43572", + "issue:43575", + "issue:43606", + "issue:43646", + "issue:43673", + "issue:43688", + "issue:43701", + "issue:43716", + "issue:43742", + "issue:43761", + "issue:44077", + "issue:44162", + "issue:44242", + "issue:44479", + "issue:44496", + "issue:44534", + "issue:44617", + "issue:44734", + "issue:44743", + "issue:44792", + "issue:44871", + "issue:44877", + "issue:44913", + "issue:45127", + "issue:45200", + "issue:45406", + "issue:45563", + "issue:45584", + "issue:45698", + "issue:45803", + "issue:46050", + "issue:46170", + "issue:46178" + ], + "soft_pairs": [ + "issue:43316|issue:43716", + "issue:43316|issue:44877", + "issue:44162|issue:44792", + "issue:43550|issue:43701", + "issue:44077|issue:44534", + "issue:43606|issue:43701", + "issue:45803|issue:46170", + "issue:43404|issue:43441", + "issue:16998|issue:22355", + "issue:44871|issue:45584", + "issue:42617|issue:43575", + "issue:43673|issue:44743", + "issue:43575|issue:43742", + "issue:43388|issue:43688", + "issue:41950|issue:43475", + "issue:43761|issue:44479", + "issue:43295|issue:44871", + "issue:43452|issue:44496", + "issue:45127|issue:45200", + "issue:44871|issue:46050", + "issue:43646|issue:44913", + "issue:43572|issue:46178", + "issue:43504|issue:45698", + "issue:44734|issue:45406", + "issue:32090|issue:35141", + "issue:44617|issue:45584", + "issue:45563|issue:45584", + "issue:43761|issue:44242" + ], + "prepared_review_unit_hash": "21a990489c997cac759d47a7a392da335992600365c86b1872f5798ff2212148", + "component_budget": { + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 28, + "serialized_chars": 36076, + "estimated_input_tokens": 9019, + "estimated_eval_tokens": 18294 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is heterogeneous: the issues span unrelated loading/config regressions, model-specific bugs, and serving/training edge cases. I don\u2019t see a single underlying duplicate bug or change set tying them together.", + "confidence": 0.96, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43316", + "right": "issue:43716", + "accept": false, + "reason": "Both are Gemma-related API complaints, but one is a config discrepancy and the other is an image/text dtype mismatch; different code paths." + }, + { + "left": "issue:43316", + "right": "issue:44877", + "accept": false, + "reason": "Config/API mismatch in Gemma3TextConfig vs strict config loading for granite_speech are unrelated failures." + }, + { + "left": "issue:44162", + "right": "issue:44792", + "accept": false, + "reason": "ESM2 training/runtime breakage and a Janus image-generation test failure are different model pipelines and bugs." + }, + { + "left": "issue:43550", + "right": "issue:43701", + "accept": false, + "reason": "torch.compile/SDPA failure in Bamba is unrelated to resume_from_checkpoint key mismatches." + }, + { + "left": "issue:44077", + "right": "issue:44534", + "accept": false, + "reason": "Optional post_init enforcement in patchtsmixer and non-persistent buffer junk in v5 are distinct initialization issues." + }, + { + "left": "issue:43606", + "right": "issue:43701", + "accept": false, + "reason": "CPU offload device mismatch and checkpoint key mismatch are separate runtime/loading problems." + }, + { + "left": "issue:45803", + "right": "issue:46170", + "accept": false, + "reason": "Bare except swallowing errors in FuyuBatchFeature is unrelated to Windows ROCm torch.distributed import/type issues." + }, + { + "left": "issue:43404", + "right": "issue:43441", + "accept": false, + "reason": "LM head tying bug in Mistral3 and FlashAttention failure in Ministral-3 are different model-specific regressions." + }, + { + "left": "issue:16998", + "right": "issue:22355", + "accept": false, + "reason": "model_max_length discussion for DeBERTa-V3 is unrelated to the missing transformers.onnx module import error." + }, + { + "left": "issue:44871", + "right": "issue:45584", + "accept": false, + "reason": "Gemma eos_token_id config inconsistency and Whisper empty-transcription failure are unrelated generation bugs." + }, + { + "left": "issue:42617", + "right": "issue:43575", + "accept": false, + "reason": "3d_parallel.py execution failure and Qwen tensor-parallel OOM are different symptoms with different causes." + }, + { + "left": "issue:43673", + "right": "issue:44743", + "accept": false, + "reason": "Chunked-prefill cache missing and recurrent state reset in modular_qwen3_5 are different cache-related defects." + }, + { + "left": "issue:43575", + "right": "issue:43742", + "accept": false, + "reason": "Tensor-parallel OOM for a large Qwen model is unrelated to a MobileLLM config loading KeyError." + }, + { + "left": "issue:43388", + "right": "issue:43688", + "accept": false, + "reason": "Metric gathering label truncation and auxiliary-loss normalization are separate training/eval bugs." + }, + { + "left": "issue:41950", + "right": "issue:43475", + "accept": false, + "reason": "Video-classification processor lookup and SAM3 video missing attribute are different video-model issues." + }, + { + "left": "issue:43761", + "right": "issue:44479", + "accept": false, + "reason": "CLIPVision hidden_states regression and Qwen video-input regression affect different architectures and code paths." + }, + { + "left": "issue:43295", + "right": "issue:44871", + "accept": false, + "reason": "processor.tokenizer regression in custom code is unrelated to Gemma eos_token_id list-vs-scalar mismatch." + }, + { + "left": "issue:43452", + "right": "issue:44496", + "accept": false, + "reason": "GGUF loading breakage and unrecognized-model/config model_type errors are both loading-related but not the same bug." + }, + { + "left": "issue:45127", + "right": "issue:45200", + "accept": false, + "reason": "LoRA merge collapse with tied embeddings and missing mm_token_type_ids default are distinct model fine-tuning problems." + }, + { + "left": "issue:44871", + "right": "issue:46050", + "accept": false, + "reason": "Gemma eos_token configuration mismatch and quantized save/load breakage are unrelated persistence issues." + }, + { + "left": "issue:43646", + "right": "issue:44913", + "accept": false, + "reason": "Custom model initialization breakage in v5 and rotary_pct reload regression are different config/init bugs." + }, + { + "left": "issue:43572", + "right": "issue:46178", + "accept": false, + "reason": "StableLmConfig pad_token_idx regression and tensorflow_text backend mapping error are unrelated." + }, + { + "left": "issue:43504", + "right": "issue:45698", + "accept": false, + "reason": "Legacy field preventing Beit preset load and wrong custom module after save/load are different serialization issues." + }, + { + "left": "issue:44734", + "right": "issue:45406", + "accept": false, + "reason": "KV-cache continuation tensor indexing crash and Gemma4Processor missing _tokenizer are separate serve bugs." + }, + { + "left": "issue:32090", + "right": "issue:35141", + "accept": false, + "reason": "GPU broadcast NoneType error in Trainer and output embedding reinit after resize are unrelated training/model-init issues." + }, + { + "left": "issue:44617", + "right": "issue:45584", + "accept": false, + "reason": "SAM3 video OOM and Whisper empty-transcription regression do not share a root cause." + }, + { + "left": "issue:45563", + "right": "issue:45584", + "accept": false, + "reason": "A stale warning in paged generate() is different from Whisper failing on empty transcription after special-token alignment." + }, + { + "left": "issue:43761", + "right": "issue:44242", + "accept": false, + "reason": "CLIPVision forward regression and load-balancing loss omission are separate model behaviors." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 17, + "review_unit_index": 31, + "review_unit_count": 50, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-428", + "nodes": [ + "issue:35141", + "issue:41720", + "issue:42490", + "issue:43065", + "issue:43335", + "issue:43525", + "issue:43531", + "issue:43540", + "issue:43550", + "issue:43644", + "issue:43827", + "issue:43906", + "issue:43931", + "issue:43976", + "issue:43994", + "issue:44060", + "issue:44077", + "issue:44079", + "issue:44112", + "issue:44190", + "issue:44462", + "issue:44464", + "issue:44479", + "issue:44521", + "issue:44534", + "issue:44610", + "issue:44734", + "issue:44737", + "issue:44792", + "issue:44855", + "issue:44913", + "issue:44933", + "issue:44964", + "issue:44977", + "issue:45127", + "issue:45245", + "issue:45290", + "issue:45357", + "issue:45362", + "issue:45381", + "issue:45563", + "issue:45588", + "issue:45600", + "issue:45803", + "issue:45907", + "issue:45987", + "issue:46082", + "issue:46121" + ], + "soft_pairs": [ + "issue:43525|issue:43531", + "issue:44792|issue:44977", + "issue:43931|issue:44060", + "issue:44479|issue:44734", + "issue:44464|issue:45563", + "issue:44534|issue:44913", + "issue:43906|issue:45290", + "issue:44190|issue:44964", + "issue:44521|issue:45381", + "issue:43827|issue:43994", + "issue:44792|issue:45600", + "issue:41720|issue:45245", + "issue:45803|issue:45987", + "issue:43065|issue:44610", + "issue:44855|issue:45588", + "issue:35141|issue:42490", + "issue:44933|issue:45362", + "issue:43335|issue:44855", + "issue:44077|issue:44079", + "issue:43976|issue:44855", + "issue:43550|issue:44190", + "issue:43644|issue:44913", + "issue:43540|issue:44734", + "issue:45127|issue:45357", + "issue:44737|issue:45803", + "issue:44462|issue:45907", + "issue:44112|issue:45600", + "issue:43335|issue:44190", + "issue:46082|issue:46121" + ], + "prepared_review_unit_hash": "7752f49269afdf16ecf863625f2503bfa6d44d1c938c35ca213b57ef31df7eee", + "component_budget": { + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 29, + "serialized_chars": 36463, + "estimated_input_tokens": 9116, + "estimated_eval_tokens": 18488 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The set is a grab bag of unrelated issues; none of the soft pairs look like the same underlying bug/change, so all soft edges should be rejected. There are no PRs in the cluster.", + "confidence": 0.98, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43525", + "right": "issue:43531", + "accept": false, + "reason": "Different model/config bugs: missing pad_token_id vs sliding_window handling." + }, + { + "left": "issue:44792", + "right": "issue:44977", + "accept": false, + "reason": "Unrelated failures: janus test failure vs Qwen3.5 flash-attention generation." + }, + { + "left": "issue:43931", + "right": "issue:44060", + "accept": false, + "reason": "Both Qwen-related, but different problems: load-time weight shape mismatch vs tied-weights warning." + }, + { + "left": "issue:44479", + "right": "issue:44734", + "accept": false, + "reason": "Different code paths: multimodal video regression vs serve KV-cache indexing." + }, + { + "left": "issue:44464", + "right": "issue:45563", + "accept": false, + "reason": "Chunked generation inconsistency is not the same as a stale warning about num_return_sequences." + }, + { + "left": "issue:44534", + "right": "issue:44913", + "accept": false, + "reason": "Different bugs: non-persistent buffer init vs GPTNeoX config reload behavior." + }, + { + "left": "issue:43906", + "right": "issue:45290", + "accept": false, + "reason": "Insufficient overlap; one is an isolated reproduction of another issue, the other is a chat template tool-call crash." + }, + { + "left": "issue:44190", + "right": "issue:44964", + "accept": false, + "reason": "Dataset-loading example failure vs model-loading failure; different subsystems." + }, + { + "left": "issue:44521", + "right": "issue:45381", + "accept": false, + "reason": "Both multimodal, but assistant mask generation and vision_position_ids are different bugs." + }, + { + "left": "issue:43827", + "right": "issue:43994", + "accept": false, + "reason": "Docs referencing removed pipeline APIs is unrelated to SigLIP inference quality." + }, + { + "left": "issue:44792", + "right": "issue:45600", + "accept": false, + "reason": "Test failure in a model path vs CI break from removed auto_mappings entries." + }, + { + "left": "issue:41720", + "right": "issue:45245", + "accept": false, + "reason": "CUDA assert on Qwen3 auto device mapping is unrelated to the 2^24 category limit error." + }, + { + "left": "issue:45803", + "right": "issue:45987", + "accept": false, + "reason": "Different behaviors: exception handling in Fuyu batching vs cache length return type." + }, + { + "left": "issue:43065", + "right": "issue:44610", + "accept": false, + "reason": "Dummy Conv2d in Sam3PixelDecoder is unrelated to OmDet-Turbo input size mismatch." + }, + { + "left": "issue:44855", + "right": "issue:45588", + "accept": false, + "reason": "Python 3.13 import/parse error is unrelated to flash_attention s_aux=None crash." + }, + { + "left": "issue:35141", + "right": "issue:42490", + "accept": false, + "reason": "Embedding reinit on resize is unrelated to path-trailing-slash import warning." + }, + { + "left": "issue:44933", + "right": "issue:45362", + "accept": false, + "reason": "Missing import from image_utils is unrelated to Qwen3.5 chat crash." + }, + { + "left": "issue:43335", + "right": "issue:44855", + "accept": false, + "reason": "SwitchTransformers sparse-layer creation bug is unrelated to Python 3.13 indentation/parsing." + }, + { + "left": "issue:44077", + "right": "issue:44079", + "accept": false, + "reason": "Different object-model issues: optional post_init vs ModelOutput key assignment." + }, + { + "left": "issue:43976", + "right": "issue:44855", + "accept": false, + "reason": "Python version incompatibility at 5.1.0 is a different issue than a decorator/comment parsing failure." + }, + { + "left": "issue:43550", + "right": "issue:44190", + "accept": false, + "reason": "torch.compile/SDPA model failure is unrelated to a local dataset loading problem." + }, + { + "left": "issue:43644", + "right": "issue:44913", + "accept": false, + "reason": "Non-persistent buffer initialization bug is unrelated to GPTNeoX config persistence." + }, + { + "left": "issue:43540", + "right": "issue:44734", + "accept": false, + "reason": "Video-input ValueError in Qwen3OmniMoe is unrelated to serve KV-cache tensor indexing." + }, + { + "left": "issue:45127", + "right": "issue:45357", + "accept": false, + "reason": "LoRA merge with extended vocab is unrelated to incorrect visual encoder keys on save_pretrained." + }, + { + "left": "issue:44737", + "right": "issue:45803", + "accept": false, + "reason": "XLNet device placement bug is unrelated to broad exception swallowing in Fuyu batching." + }, + { + "left": "issue:44462", + "right": "issue:45907", + "accept": false, + "reason": "AutoTokenizer ignoring tokenizer.json is unrelated to list_repo_files not forwarding revision." + }, + { + "left": "issue:44112", + "right": "issue:45600", + "accept": false, + "reason": "Both are CI-related, but one is a stale GraniteSpeech test and the other is a broken auto_mappings reference." + }, + { + "left": "issue:43335", + "right": "issue:44190", + "accept": false, + "reason": "SwitchTransformers config bug is unrelated to dataset loading in the training script." + }, + { + "left": "issue:46082", + "right": "issue:46121", + "accept": false, + "reason": "Llama head_dim validation and rope-param JSON list handling are different configuration validation bugs." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 17, + "review_unit_index": 32, + "review_unit_count": 50, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-428", + "nodes": [ + "issue:36246", + "issue:38175", + "issue:38617", + "issue:39692", + "issue:41628", + "issue:41950", + "issue:42175", + "issue:42490", + "issue:42907", + "issue:43065", + "issue:43295", + "issue:43404", + "issue:43441", + "issue:43479", + "issue:43526", + "issue:43531", + "issue:43643", + "issue:43650", + "issue:43854", + "issue:43901", + "issue:43976", + "issue:44016", + "issue:44062", + "issue:44112", + "issue:44186", + "issue:44190", + "issue:44220", + "issue:44246", + "issue:44279", + "issue:44496", + "issue:44561", + "issue:44683", + "issue:44734", + "issue:44938", + "issue:44964", + "issue:44987", + "issue:45003", + "issue:45042", + "issue:45216", + "issue:45245", + "issue:45276", + "issue:45335", + "issue:45375", + "issue:45446", + "issue:45542", + "issue:45584", + "issue:45600", + "issue:45632" + ], + "soft_pairs": [ + "issue:44938|issue:45542", + "issue:39692|issue:44190", + "issue:43479|issue:45375", + "issue:45276|issue:45584", + "issue:43295|issue:45276", + "issue:36246|issue:41628", + "issue:43650|issue:44220", + "issue:45245|issue:45584", + "issue:41950|issue:44016", + "issue:44186|issue:44734", + "issue:36246|issue:38617", + "issue:42175|issue:44279", + "issue:44683|issue:45446", + "issue:44246|issue:45003", + "issue:43404|issue:45216", + "issue:44016|issue:44220", + "issue:38175|issue:43065", + "issue:43643|issue:44496", + "issue:44734|issue:45632", + "issue:43854|issue:44964", + "issue:44964|issue:45600", + "issue:44561|issue:45632", + "issue:44112|issue:44964", + "issue:43976|issue:44987", + "issue:42907|issue:43441", + "issue:42490|issue:45632", + "issue:43531|issue:45600", + "issue:43295|issue:45335", + "issue:43526|issue:43901", + "issue:44496|issue:45042", + "issue:44016|issue:44062" + ], + "prepared_review_unit_hash": "5ae5dc67ee6c80ff190e2d8f064a55eb7be17438d1b67f1902c63307126a99da", + "component_budget": { + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 31, + "serialized_chars": 36923, + "estimated_input_tokens": 9231, + "estimated_eval_tokens": 18718 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is heterogeneous: most items are unrelated model-loading/import/config regressions across different models and subsystems. I did not find any soft pair that looks like the same concrete bug/change, so all soft edges are rejected.", + "confidence": 0.94, + "canonical_issue_reason": "Issue 44561 is the broadest, most cross-cutting regression-style report in the set: a v5 compatibility break impacting trust_remote_code models. It is a better anchor than the many model-specific import/load/config complaints here.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 44561 is the best representative of the cluster\u2019s main theme (Transformers v5 regressions affecting remote-code/model-loading paths) and has the clearest general duplicate-target potential.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44938", + "right": "issue:45542", + "accept": false, + "reason": "Different failure domains: Python 3.14 import/load issue vs missing TensorFlow backend when only tensorboard is installed." + }, + { + "left": "issue:39692", + "right": "issue:44190", + "accept": false, + "reason": "Different problems: SigLIP2 docs/example errors vs local dataset loading in a training script." + }, + { + "left": "issue:43479", + "right": "issue:45375", + "accept": false, + "reason": "Both are config-related, but they affect different models and different fields; not the same underlying bug." + }, + { + "left": "issue:45276", + "right": "issue:45584", + "accept": false, + "reason": "Unrelated code paths: token embedding resize behavior in Gemma4 vs Whisper generation after special-token alignment." + }, + { + "left": "issue:43295", + "right": "issue:45276", + "accept": false, + "reason": "Different regressions: processor/tokenizer API behavior vs embedding resize propagation." + }, + { + "left": "issue:36246", + "right": "issue:41628", + "accept": false, + "reason": "Both are import errors, but for different symbols and different modules; not the same bug." + }, + { + "left": "issue:43650", + "right": "issue:44220", + "accept": false, + "reason": "One is a placeholder/data request issue; the other is a specific audio feature extraction bug." + }, + { + "left": "issue:45245", + "right": "issue:45584", + "accept": false, + "reason": "Category-count runtime error is unrelated to Whisper empty-transcription generation failure." + }, + { + "left": "issue:41950", + "right": "issue:44016", + "accept": false, + "reason": "Video-classification processor lookup bug vs notebook syntax error; no shared code-path." + }, + { + "left": "issue:44186", + "right": "issue:44734", + "accept": false, + "reason": "Tokenizer/padding crash and server KV-cache indexing crash are different components and failures." + }, + { + "left": "issue:36246", + "right": "issue:38617", + "accept": false, + "reason": "Both are import-related but for different missing names and different areas of the package." + }, + { + "left": "issue:42175", + "right": "issue:44279", + "accept": false, + "reason": "Packaging/backends dependency complaint vs a generic dependency issue; too vague and not the same concrete problem." + }, + { + "left": "issue:44683", + "right": "issue:45446", + "accept": false, + "reason": "Both touch flex_attention, but one is a compiled-kernel failure on newer torch and the other is a version-check/import guard bug." + }, + { + "left": "issue:44246", + "right": "issue:45003", + "accept": false, + "reason": "Import performance regression and unsafe sys.modules access are not the same concrete bug." + }, + { + "left": "issue:43404", + "right": "issue:45216", + "accept": false, + "reason": "Different models and failures: untied lm_head weights vs incorrect save_pretrained checkpoint serialization." + }, + { + "left": "issue:44016", + "right": "issue:44220", + "accept": false, + "reason": "Notebook syntax error is unrelated to audio feature extraction code." + }, + { + "left": "issue:38175", + "right": "issue:43065", + "accept": false, + "reason": "Different model families and different issues: SigLIP2 zero probabilities vs SAM3 decoder dummy layer." + }, + { + "left": "issue:43643", + "right": "issue:44496", + "accept": false, + "reason": "Both involve config/model loading, but one is missing fields with trust_remote_code and the other is an unrecognized model_type error; distinct root causes." + }, + { + "left": "issue:44734", + "right": "issue:45632", + "accept": false, + "reason": "Server tensor indexing crash vs local trust_remote_code cache-path collision are different failure modes." + }, + { + "left": "issue:43854", + "right": "issue:44964", + "accept": false, + "reason": "Different models and symptoms: GLM unit-test loading issue vs Phi-4 multimodal loading issue." + }, + { + "left": "issue:44964", + "right": "issue:45600", + "accept": false, + "reason": "Model loading failure for Phi-4 multimodal is unrelated to stale auto_mappings references and CI breakage." + }, + { + "left": "issue:44561", + "right": "issue:45632", + "accept": false, + "reason": "Both mention trust_remote_code, but one is a v5 API removal regression and the other is a cache-path collision; not the same bug." + }, + { + "left": "issue:44112", + "right": "issue:44964", + "accept": false, + "reason": "CI test flake in GraniteSpeech is unrelated to Phi-4 multimodal model loading." + }, + { + "left": "issue:43976", + "right": "issue:44987", + "accept": false, + "reason": "General Python-version compatibility complaint vs a specific model-loading failure; not the same underlying issue." + }, + { + "left": "issue:42907", + "right": "issue:43441", + "accept": false, + "reason": "Different Ministral issues: saving dequantized models vs FlashAttention failure." + }, + { + "left": "issue:42490", + "right": "issue:45632", + "accept": false, + "reason": "Both are local-path related, but one is a warning on trailing slashes and the other is a cache collision for same leaf names." + }, + { + "left": "issue:43531", + "right": "issue:45600", + "accept": false, + "reason": "Qwen3-MoE sliding-window behavior vs Sam3LiteText auto-mapping removal are unrelated." + }, + { + "left": "issue:43295", + "right": "issue:45335", + "accept": false, + "reason": "Both involve resize_token_embeddings, but on different architectures and different propagation bugs." + }, + { + "left": "issue:43526", + "right": "issue:43901", + "accept": false, + "reason": "Image processor label reduction bug vs text-classification docs/behavior mismatch are unrelated." + }, + { + "left": "issue:44496", + "right": "issue:45042", + "accept": false, + "reason": "Unrecognized model config loading vs PIL image processors requiring torchvision are different code paths." + }, + { + "left": "issue:44016", + "right": "issue:44062", + "accept": false, + "reason": "Notebook syntax error and AddedToken keyword collision are unrelated." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 17, + "review_unit_index": 33, + "review_unit_count": 50, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-428", + "nodes": [ + "issue:38175", + "issue:39692", + "issue:41950", + "issue:43295", + "issue:43299", + "issue:43335", + "issue:43381", + "issue:43450", + "issue:43479", + "issue:43525", + "issue:43577", + "issue:43650", + "issue:43673", + "issue:43824", + "issue:43854", + "issue:43901", + "issue:43976", + "issue:44075", + "issue:44242", + "issue:44279", + "issue:44297", + "issue:44336", + "issue:44351", + "issue:44368", + "issue:44462", + "issue:44496", + "issue:44521", + "issue:44617", + "issue:44683", + "issue:44779", + "issue:44933", + "issue:45020", + "issue:45081", + "issue:45200", + "issue:45216", + "issue:45278", + "issue:45325", + "issue:45335", + "issue:45357", + "issue:45405", + "issue:45542", + "issue:45584", + "issue:45593", + "issue:45759", + "issue:45812", + "issue:45823", + "issue:46132" + ], + "soft_pairs": [ + "issue:43673|issue:44336", + "issue:43335|issue:45593", + "issue:43824|issue:44933", + "issue:44075|issue:44368", + "issue:43299|issue:45216", + "issue:43976|issue:44279", + "issue:43381|issue:45593", + "issue:43901|issue:44496", + "issue:44496|issue:45759", + "issue:45335|issue:45584", + "issue:43479|issue:43525", + "issue:44521|issue:45325", + "issue:44242|issue:45200", + "issue:43299|issue:45357", + "issue:44351|issue:45278", + "issue:43577|issue:45216", + "issue:41950|issue:45020", + "issue:43976|issue:45405", + "issue:38175|issue:44617", + "issue:41950|issue:45823", + "issue:44779|issue:45812", + "issue:39692|issue:43901", + "issue:44279|issue:45405", + "issue:44683|issue:45542", + "issue:43295|issue:44779", + "issue:43299|issue:43854", + "issue:43650|issue:46132", + "issue:44462|issue:45081", + "issue:43450|issue:44279", + "issue:44297|issue:44462" + ], + "prepared_review_unit_hash": "2b57409df080534f53c133ed5ca0b285ae4fdfa64e2a3c591fd3072a1d1b6e6a", + "component_budget": { + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 30, + "serialized_chars": 36129, + "estimated_input_tokens": 9033, + "estimated_eval_tokens": 18322 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is highly heterogeneous: it mixes unrelated bugs in tokenizer handling, multimodal configs, model loading/saving, pipelines, docs, and dependency/version compatibility. I found no true duplicate pairs; all soft edges are false positives.", + "confidence": 0.93, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": "issue:43299 is the closest representative because it is a concrete, high-signal regression in a major model-loading path, but it still does not serve as a true duplicate hub for this mixed cluster.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43673", + "right": "issue:44336", + "accept": false, + "reason": "Different subsystems and symptoms: chunked_prefill cache handling vs ANSI formatting in loading_report." + }, + { + "left": "issue:43335", + "right": "issue:45593", + "accept": false, + "reason": "Unrelated model logic: SwitchTransformers sparse-layer initialization vs D-FINE auxiliary-loss behavior." + }, + { + "left": "issue:43824", + "right": "issue:44933", + "accept": false, + "reason": "Both are import problems, but for different missing symbols in different modules and with different causes." + }, + { + "left": "issue:44075", + "right": "issue:44368", + "accept": false, + "reason": "Optimizer argument handling is unrelated to the tie_word_embeddings warning in a Qwen LoRA setup." + }, + { + "left": "issue:43299", + "right": "issue:45216", + "accept": false, + "reason": "One is a model-loading failure; the other is a save_pretrained checkpoint serialization regression." + }, + { + "left": "issue:43976", + "right": "issue:44279", + "accept": false, + "reason": "Python-version compatibility and a generic dependency issue are too broad to be the same bug." + }, + { + "left": "issue:43381", + "right": "issue:45593", + "accept": false, + "reason": "Gradient-checkpointing eval-mode behavior is unrelated to D-FINE denoising/auxiliary losses." + }, + { + "left": "issue:43901", + "right": "issue:44496", + "accept": false, + "reason": "Docs mismatch vs a model/config loading error; different failure modes and code paths." + }, + { + "left": "issue:44496", + "right": "issue:45759", + "accept": false, + "reason": "Both involve model configuration, but one is unrecognized-model config metadata and the other is from_config unwrapping for composite Qwen configs." + }, + { + "left": "issue:45335", + "right": "issue:45584", + "accept": false, + "reason": "T5Gemma embedding resizing and Whisper empty-transcription generation are unrelated bugs." + }, + { + "left": "issue:43479", + "right": "issue:43525", + "accept": false, + "reason": "Different config fields and failures: missing default vision/audio configs vs missing pad_token_id." + }, + { + "left": "issue:44521", + "right": "issue:45325", + "accept": false, + "reason": "Both are multimodal, but assistant mask generation and rope-index scaling are separate code paths." + }, + { + "left": "issue:44242", + "right": "issue:45200", + "accept": false, + "reason": "MoE load-balancing loss logic is unrelated to Gemma4 mm_token_type_ids defaults." + }, + { + "left": "issue:43299", + "right": "issue:45357", + "accept": false, + "reason": "Loading regression vs save_pretrained writing incorrect visual-encoder keys; not the same concrete bug." + }, + { + "left": "issue:44351", + "right": "issue:45278", + "accept": false, + "reason": "A specific HybridCache import error is not the same as broad import errors after version upgrades." + }, + { + "left": "issue:43577", + "right": "issue:45216", + "accept": false, + "reason": "Blip2 dtype propagation and Qwen3.5 checkpoint saving are different subsystems and symptoms." + }, + { + "left": "issue:41950", + "right": "issue:45020", + "accept": false, + "reason": "Video-classification processor selection and remote_code model breakage are unrelated." + }, + { + "left": "issue:43976", + "right": "issue:45405", + "accept": false, + "reason": "Python support/versioning and a PEFT minimum-version bump are not the same bug." + }, + { + "left": "issue:38175", + "right": "issue:44617", + "accept": false, + "reason": "SigLIP2 zero probabilities and Sam3Video OOM are completely different problems." + }, + { + "left": "issue:41950", + "right": "issue:45823", + "accept": false, + "reason": "Pipeline processor lookup vs Gemma4 device placement mismatch are unrelated failures." + }, + { + "left": "issue:44779", + "right": "issue:45812", + "accept": false, + "reason": "Both are tokenizer regressions, but they affect different model families and likely different tokenizer code paths." + }, + { + "left": "issue:39692", + "right": "issue:43901", + "accept": false, + "reason": "SigLIP2 docs example errors and a TextClassificationPipeline docs note are unrelated documentation issues." + }, + { + "left": "issue:44279", + "right": "issue:45405", + "accept": false, + "reason": "Generic dependency trouble and a specific unreleased PEFT version bump are different issues." + }, + { + "left": "issue:44683", + "right": "issue:45542", + "accept": false, + "reason": "torch flex_attention compilation failure is unrelated to the TensorFlow/tensorboard backend error." + }, + { + "left": "issue:43295", + "right": "issue:44779", + "accept": false, + "reason": "Custom model processor/tokenizer regression is different from Deepseek tokenizer output corruption." + }, + { + "left": "issue:43299", + "right": "issue:43854", + "accept": false, + "reason": "They are both model-loading reports, but for different models and different failure points." + }, + { + "left": "issue:43650", + "right": "issue:46132", + "accept": false, + "reason": "Noise/placeholder issue title vs an AttentionInterface.register behavior change; no shared bug." + }, + { + "left": "issue:44462", + "right": "issue:45081", + "accept": false, + "reason": "AutoTokenizer ignoring tokenizer.json and a Mistral regex patch crash are separate tokenizer-loading bugs." + }, + { + "left": "issue:43450", + "right": "issue:44279", + "accept": false, + "reason": "Batched video processor output shape and a dependency issue have no concrete overlap." + }, + { + "left": "issue:44297", + "right": "issue:44462", + "accept": false, + "reason": "Saving the wrong tokenizer_class and ignoring tokenizer.json are different serialization/loading defects." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 17, + "review_unit_index": 34, + "review_unit_count": 50, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-428", + "nodes": [ + "issue:33290", + "issue:41950", + "issue:42175", + "issue:42994", + "issue:43065", + "issue:43262", + "issue:43317", + "issue:43441", + "issue:43525", + "issue:43650", + "issue:43653", + "issue:43673", + "issue:43723", + "issue:43761", + "issue:43784", + "issue:43824", + "issue:43881", + "issue:43901", + "issue:43976", + "issue:43994", + "issue:44016", + "issue:44162", + "issue:44261", + "issue:44263", + "issue:44291", + "issue:44297", + "issue:44361", + "issue:44466", + "issue:44521", + "issue:44623", + "issue:44671", + "issue:44683", + "issue:44749", + "issue:44779", + "issue:44898", + "issue:45081", + "issue:45276", + "issue:45290", + "issue:45397", + "issue:45542", + "issue:45593", + "issue:45600", + "issue:45656", + "issue:45712", + "issue:45823", + "issue:45834", + "issue:45907" + ], + "soft_pairs": [ + "issue:44016|issue:44291", + "issue:45081|issue:45907", + "issue:41950|issue:43650", + "issue:43441|issue:45834", + "issue:44466|issue:45276", + "issue:45600|issue:45712", + "issue:42994|issue:43317", + "issue:33290|issue:45656", + "issue:43650|issue:45290", + "issue:44623|issue:44749", + "issue:44623|issue:45397", + "issue:43653|issue:45593", + "issue:43065|issue:44898", + "issue:44261|issue:44263", + "issue:43441|issue:44683", + "issue:44016|issue:45823", + "issue:43262|issue:44521", + "issue:44361|issue:45593", + "issue:42175|issue:43976", + "issue:43723|issue:44162", + "issue:44671|issue:44779", + "issue:43901|issue:43994", + "issue:43673|issue:43761", + "issue:43784|issue:45542", + "issue:43824|issue:43976", + "issue:44297|issue:45397", + "issue:43525|issue:43881" + ], + "prepared_review_unit_hash": "fe8d52f42a9911139d324c5f7fcde60063736d008beb26e3fa11ae2d96915bfc", + "component_budget": { + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 27, + "serialized_chars": 35215, + "estimated_input_tokens": 8804, + "estimated_eval_tokens": 17864 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is highly heterogeneous: the soft-linked issues cover unrelated tokenizer, model, pipeline, packaging, CI, and notebook problems. None of the candidate soft edges look like true duplicates or a single shared fix.", + "confidence": 0.98, + "canonical_issue_reason": "No clear canonical issue: the cluster mixes unrelated bugs across many subsystems, so there is no single issue that represents the underlying problem.", + "canonical_pr_reason": null, + "best_issue_reason": "No strong best issue candidate for deduplication purposes; the issues are too diverse to safely treat one as the representative fix target.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44016", + "right": "issue:44291", + "accept": false, + "reason": "Notebook syntax error vs. init_empty_weights argument mismatch; different failure modes and code paths." + }, + { + "left": "issue:45081", + "right": "issue:45907", + "accept": false, + "reason": "Mistral tokenizer regex crash vs. list_repo_files revision forwarding; unrelated areas." + }, + { + "left": "issue:41950", + "right": "issue:43650", + "accept": false, + "reason": "Video-classification processor bug vs. an unrelated issue titled \"ADD THE DATA\"; no shared bug." + }, + { + "left": "issue:43441", + "right": "issue:45834", + "accept": false, + "reason": "Ministral FlashAttention regression vs. Kosmos2.5 OCR index error; different models and symptoms." + }, + { + "left": "issue:44466", + "right": "issue:45276", + "accept": false, + "reason": "lm_head.weight serialization inconsistency vs. Gemma4 embedding resize behavior; not the same bug." + }, + { + "left": "issue:45600", + "right": "issue:45712", + "accept": false, + "reason": "Sam3LiteText auto-mapping CI break vs. leftover dummy classes in dummy_pt_objects; separate cleanup issues." + }, + { + "left": "issue:42994", + "right": "issue:43317", + "accept": false, + "reason": "Quantized model saving failure vs. dequantized model loading/offload failure; opposite workflows, different bugs." + }, + { + "left": "issue:33290", + "right": "issue:45656", + "accept": false, + "reason": "Adafactor OOM under deepspeed vs. optimizer.step called twice under deepspeed; related subsystem but different concrete defects." + }, + { + "left": "issue:43650", + "right": "issue:45290", + "accept": false, + "reason": "Unrelated issue text vs. apply_chat_template crash on tool-call assistant messages." + }, + { + "left": "issue:44623", + "right": "issue:44749", + "accept": false, + "reason": "processor.save_pretrained missing files vs. Chinese-language performance regression report; unrelated." + }, + { + "left": "issue:44623", + "right": "issue:45397", + "accept": false, + "reason": "Missing processor files on save vs. gemma-4 zero3 from_pretrained failure; different code paths." + }, + { + "left": "issue:43653", + "right": "issue:45593", + "accept": false, + "reason": "BigBirdTokenizer special-token registration bug vs. D-FINE auxiliary-loss behavior; unrelated models/features." + }, + { + "left": "issue:43065", + "right": "issue:44898", + "accept": false, + "reason": "Sam3PixelDecoder dummy Conv2d vs. Perceiver non-default-resolution interpolation bug; different components." + }, + { + "left": "issue:44261", + "right": "issue:44263", + "accept": false, + "reason": "MLA q_a_layernorm precision/config issue vs. torch.split return values in GlmMoeDsaIndexer; unrelated defects." + }, + { + "left": "issue:43441", + "right": "issue:44683", + "accept": false, + "reason": "FlashAttention failure for Ministral vs. compiled flex_attention failure on torch>=2.9; both attention-related but not the same bug." + }, + { + "left": "issue:44016", + "right": "issue:45823", + "accept": false, + "reason": "Notebook syntax error vs. Gemma4 PLE device mismatch; no shared underlying issue." + }, + { + "left": "issue:43262", + "right": "issue:44521", + "accept": false, + "reason": "Audio apply_chat_template sampling-rate default vs. multimodal assistant_masks bug; same API surface, different bug." + }, + { + "left": "issue:44361", + "right": "issue:45593", + "accept": false, + "reason": "MLukeTokenizer AttributeError vs. D-FINE loss configuration issue; unrelated." + }, + { + "left": "issue:42175", + "right": "issue:43976", + "accept": false, + "reason": "Missing TensorFlow in torch-only install vs. Python version compatibility failure; different packaging problems." + }, + { + "left": "issue:43723", + "right": "issue:44162", + "accept": false, + "reason": "AutoTokenizer loading problem in v5 vs. ESM2 broken inference; distinct models and symptoms." + }, + { + "left": "issue:44671", + "right": "issue:44779", + "accept": false, + "reason": "CamemBERT masked-LM regression vs. Deepseek tokenizer regression; both v5 regressions, but different code paths." + }, + { + "left": "issue:43901", + "right": "issue:43994", + "accept": false, + "reason": "Documentation mismatch for return_all_scores vs. SigLIP2 bad model/pipeline outputs; unrelated." + }, + { + "left": "issue:43673", + "right": "issue:43761", + "accept": false, + "reason": "GenerationMixin chunked_prefill cache issue vs. CLIPVisionModel hidden_states bug; unrelated model internals." + }, + { + "left": "issue:43784", + "right": "issue:45542", + "accept": false, + "reason": "Missing nn import in sentence-transformers integration vs. TensorBoard/TensorFlow backend confusion; different dependency issues." + }, + { + "left": "issue:43824", + "right": "issue:43976", + "accept": false, + "reason": "Qwen2.5-VL import error vs. Python 3.9/3.10 compatibility issue; unrelated." + }, + { + "left": "issue:44297", + "right": "issue:45397", + "accept": false, + "reason": "tokenizer.save_pretrained metadata mismatch vs. gemma-4 zero3 loading failure; different functionality." + }, + { + "left": "issue:43525", + "right": "issue:43881", + "accept": false, + "reason": "Llama4Config missing pad_token_id vs. GLM-4V-9B loading failure; different models and bugs." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 17, + "review_unit_index": 35, + "review_unit_count": 50, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-428", + "nodes": [ + "issue:36246", + "issue:41628", + "issue:41950", + "issue:42491", + "issue:42757", + "issue:42886", + "issue:43299", + "issue:43352", + "issue:43450", + "issue:43525", + "issue:43526", + "issue:43531", + "issue:43596", + "issue:43643", + "issue:43673", + "issue:43901", + "issue:43976", + "issue:44162", + "issue:44188", + "issue:44220", + "issue:44242", + "issue:44246", + "issue:44295", + "issue:44336", + "issue:44351", + "issue:44373", + "issue:44485", + "issue:44496", + "issue:44561", + "issue:44589", + "issue:44617", + "issue:44749", + "issue:44779", + "issue:44857", + "issue:44871", + "issue:45003", + "issue:45200", + "issue:45250", + "issue:45292", + "issue:45362", + "issue:45397", + "issue:45412", + "issue:45561", + "issue:45563", + "issue:45593", + "issue:46123", + "issue:46132" + ], + "soft_pairs": [ + "issue:43450|issue:44617", + "issue:43525|issue:43643", + "issue:43901|issue:44246", + "issue:44188|issue:44336", + "issue:41628|issue:43976", + "issue:44589|issue:45362", + "issue:45561|issue:45563", + "issue:42491|issue:43299", + "issue:44336|issue:44373", + "issue:44295|issue:45397", + "issue:44242|issue:45593", + "issue:44246|issue:44496", + "issue:41950|issue:44220", + "issue:43531|issue:45362", + "issue:43526|issue:45003", + "issue:42886|issue:46132", + "issue:36246|issue:42757", + "issue:36246|issue:44351", + "issue:44336|issue:44485", + "issue:45003|issue:45561", + "issue:44246|issue:44871", + "issue:43673|issue:44561", + "issue:44242|issue:44857", + "issue:44295|issue:44749", + "issue:36246|issue:43596", + "issue:43352|issue:45250", + "issue:44779|issue:45292", + "issue:44162|issue:45200", + "issue:43901|issue:45003", + "issue:45412|issue:46123" + ], + "prepared_review_unit_hash": "053fcbee34f06dc052053ce62370aee7aaad2420619ab2ac9f97aa5d46bb8d4e", + "component_budget": { + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 30, + "serialized_chars": 35818, + "estimated_input_tokens": 8955, + "estimated_eval_tokens": 18166 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "Mostly a loose grab-bag of unrelated issues: import/export regressions, model-specific loading bugs, pipeline behavior, performance, docs, and test infrastructure. There is no clear duplicate cluster; only a few items are superficially similar around import failures or version regressions.", + "confidence": 0.69, + "canonical_issue_reason": "41628 is the closest broad representative of the import/export regression subset (`Cannot import name 'AutoImageProcessor' from 'transformers'`), which is the nearest shared theme here, though the overall cluster is still heterogeneous.", + "canonical_pr_reason": null, + "best_issue_reason": "41628 is the best global issue candidate because it is a generic top-level import failure and most plausibly serves as the broad representative of the small import-API regression subset.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43450", + "right": "issue:44617", + "accept": false, + "reason": "Both involve video/model behavior, but one is a batched video-processor shape bug and the other is a Sam3Video OOM issue; different failure modes and fixes." + }, + { + "left": "issue:43525", + "right": "issue:43643", + "accept": false, + "reason": "A missing `pad_token_id` attribute in `Llama4Config` is not the same as missing fields from `AutoConfig.from_pretrained(..., trust_remote_code=True)`." + }, + { + "left": "issue:43901", + "right": "issue:44246", + "accept": false, + "reason": "Docs mismatch for `return_all_scores` and intermittent import slowness are unrelated." + }, + { + "left": "issue:44188", + "right": "issue:44336", + "accept": false, + "reason": "Attention-kernel divergence under `torch.compile` is unrelated to ANSI codes emitted by loading report utilities." + }, + { + "left": "issue:41628", + "right": "issue:43976", + "accept": false, + "reason": "Both mention compatibility/import problems, but one is a missing `AutoImageProcessor` export and the other is a Python-version support issue." + }, + { + "left": "issue:44589", + "right": "issue:45362", + "accept": false, + "reason": "A missing float8 storage type and a Qwen3.5 chat crash are different concrete bugs." + }, + { + "left": "issue:45561", + "right": "issue:45563", + "accept": false, + "reason": "Pytest-xdist file race and a stale `num_return_sequences` warning do not share the same code path." + }, + { + "left": "issue:42491", + "right": "issue:43299", + "accept": false, + "reason": "Both are Qwen/MoE-related regressions, but they affect different model families and loading paths; not the same bug." + }, + { + "left": "issue:44336", + "right": "issue:44373", + "accept": false, + "reason": "ANSI escape codes in loading reports and a wrong docstring for `position_ids` are unrelated." + }, + { + "left": "issue:44295", + "right": "issue:45397", + "accept": false, + "reason": "A `position_ids` buffer access error and a Gemma-4 Zero3 `from_pretrained` issue are distinct failures." + }, + { + "left": "issue:44242", + "right": "issue:45593", + "accept": false, + "reason": "Missing auxiliary loss in MoE routing and missing auxiliary losses for D-FINE denoising off are different model-specific training bugs." + }, + { + "left": "issue:44246", + "right": "issue:44496", + "accept": false, + "reason": "Import performance and an unrecognized model/config error are unrelated." + }, + { + "left": "issue:41950", + "right": "issue:44220", + "accept": false, + "reason": "Video-classification processor lookup and `_torch_extract_fbank_features()` are different subsystems and bugs." + }, + { + "left": "issue:43531", + "right": "issue:45362", + "accept": false, + "reason": "Both reference Qwen, but one is a `sliding_window` bug in MoE models and the other is a chat crash on Qwen3.5-35B; not the same fix." + }, + { + "left": "issue:43526", + "right": "issue:45003", + "accept": false, + "reason": "A Beit image processor label-reduction bug is unrelated to unsafe `sys.modules[]` access in `modeling_utils`." + }, + { + "left": "issue:42886", + "right": "issue:46132", + "accept": false, + "reason": "Offline cache loading and `AttentionInterface.register` behavior changes are different problems." + }, + { + "left": "issue:36246", + "right": "issue:42757", + "accept": false, + "reason": "Both are import errors, but they come from different packages and symbols (`transformers.models.qwen2_5_vl` vs `huggingface_hub`)." + }, + { + "left": "issue:36246", + "right": "issue:44351", + "accept": false, + "reason": "Missing `Qwen2_5_VLImageProcessor` and missing `HybridCache` exports are separate API regressions." + }, + { + "left": "issue:44336", + "right": "issue:44485", + "accept": false, + "reason": "Loading-report ANSI output and GLM-5 RoPE implementation are unrelated." + }, + { + "left": "issue:45003", + "right": "issue:45561", + "accept": false, + "reason": "Unsafe `sys.modules[]` access and a pytest-xdist file race do not describe the same bug." + }, + { + "left": "issue:44246", + "right": "issue:44871", + "accept": false, + "reason": "Import slowness is unrelated to Gemma-3 `eos_token_id` config inconsistency." + }, + { + "left": "issue:43673", + "right": "issue:44561", + "accept": false, + "reason": "Missing GenerationMixin cache in chunked prefill and removal of `is_torch_fx_available` breaking trust-remote-code models are different regressions." + }, + { + "left": "issue:44242", + "right": "issue:44857", + "accept": false, + "reason": "Different auxiliary-loss/loss-scaling bugs in different models; no shared concrete code path." + }, + { + "left": "issue:44295", + "right": "issue:44749", + "accept": false, + "reason": "A `position_ids` buffer error and a slow data-filtering regression after upgrade are unrelated." + }, + { + "left": "issue:36246", + "right": "issue:43596", + "accept": false, + "reason": "An import error for a vision processor and an index error in `BertModel` are unrelated." + }, + { + "left": "issue:43352", + "right": "issue:45250", + "accept": false, + "reason": "One is a model-specific unsupported Flash Attention 2 issue; the other is only a generic Flash Attention 2 topic." + }, + { + "left": "issue:44779", + "right": "issue:45292", + "accept": false, + "reason": "Deepseek tokenizer correctness and `resize_token_embeddings` not updating output embeddings are different bugs." + }, + { + "left": "issue:44162", + "right": "issue:45200", + "accept": false, + "reason": "ESM2 breakage and Gemma 4 `mm_token_type_ids` defaults concern different models and code paths." + }, + { + "left": "issue:43901", + "right": "issue:45003", + "accept": false, + "reason": "A pipeline docs mismatch and unsafe `sys.modules[]` access are unrelated." + }, + { + "left": "issue:45412", + "right": "issue:46123", + "accept": false, + "reason": "RT-DETR memory release and MaskGenerationPipeline dropping partial-batch results are different runtime bugs." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 17, + "review_unit_index": 36, + "review_unit_count": 50, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-428", + "nodes": [ + "issue:42886", + "issue:43065", + "issue:43232", + "issue:43257", + "issue:43262", + "issue:43408", + "issue:43475", + "issue:43540", + "issue:43653", + "issue:43673", + "issue:43931", + "issue:44060", + "issue:44077", + "issue:44162", + "issue:44246", + "issue:44336", + "issue:44368", + "issue:44466", + "issue:44479", + "issue:44496", + "issue:44530", + "issue:44734", + "issue:44877", + "issue:44933", + "issue:44977", + "issue:44987", + "issue:44995", + "issue:45003", + "issue:45137", + "issue:45200", + "issue:45229", + "issue:45292", + "issue:45335", + "issue:45375", + "issue:45397", + "issue:45459", + "issue:45563", + "issue:45584", + "issue:45588", + "issue:45600", + "issue:45656", + "issue:45701", + "issue:45759", + "issue:45987", + "issue:46082", + "issue:46123", + "issue:46170" + ], + "soft_pairs": [ + "issue:44877|issue:45375", + "issue:45987|issue:46082", + "issue:44496|issue:45003", + "issue:45375|issue:45759", + "issue:43262|issue:44246", + "issue:45292|issue:45584", + "issue:43257|issue:45137", + "issue:45292|issue:45701", + "issue:44987|issue:45459", + "issue:43232|issue:44336", + "issue:43065|issue:43408", + "issue:44995|issue:45200", + "issue:44479|issue:46170", + "issue:43475|issue:43540", + "issue:42886|issue:45229", + "issue:44466|issue:45335", + "issue:45397|issue:45656", + "issue:44060|issue:44077", + "issue:43653|issue:44162", + "issue:44368|issue:44977", + "issue:44995|issue:45588", + "issue:44368|issue:45563", + "issue:45137|issue:45600", + "issue:44877|issue:46170", + "issue:43673|issue:44734", + "issue:43931|issue:44368", + "issue:43065|issue:46123", + "issue:44530|issue:46170", + "issue:44933|issue:45459" + ], + "prepared_review_unit_hash": "1b457941161f3e28ab34de6e36bf0d02941ed7e7990abf66dfc9b0d6cd4e3e73", + "component_budget": { + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 29, + "serialized_chars": 35852, + "estimated_input_tokens": 8963, + "estimated_eval_tokens": 18182 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This is not a true duplicate cluster: the items span many unrelated bugs and features (tokenizers, generation/cache, SAM3, Qwen, DeepSpeed, config loading, etc.). The soft pairs mostly share broad subsystems or keywords, but not the same concrete bug or change.", + "confidence": 0.95, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44877", + "right": "issue:45375", + "accept": false, + "reason": "Both mention config strictness, but they are different models and different missing-field failures." + }, + { + "left": "issue:45987", + "right": "issue:46082", + "accept": false, + "reason": "Cache return type and Llama head_dim validation are unrelated issues." + }, + { + "left": "issue:44496", + "right": "issue:45003", + "accept": false, + "reason": "Model auto-loading failure and unsafe sys.modules access are different problems." + }, + { + "left": "issue:45375", + "right": "issue:45759", + "accept": false, + "reason": "Both are multimodal Qwen config issues, but they affect different fields and code paths." + }, + { + "left": "issue:43262", + "right": "issue:44246", + "accept": false, + "reason": "Audio sampling-rate handling and import slowness are unrelated." + }, + { + "left": "issue:45292", + "right": "issue:45584", + "accept": false, + "reason": "resize_token_embeddings bug and Whisper empty-transcription failure are different bugs." + }, + { + "left": "issue:43257", + "right": "issue:45137", + "accept": false, + "reason": "DeepSpeed appears in both, but one is weight conversion and the other is a ZeRO3 deque error." + }, + { + "left": "issue:45292", + "right": "issue:45701", + "accept": false, + "reason": "Embedding resize behavior and tokenization version drift are not the same issue." + }, + { + "left": "issue:44987", + "right": "issue:45459", + "accept": false, + "reason": "Model loading regression and protobuf decode-error masking are unrelated." + }, + { + "left": "issue:43232", + "right": "issue:44336", + "accept": false, + "reason": "Generation kwargs handling and ANSI output formatting are unrelated." + }, + { + "left": "issue:43065", + "right": "issue:43408", + "accept": false, + "reason": "Both are SAM3-related, but one is a dummy conv module and the other is a model-type mismatch warning." + }, + { + "left": "issue:44995", + "right": "issue:45200", + "accept": false, + "reason": "A stale indexer cache crash and missing mm_token_type_ids are different bugs." + }, + { + "left": "issue:44479", + "right": "issue:46170", + "accept": false, + "reason": "A video-input regression and a Windows/ROCm distributed import error are unrelated." + }, + { + "left": "issue:43475", + "right": "issue:43540", + "accept": false, + "reason": "SAM3 vision encoder output and Qwen3OmniMoe video processing failures are different code paths." + }, + { + "left": "issue:42886", + "right": "issue:45229", + "accept": false, + "reason": "Tokenizer offline-cache loading and Gemma4 CUDA OOM are unrelated." + }, + { + "left": "issue:44466", + "right": "issue:45335", + "accept": false, + "reason": "Both involve tied/embedding weights, but the concrete failures are different." + }, + { + "left": "issue:45397", + "right": "issue:45656", + "accept": false, + "reason": "Both involve DeepSpeed, but one is from_pretrained loading and the other is double optimizer stepping." + }, + { + "left": "issue:44060", + "right": "issue:44077", + "accept": false, + "reason": "Qwen3-Next tied-weight warning and PatchTSMixer post_init policy are unrelated." + }, + { + "left": "issue:43653", + "right": "issue:44162", + "accept": false, + "reason": "BigBirdTokenizer special-token registration and ESM2 breakage are different issues." + }, + { + "left": "issue:44368", + "right": "issue:44977", + "accept": false, + "reason": "One is a config warning about tie_word_embeddings; the other is a flash-attention generation regression." + }, + { + "left": "issue:44995", + "right": "issue:45588", + "accept": false, + "reason": "Stale cache crash and flash_attention s_aux=None crash are unrelated." + }, + { + "left": "issue:44368", + "right": "issue:45563", + "accept": false, + "reason": "A tied-weights warning and a stale num_return_sequences warning are different bugs." + }, + { + "left": "issue:45137", + "right": "issue:45600", + "accept": false, + "reason": "DeepSpeed ZeRO3 deque failure and Sam3LiteText auto-mapping CI breakage are unrelated." + }, + { + "left": "issue:44877", + "right": "issue:46170", + "accept": false, + "reason": "Strict config loading and Windows ROCm distributed import errors are unrelated." + }, + { + "left": "issue:43673", + "right": "issue:44734", + "accept": false, + "reason": "Both mention cache/generation, but they fail in different components and modes." + }, + { + "left": "issue:43931", + "right": "issue:44368", + "accept": false, + "reason": "A Qwen3-VL weight-shape mismatch and a tie_word_embeddings warning are unrelated." + }, + { + "left": "issue:43065", + "right": "issue:46123", + "accept": false, + "reason": "SAM3 dummy layer handling and MaskGenerationPipeline partial-batch behavior are unrelated." + }, + { + "left": "issue:44530", + "right": "issue:46170", + "accept": false, + "reason": "PagedAttentionCache linear_attention error and Windows ROCm torch.distributed import failure are unrelated." + }, + { + "left": "issue:44933", + "right": "issue:45459", + "accept": false, + "reason": "Missing image_utils import and protobuf decode-error masking are unrelated." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 17, + "review_unit_index": 37, + "review_unit_count": 50, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-428", + "nodes": [ + "issue:28282", + "issue:30990", + "issue:36010", + "issue:41628", + "issue:42371", + "issue:42673", + "issue:42831", + "issue:42994", + "issue:43232", + "issue:43475", + "issue:43482", + "issue:43525", + "issue:43643", + "issue:43646", + "issue:43673", + "issue:43701", + "issue:43704", + "issue:43873", + "issue:44038", + "issue:44162", + "issue:44188", + "issue:44222", + "issue:44297", + "issue:44351", + "issue:44361", + "issue:44623", + "issue:44661", + "issue:44749", + "issue:44869", + "issue:44871", + "issue:44933", + "issue:44945", + "issue:45081", + "issue:45092", + "issue:45103", + "issue:45230", + "issue:45278", + "issue:45310", + "issue:45356", + "issue:45375", + "issue:45381", + "issue:45397", + "issue:45478", + "issue:45656", + "issue:45753", + "issue:46077", + "issue:46123", + "issue:46132" + ], + "soft_pairs": [ + "issue:44297|issue:44623", + "issue:44869|issue:45356", + "issue:44361|issue:44749", + "issue:43646|issue:44162", + "issue:43232|issue:43673", + "issue:44623|issue:45656", + "issue:44749|issue:45656", + "issue:43673|issue:45278", + "issue:44871|issue:46132", + "issue:43701|issue:44749", + "issue:43475|issue:45381", + "issue:42994|issue:43482", + "issue:45230|issue:45397", + "issue:43475|issue:43643", + "issue:28282|issue:30990", + "issue:44038|issue:44623", + "issue:46077|issue:46123", + "issue:41628|issue:44933", + "issue:42371|issue:45092", + "issue:43475|issue:44661", + "issue:45375|issue:45478", + "issue:45081|issue:45103", + "issue:42994|issue:43525", + "issue:45310|issue:45375", + "issue:43873|issue:44945", + "issue:43232|issue:44188", + "issue:41628|issue:44351", + "issue:42673|issue:43704", + "issue:42831|issue:44222", + "issue:44945|issue:45753", + "issue:36010|issue:41628", + "issue:44222|issue:45310" + ], + "prepared_review_unit_hash": "cd1499c83d15e10908153d230f2a7e0b1ac66d58262725a155d9dacd2a936356", + "component_budget": { + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 32, + "serialized_chars": 36963, + "estimated_input_tokens": 9241, + "estimated_eval_tokens": 18738 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly unrelated Transformers issues connected only by broad themes like imports, generation, saving/loading, or model-family regressions. There are two clear exact duplicate pairs in the packet: issue:42673/issue:43704 (Qwen3ForCausalLM VRAM leak in multiple dataloader threads) and issue:45310/issue:45478 (Qwen3.5 MoE from_pretrained error). The remaining soft edges look like conservative false positives rather than true duplicates.", + "confidence": 0.91, + "canonical_issue_reason": "issue:42673 is the strongest canonical for its exact duplicate pair because it is the earlier, fuller report of the Qwen3ForCausalLM VRAM leak; issue:43704 is an exact duplicate. Note that the packet also contains a separate exact duplicate pair, issue:45310/issue:45478, so this cluster is not a single duplicate set.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:42673 is the best representative overall because it has the fullest discussion/history among the exact duplicate pairs and is the clearest stable reference point for triage.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44297", + "right": "issue:44623", + "accept": false, + "reason": "Both mention save_pretrained, but one is a tokenizer_config mismatch and the other is missing processor files; different failure modes and code paths." + }, + { + "left": "issue:44869", + "right": "issue:45356", + "accept": false, + "reason": "Both are tokenizer/codec-related, but one is a Whisper timestamp decode crash and the other is a Kimi-K2.5 tokenizer regression with a different warning/codec issue." + }, + { + "left": "issue:44361", + "right": "issue:44749", + "accept": false, + "reason": "Different problems: MLukeTokenizer task AttributeError vs a general performance slowdown after upgrade." + }, + { + "left": "issue:43646", + "right": "issue:44162", + "accept": false, + "reason": "Both are Transformers v5 regressions, but one is custom model initialization and the other is ESM2-specific breakage; not the same underlying bug." + }, + { + "left": "issue:43232", + "right": "issue:43673", + "accept": false, + "reason": "Both involve generation internals, but one is sync_gpus/update_model_kwargs behavior and the other is chunked_prefill cache handling; distinct code paths." + }, + { + "left": "issue:44623", + "right": "issue:45656", + "accept": false, + "reason": "processor.save_pretrained file issues are unrelated to a deepspeed optimizer-step counting bug." + }, + { + "left": "issue:44749", + "right": "issue:45656", + "accept": false, + "reason": "A data-filtering slowdown after upgrade is unrelated to optimizer.step being called twice." + }, + { + "left": "issue:43673", + "right": "issue:45278", + "accept": false, + "reason": "One is a specific GenerationMixin cache regression; the other is a broad import-error report after upgrade." + }, + { + "left": "issue:44871", + "right": "issue:46132", + "accept": false, + "reason": "Gemma eos_token_id inconsistency is unrelated to AttentionInterface.register changing registered-function behavior." + }, + { + "left": "issue:43701", + "right": "issue:44749", + "accept": false, + "reason": "Checkpoint resume key mismatch is unrelated to dataset filtering performance." + }, + { + "left": "issue:43475", + "right": "issue:45381", + "accept": false, + "reason": "Both are vision/video model issues, but one is a missing SAM3 attribute and the other is incorrect Qwen2.5-VL video position IDs." + }, + { + "left": "issue:42994", + "right": "issue:43482", + "accept": false, + "reason": "Quantized model saving failure is different from Qwen2.5-GGUF loading failure." + }, + { + "left": "issue:45230", + "right": "issue:45397", + "accept": false, + "reason": "issue:45230 is too generic to establish a shared bug, while 45397 is a specific gemma-4 zero3 from_pretrained problem." + }, + { + "left": "issue:43475", + "right": "issue:43643", + "accept": false, + "reason": "Different failures: SAM3 output attribute missing vs trust_remote_code returning missing fields from AutoConfig." + }, + { + "left": "issue:28282", + "right": "issue:30990", + "accept": false, + "reason": "An ImportError about PyTorch missing is unrelated to Sentence Transformers getting stuck during load." + }, + { + "left": "issue:44038", + "right": "issue:44623", + "accept": false, + "reason": "Qwen3-VL-Moe v5 incompatibility and processor.save_pretrained missing files are unrelated issues." + }, + { + "left": "issue:46077", + "right": "issue:46123", + "accept": false, + "reason": "Gemma4 causal-mask construction and MaskGenerationPipeline final-batch handling are different bugs." + }, + { + "left": "issue:41628", + "right": "issue:44933", + "accept": false, + "reason": "Both are import errors, but for different symbols/modules and likely different fixes." + }, + { + "left": "issue:42371", + "right": "issue:45092", + "accept": false, + "reason": "TF32 API guidance is unrelated to remote-code checkpoint meta-initialization incompatibility." + }, + { + "left": "issue:43475", + "right": "issue:44661", + "accept": false, + "reason": "A SAM3 video attribute error is unrelated to add-new-model-like failing on TOKENIZER_MAPPING_NAMES." + }, + { + "left": "issue:45375", + "right": "issue:45478", + "accept": false, + "reason": "Both involve Qwen3.5 MoE, but one is a missing vision-config field and the other is a general from_pretrained error; not enough to treat as the same bug." + }, + { + "left": "issue:45081", + "right": "issue:45103", + "accept": false, + "reason": "Mistral regex patching and auto_docstring future-annotations crashes are unrelated AttributeErrors in different subsystems." + }, + { + "left": "issue:42994", + "right": "issue:43525", + "accept": false, + "reason": "Quantized model saving failure is different from Llama4Config missing pad_token_id." + }, + { + "left": "issue:45310", + "right": "issue:45375", + "accept": false, + "reason": "Both are Qwen3.5 MoE-related, but one is a from_pretrained error and the other is a missing vision-config field; not clearly the same underlying defect." + }, + { + "left": "issue:43873", + "right": "issue:44945", + "accept": false, + "reason": "Quantization/offloading behavior is unrelated to incorrect pipeline-parallel LLM outputs." + }, + { + "left": "issue:43232", + "right": "issue:44188", + "accept": false, + "reason": "Generation kwargs after sync_gpus is unrelated to attention-kernel divergence under torch.compile." + }, + { + "left": "issue:41628", + "right": "issue:44351", + "accept": false, + "reason": "Different missing imports: AutoImageProcessor vs HybridCache." + }, + { + "left": "issue:42673", + "right": "issue:43704", + "accept": true, + "reason": "Exact same title and same reported bug: Qwen3ForCausalLM VRAM leak in multiple dataloader threads." + }, + { + "left": "issue:42831", + "right": "issue:44222", + "accept": false, + "reason": "Both are FP8-related, but one is an accuracy issue and the other is save_pretrained for moe models." + }, + { + "left": "issue:44945", + "right": "issue:45753", + "accept": false, + "reason": "Pipeline-parallel wrong output is unrelated to a Qwen3_5 infinite loop on a specific image." + }, + { + "left": "issue:36010", + "right": "issue:41628", + "accept": false, + "reason": "Different import failures: GenerationMixin import vs AutoImageProcessor import." + }, + { + "left": "issue:44222", + "right": "issue:45310", + "accept": false, + "reason": "FP8 moe save_pretrained and Qwen3.5-Moe from_pretrained errors are different bugs." + } + ] + }, + "evaluator_result": { + "accept": false, + "feedback": "The summary is not fully grounded: the packet contains another exact duplicate pair besides 42673/43704, namely issue:45310 and issue:45478, which have the same title and clearly describe the same Qwen3.5 MoE from_pretrained error. Because the response says there is only one clear duplicate pair, it overstates the evidence. The other soft-edge rejections are generally conservative, but the missing duplicate makes the overall assessment imprecise." + }, + "evaluator_used": true, + "retried": true, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 17, + "review_unit_index": 38, + "review_unit_count": 50, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-428", + "nodes": [ + "issue:30333", + "issue:30990", + "issue:33290", + "issue:35141", + "issue:42175", + "issue:42222", + "issue:42898", + "issue:42994", + "issue:43232", + "issue:43278", + "issue:43377", + "issue:43479", + "issue:43504", + "issue:43526", + "issue:43673", + "issue:43723", + "issue:43784", + "issue:43824", + "issue:43874", + "issue:43976", + "issue:44038", + "issue:44060", + "issue:44079", + "issue:44188", + "issue:44242", + "issue:44373", + "issue:44448", + "issue:44496", + "issue:44568", + "issue:44623", + "issue:44734", + "issue:44779", + "issue:44792", + "issue:44869", + "issue:44933", + "issue:44964", + "issue:45092", + "issue:45103", + "issue:45125", + "issue:45137", + "issue:45200", + "issue:45310", + "issue:45356", + "issue:45405", + "issue:45561", + "issue:45712", + "issue:45812" + ], + "soft_pairs": [ + "issue:42994|issue:43824", + "issue:44964|issue:45137", + "issue:43232|issue:44792", + "issue:43673|issue:44188", + "issue:44038|issue:45310", + "issue:42175|issue:45712", + "issue:43526|issue:45561", + "issue:43504|issue:45561", + "issue:44496|issue:45561", + "issue:43377|issue:44242", + "issue:42175|issue:44933", + "issue:42222|issue:45200", + "issue:42898|issue:44568", + "issue:43976|issue:44933", + "issue:30990|issue:35141", + "issue:43479|issue:44869", + "issue:43479|issue:44079", + "issue:43278|issue:45137", + "issue:44623|issue:45310", + "issue:42898|issue:45812", + "issue:44188|issue:44373", + "issue:30333|issue:33290", + "issue:43784|issue:43824", + "issue:44734|issue:45103", + "issue:43723|issue:44568", + "issue:42898|issue:45356", + "issue:42898|issue:44779", + "issue:43874|issue:45092", + "issue:45405|issue:45712", + "issue:42898|issue:44448", + "issue:44060|issue:45125" + ], + "prepared_review_unit_hash": "673af8d6d3170d33f11597bec5cacef15485e4eaa6cfc5e2c4b19ee73c0b8383", + "component_budget": { + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 31, + "serialized_chars": 36549, + "estimated_input_tokens": 9138, + "estimated_eval_tokens": 18532 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The cluster is mostly a grab bag of unrelated issues, with a loose concentration around tokenizer/v5 regressions and multimodal/model-loading bugs; none of the soft pairs look like true duplicates. I would anchor the issue cluster on the broad tokenizer regression report, with a more concrete tokenizer bug as the strongest representative.", + "confidence": 0.74, + "canonical_issue_reason": "Issue #42898 is the broadest and most central regression report in the set: it describes a v5 tokenizer behavior change and is thematically closest to several nearby tokenizer-related reports.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue #45812 is the cleanest high-signal report: it has a concrete, model-specific tokenizer failure with a clear silent regression from v4 to v5.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:42994", + "right": "issue:43824", + "accept": false, + "reason": "Quantized model saving and a missing Qwen2.5-VL import symbol are unrelated failures." + }, + { + "left": "issue:44964", + "right": "issue:45137", + "accept": false, + "reason": "Phi-4 multimodal loading and a DeepSpeed ZeRO3 deque error affect different code paths." + }, + { + "left": "issue:43232", + "right": "issue:44792", + "accept": false, + "reason": "Generation kwargs syncing and the janus image-generation test failure are different bugs." + }, + { + "left": "issue:43673", + "right": "issue:44188", + "accept": false, + "reason": "Chunked-prefill cache handling is not the same as the torch.compile attention-kernel branching issue." + }, + { + "left": "issue:44038", + "right": "issue:45310", + "accept": false, + "reason": "Both involve Qwen3 VL/MoE models, but they report different versions and distinct from_pretrained failures." + }, + { + "left": "issue:42175", + "right": "issue:45712", + "accept": false, + "reason": "Missing tensorflow in the torch extra and leftover dummy PT classes are separate packaging problems." + }, + { + "left": "issue:43526", + "right": "issue:45561", + "accept": false, + "reason": "A BeitImageProcessorFast label bug is unrelated to the xdist captured_info.txt race." + }, + { + "left": "issue:43504", + "right": "issue:45561", + "accept": false, + "reason": "Legacy-field model loading and a pytest-xdist file race are not the same issue." + }, + { + "left": "issue:44496", + "right": "issue:45561", + "accept": false, + "reason": "An unrecognized-model config error is unrelated to the testing-utils race condition." + }, + { + "left": "issue:43377", + "right": "issue:44242", + "accept": false, + "reason": "Missing padding-mask support in MIMI encoding is different from the load-balancing-loss gating bug." + }, + { + "left": "issue:42175", + "right": "issue:44933", + "accept": false, + "reason": "Backend dependency resolution and a missing image_utils import are unrelated." + }, + { + "left": "issue:42222", + "right": "issue:45200", + "accept": false, + "reason": "Broken vitpose models and Gemma 4 mm_token_type_ids defaulting are different multimodal bugs." + }, + { + "left": "issue:42898", + "right": "issue:44568", + "accept": false, + "reason": "Both are tokenizer regressions in v5, but one is clean_up_tokenization_spaces behavior and the other is add_special_tokens BOS/EOS handling." + }, + { + "left": "issue:43976", + "right": "issue:44933", + "accept": false, + "reason": "Python version compatibility and a missing import are unrelated." + }, + { + "left": "issue:30990", + "right": "issue:35141", + "accept": false, + "reason": "Sentence-transformer loading hangs and embedding reinitialization in post_init are different problems." + }, + { + "left": "issue:43479", + "right": "issue:44869", + "accept": false, + "reason": "Default multimodal config initialization and Whisper timestamp decoding crash are unrelated." + }, + { + "left": "issue:43479", + "right": "issue:44079", + "accept": false, + "reason": "Config defaulting for multimodal submodules does not match ModelOutput key assignment logic." + }, + { + "left": "issue:43278", + "right": "issue:45137", + "accept": false, + "reason": "A dtype mismatch in evaluation is unrelated to the DeepSpeed deque pop error." + }, + { + "left": "issue:44623", + "right": "issue:45310", + "accept": false, + "reason": "Missing files on processor.save_pretrained and Qwen3.5 MoE from_pretrained failure are different load/save bugs." + }, + { + "left": "issue:42898", + "right": "issue:45812", + "accept": false, + "reason": "Both are tokenizer regressions, but they affect different behaviors and model families, so they are not the same concrete bug." + }, + { + "left": "issue:44188", + "right": "issue:44373", + "accept": false, + "reason": "A torch.compile attention-kernel divergence is unrelated to a docstring typo." + }, + { + "left": "issue:30333", + "right": "issue:33290", + "accept": false, + "reason": "MLflow reporting on failed training jobs and AdaFactor OOM in DeepSpeed are separate training issues." + }, + { + "left": "issue:43784", + "right": "issue:43824", + "accept": false, + "reason": "A sentence-transformers import NameError and a Qwen2.5-VL import error are different import failures." + }, + { + "left": "issue:44734", + "right": "issue:45103", + "accept": false, + "reason": "KV-cache continuation indexing in transformers serve is unrelated to the auto_docstring future-annotations crash." + }, + { + "left": "issue:43723", + "right": "issue:44568", + "accept": false, + "reason": "Tokenizer loading in v5 and BOS/EOS add_special_tokens behavior are different tokenizer regressions." + }, + { + "left": "issue:42898", + "right": "issue:45356", + "accept": false, + "reason": "Clean-up spacing behavior in v5 and Kimi-K2.5 codec handling are separate tokenizer issues." + }, + { + "left": "issue:42898", + "right": "issue:44779", + "accept": false, + "reason": "These are both tokenizer regressions, but for different models and different failure modes." + }, + { + "left": "issue:43874", + "right": "issue:45092", + "accept": false, + "reason": "A missing image-patch count method and an InternVL2 meta-init incompatibility are not the same bug." + }, + { + "left": "issue:45405", + "right": "issue:45712", + "accept": false, + "reason": "A bumped minimum PEFT version and dummy PT class cleanup are unrelated." + }, + { + "left": "issue:42898", + "right": "issue:44448", + "accept": false, + "reason": "A tokenizer behavior regression and a model output discrepancy for Pegasus are different issues." + }, + { + "left": "issue:44060", + "right": "issue:45125", + "accept": false, + "reason": "An incorrect tied-weights warning in Qwen3-Next is not the same as a missing tensor-parallel plan in Qwen3.5 MoE." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 17, + "review_unit_index": 39, + "review_unit_count": 50, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-428", + "nodes": [ + "issue:22355", + "issue:28282", + "issue:30333", + "issue:30990", + "issue:34689", + "issue:35141", + "issue:36331", + "issue:38175", + "issue:41628", + "issue:42757", + "issue:42831", + "issue:42907", + "issue:43116", + "issue:43232", + "issue:43479", + "issue:43644", + "issue:43723", + "issue:43761", + "issue:44190", + "issue:44206", + "issue:44222", + "issue:44263", + "issue:44295", + "issue:44297", + "issue:44360", + "issue:44373", + "issue:44448", + "issue:44483", + "issue:44484", + "issue:44485", + "issue:44514", + "issue:44623", + "issue:44871", + "issue:44933", + "issue:44977", + "issue:45071", + "issue:45092", + "issue:45103", + "issue:45245", + "issue:45310", + "issue:45357", + "issue:45406", + "issue:45656", + "issue:45712", + "issue:45834", + "issue:46050", + "issue:46082", + "issue:46132" + ], + "soft_pairs": [ + "issue:44295|issue:45656", + "issue:44448|issue:46132", + "issue:44933|issue:45712", + "issue:44222|issue:44297", + "issue:44297|issue:45656", + "issue:43116|issue:44190", + "issue:43479|issue:43644", + "issue:44190|issue:46082", + "issue:44222|issue:44623", + "issue:30333|issue:36331", + "issue:42831|issue:44623", + "issue:30333|issue:34689", + "issue:45245|issue:46132", + "issue:45071|issue:46082", + "issue:22355|issue:28282", + "issue:41628|issue:42757", + "issue:44206|issue:45103", + "issue:28282|issue:30333", + "issue:30333|issue:35141", + "issue:44297|issue:45310", + "issue:43723|issue:44977", + "issue:44222|issue:44484", + "issue:43232|issue:45834", + "issue:43761|issue:46082", + "issue:44263|issue:45092", + "issue:44483|issue:45406", + "issue:44295|issue:44623", + "issue:44373|issue:44485", + "issue:44514|issue:45103", + "issue:43723|issue:44871", + "issue:45092|issue:45357", + "issue:44263|issue:44360", + "issue:30990|issue:38175", + "issue:42907|issue:46050" + ], + "prepared_review_unit_hash": "5a8150cde9b810f48fd881f556beaf5d795db9dcb8a9bdef078c070f149a7558", + "component_budget": { + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 34, + "serialized_chars": 37629, + "estimated_input_tokens": 9408, + "estimated_eval_tokens": 19072 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The items are a heterogeneous set of unrelated Transformers issues; no clear duplicate cluster emerges, and all soft candidate pairs should be rejected.", + "confidence": 0.24, + "canonical_issue_reason": "No real duplicate cluster is present; if a representative issue is needed, #44297 is the most concrete, narrowly scoped serialization bug among otherwise unrelated reports.", + "canonical_pr_reason": null, + "best_issue_reason": "#44297 is the best triage candidate because it has a specific, reproducible save_pretrained symptom and a clear code path, making it more actionable than the other heterogeneous issues.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44295", + "right": "issue:45656", + "accept": false, + "reason": "Different bugs: buffer registration/access error versus optimizer stepping twice under deepspeed." + }, + { + "left": "issue:44448", + "right": "issue:46132", + "accept": false, + "reason": "Both touch model behavior, but one is a Pegasus output change and the other is an AttentionInterface registration side effect." + }, + { + "left": "issue:44933", + "right": "issue:45712", + "accept": false, + "reason": "Both are import/export related, but one is a missing image_utils import and the other is leftover dummy PT classes leaking into dir()." + }, + { + "left": "issue:44222", + "right": "issue:44297", + "accept": false, + "reason": "Both involve save_pretrained, but one is FP8 MoE model saving and the other is tokenizer metadata mismatch." + }, + { + "left": "issue:44297", + "right": "issue:45656", + "accept": false, + "reason": "Tokenizer serialization bug is unrelated to a deepspeed optimizer double-step issue." + }, + { + "left": "issue:43116", + "right": "issue:44190", + "accept": false, + "reason": "Example-script failures in different workflows: multi-label classification output versus local dataset loading." + }, + { + "left": "issue:43479", + "right": "issue:43644", + "accept": false, + "reason": "Config default initialization bug is unrelated to non-persistent buffer corruption." + }, + { + "left": "issue:44190", + "right": "issue:46082", + "accept": false, + "reason": "Local dataset loading issue is unrelated to LlamaConfig head_dim validation." + }, + { + "left": "issue:44222", + "right": "issue:44623", + "accept": false, + "reason": "Different save-pretrained problems affecting different objects: FP8 MoE model versus processor artifact files." + }, + { + "left": "issue:30333", + "right": "issue:36331", + "accept": false, + "reason": "Both are training-related, but MLflow reporting on failure is separate from a compute_loss signature break." + }, + { + "left": "issue:42831", + "right": "issue:44623", + "accept": false, + "reason": "Accuracy regression in FineGrainedFP8 is unrelated to missing processor save files." + }, + { + "left": "issue:30333", + "right": "issue:34689", + "accept": false, + "reason": "Training-job reporting and a Llama 3.2 model-loading regression are different code paths." + }, + { + "left": "issue:45245", + "right": "issue:46132", + "accept": false, + "reason": "Category-count runtime error is unrelated to AttentionInterface registration behavior." + }, + { + "left": "issue:45071", + "right": "issue:46082", + "accept": false, + "reason": "PretrainedConfig type-checking break is not the same as LlamaConfig head_dim rejection." + }, + { + "left": "issue:22355", + "right": "issue:28282", + "accept": false, + "reason": "Both are import errors, but one is a missing transformers.onnx module and the other is missing PyTorch dependency detection." + }, + { + "left": "issue:41628", + "right": "issue:42757", + "accept": false, + "reason": "Different missing symbols from different packages; not the same underlying failure." + }, + { + "left": "issue:44206", + "right": "issue:45103", + "accept": false, + "reason": "Feature-extractor argument crash and auto-docstring annotation parsing crash are unrelated." + }, + { + "left": "issue:28282", + "right": "issue:30333", + "accept": false, + "reason": "ImportError for PyTorch dependency is unrelated to MLflow reporting on failed training jobs." + }, + { + "left": "issue:30333", + "right": "issue:35141", + "accept": false, + "reason": "Training reporting bug is unrelated to embedding reinitialization after resize_token_embeddings." + }, + { + "left": "issue:44297", + "right": "issue:45310", + "accept": false, + "reason": "Tokenizer_class mismatch in save_pretrained is not the same as Qwen3.5 MoE from_pretrained failure." + }, + { + "left": "issue:43723", + "right": "issue:44977", + "accept": false, + "reason": "Tokenizer loading regression and Qwen3.5 flash-attention generation bug are different problems." + }, + { + "left": "issue:44222", + "right": "issue:44484", + "accept": false, + "reason": "Both mention save_pretrained, but one is FP8 MoE saving and the other is shard-size default behavior." + }, + { + "left": "issue:43232", + "right": "issue:45834", + "accept": false, + "reason": "Generation kwargs update after sync_gpus is unrelated to long OCR index errors." + }, + { + "left": "issue:43761", + "right": "issue:46082", + "accept": false, + "reason": "CLIPVisionModel hidden_states regression is unrelated to LlamaConfig head_dim validation." + }, + { + "left": "issue:44263", + "right": "issue:45092", + "accept": false, + "reason": "GlmMoeDsaIndexer torch.split issue is unrelated to remote-code checkpoint meta-init incompatibility." + }, + { + "left": "issue:44483", + "right": "issue:45406", + "accept": false, + "reason": "API request rejection and Gemma4Processor crash are different service-level failures." + }, + { + "left": "issue:44295", + "right": "issue:44623", + "accept": false, + "reason": "Position_ids buffer access error is unrelated to missing processor save artifacts." + }, + { + "left": "issue:44373", + "right": "issue:44485", + "accept": false, + "reason": "Wrong docstring for position_ids is not the same as the GLM-5 RoPE implementation discussion." + }, + { + "left": "issue:44514", + "right": "issue:45103", + "accept": false, + "reason": "Chat-template batching crash and auto-docstring annotation crash are unrelated." + }, + { + "left": "issue:43723", + "right": "issue:44871", + "accept": false, + "reason": "Tokenizer loading regression is unrelated to eos_token_id configuration mismatch." + }, + { + "left": "issue:45092", + "right": "issue:45357", + "accept": false, + "reason": "Old InternVL2 meta-init incompatibility is unrelated to Qwen3.5 save_pretrained visual-key regression." + }, + { + "left": "issue:44263", + "right": "issue:44360", + "accept": false, + "reason": "Different GLM-5 DSA topics: a torch.split bug versus a missing ReLU discussion." + }, + { + "left": "issue:30990", + "right": "issue:38175", + "accept": false, + "reason": "Sentence Transformers loading hang and SigLIP2 zero-probability outputs are unrelated." + }, + { + "left": "issue:42907", + "right": "issue:46050", + "accept": false, + "reason": "Dequantized model save failure is unrelated to sinq quantization breakage after save/load." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 17, + "review_unit_index": 40, + "review_unit_count": 50, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-428", + "nodes": [ + "issue:22355", + "issue:28218", + "issue:30990", + "issue:33357", + "issue:39401", + "issue:41628", + "issue:42371", + "issue:42757", + "issue:42831", + "issue:42886", + "issue:42907", + "issue:42994", + "issue:43066", + "issue:43122", + "issue:43232", + "issue:43479", + "issue:43576", + "issue:43646", + "issue:43723", + "issue:43756", + "issue:43784", + "issue:43824", + "issue:44038", + "issue:44060", + "issue:44117", + "issue:44261", + "issue:44263", + "issue:44295", + "issue:44297", + "issue:44373", + "issue:44442", + "issue:44521", + "issue:44568", + "issue:44671", + "issue:44704", + "issue:44743", + "issue:44829", + "issue:44913", + "issue:44945", + "issue:44991", + "issue:45092", + "issue:45278", + "issue:45290", + "issue:45310", + "issue:45593", + "issue:45753", + "issue:46002", + "issue:46132" + ], + "soft_pairs": [ + "issue:43723|issue:44991", + "issue:44829|issue:46002", + "issue:43723|issue:44442", + "issue:44261|issue:45092", + "issue:42831|issue:45310", + "issue:43479|issue:44913", + "issue:41628|issue:43824", + "issue:42831|issue:44038", + "issue:44263|issue:44373", + "issue:39401|issue:44568", + "issue:42757|issue:45278", + "issue:43232|issue:45593", + "issue:42886|issue:42994", + "issue:42831|issue:45753", + "issue:42886|issue:43066", + "issue:22355|issue:28218", + "issue:43756|issue:44060", + "issue:44671|issue:46132", + "issue:43122|issue:44521", + "issue:44117|issue:44704", + "issue:43576|issue:43784", + "issue:43646|issue:46002", + "issue:30990|issue:33357", + "issue:42371|issue:44263", + "issue:44295|issue:44297", + "issue:44704|issue:45290", + "issue:42907|issue:44945", + "issue:41628|issue:45278", + "issue:43232|issue:44743" + ], + "prepared_review_unit_hash": "54b43756e2184666db45bf5017d7e95f09833010e9cdac1f4901bb6c68b5db90", + "component_budget": { + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 29, + "serialized_chars": 35956, + "estimated_input_tokens": 8989, + "estimated_eval_tokens": 18234 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This set is mostly a grab bag of unrelated transformer regressions. The only recurring theme is tokenizer/import breakage in v5, but the soft pairs still point to different concrete failures, so I would not merge any of them.", + "confidence": 0.9, + "canonical_issue_reason": "Issue 43723 is the broadest tokenizer-loading regression in v5 and is the closest hub for the tokenizer-related reports, but the overall set is too heterogeneous for a true duplicate canonical.", + "canonical_pr_reason": null, + "best_issue_reason": "43723 is the most representative issue for the tokenizer-loading regression subset because it targets the generic AutoTokenizer.from_pretrained failure in v5.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43723", + "right": "issue:44991", + "accept": false, + "reason": "Both are tokenizer-loading complaints, but one is a generic v5 AutoTokenizer regression and the other is a model-specific failure for EMBEDDIA/est-roberta; not enough evidence of the same bug." + }, + { + "left": "issue:44829", + "right": "issue:46002", + "accept": false, + "reason": "Different problem domains: flash_attention_3 training degradation vs NaNs after checkout/init in classification heads." + }, + { + "left": "issue:43723", + "right": "issue:44442", + "accept": false, + "reason": "Shared surface area around AutoTokenizer, but one is a general v5 loading regression and the other is a specific FastSpeech2ConformerTokenizer export/loading failure." + }, + { + "left": "issue:44261", + "right": "issue:45092", + "accept": false, + "reason": "MLA layernorm precision/config issue is unrelated to InternVL2 remote-code meta-initialization incompatibility." + }, + { + "left": "issue:42831", + "right": "issue:45310", + "accept": false, + "reason": "FineGrainedFP8 accuracy regression and Qwen3.5-Moe from_pretrained error are distinct failures." + }, + { + "left": "issue:43479", + "right": "issue:44913", + "accept": false, + "reason": "Different configs and symptoms: Phi4Multimodal default-init bug vs GPTNeoX rotary_pct not persisting on reload." + }, + { + "left": "issue:41628", + "right": "issue:43824", + "accept": false, + "reason": "Both are import errors, but they concern different missing symbols and likely different registration/export paths." + }, + { + "left": "issue:42831", + "right": "issue:44038", + "accept": false, + "reason": "FineGrainedFP8 accuracy problems are not the same as Qwen3/VL-Moe load errors." + }, + { + "left": "issue:44263", + "right": "issue:44373", + "accept": false, + "reason": "The first is a code bug in torch.split handling; the second is just a wrong docstring." + }, + { + "left": "issue:39401", + "right": "issue:44568", + "accept": false, + "reason": "Qwen3 offset_mapping correctness and mdeberta add_special_tokens BOS/EOS omission are different tokenizer behaviors." + }, + { + "left": "issue:42757", + "right": "issue:45278", + "accept": false, + "reason": "Missing is_offline_mode from huggingface_hub is a specific dependency import issue; 45278 is a broad collection of import errors after upgrade." + }, + { + "left": "issue:43232", + "right": "issue:45593", + "accept": false, + "reason": "Generation state handling after sync_gpus is unrelated to D-FINE auxiliary-loss behavior when denoising is off." + }, + { + "left": "issue:42886", + "right": "issue:42994", + "accept": false, + "reason": "Offline tokenizer cache loading and quantized model saving are separate code paths and symptoms." + }, + { + "left": "issue:42831", + "right": "issue:45753", + "accept": false, + "reason": "FP8 accuracy regression does not match a Qwen3_5 infinite-loop bug on a specific image." + }, + { + "left": "issue:42886", + "right": "issue:43066", + "accept": false, + "reason": "Offline cache loading failure and wrong tokenizer decoder type in v5 are different tokenizer subsystems." + }, + { + "left": "issue:22355", + "right": "issue:28218", + "accept": false, + "reason": "Missing transformers.onnx is an import/package issue; added-token spacing is a tokenizer behavior bug." + }, + { + "left": "issue:43756", + "right": "issue:44060", + "accept": false, + "reason": "Smollm3 RoPE-layer dropping and Qwen3-Next tied-weights warning are unrelated model-architecture issues." + }, + { + "left": "issue:44671", + "right": "issue:46132", + "accept": false, + "reason": "CamemBERT MLM prediction regression is not the same as AttentionInterface.register changing callback behavior." + }, + { + "left": "issue:43122", + "right": "issue:44521", + "accept": false, + "reason": "Both involve tokenization, but one is a general tokenization drift across versions while the other is multimodal assistant_mask generation." + }, + { + "left": "issue:44117", + "right": "issue:44704", + "accept": false, + "reason": "TOKENIZER_MAPPING_NAMES None handling and AutoProcessor kwargs propagation are separate loader plumbing bugs." + }, + { + "left": "issue:43576", + "right": "issue:43784", + "accept": false, + "reason": "Broken transformers env command and a sentence-transformers import NameError are unrelated failures." + }, + { + "left": "issue:43646", + "right": "issue:46002", + "accept": false, + "reason": "Custom model initialization breakage is not the same as NaNs appearing in classification heads after checkout/init." + }, + { + "left": "issue:30990", + "right": "issue:33357", + "accept": false, + "reason": "Sentence Transformers loading stalls and a MacOS bus error on CLIP are different runtime failures." + }, + { + "left": "issue:42371", + "right": "issue:44263", + "accept": false, + "reason": "TF32 API guidance and a torch.split indexing bug do not describe the same issue." + }, + { + "left": "issue:44295", + "right": "issue:44297", + "accept": false, + "reason": "Registering position_ids as a buffer causing read errors is unrelated to tokenizer_class metadata mismatch on save." + }, + { + "left": "issue:44704", + "right": "issue:45290", + "accept": false, + "reason": "Passing kwargs to cached_file and crashing on tool-call assistant messages are different apply_chat_template/processor bugs." + }, + { + "left": "issue:42907", + "right": "issue:44945", + "accept": false, + "reason": "Saving dequantized models and incorrect output under pipeline parallelism are unrelated model-save vs inference-parallelism issues." + }, + { + "left": "issue:41628", + "right": "issue:45278", + "accept": false, + "reason": "AutoImageProcessor import failure is one symbol missing; 45278 reports many import errors after upgrade, so they are not the same concrete bug." + }, + { + "left": "issue:43232", + "right": "issue:44743", + "accept": false, + "reason": "Generation kwargs update after sync_gpus and Qwen3.5 recurrent-state reset are different generation-path bugs." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 17, + "review_unit_index": 41, + "review_unit_count": 50, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-428", + "nodes": [ + "issue:22355", + "issue:28218", + "issue:29159", + "issue:30990", + "issue:36296", + "issue:39401", + "issue:42175", + "issue:42673", + "issue:42757", + "issue:42907", + "issue:43116", + "issue:43122", + "issue:43278", + "issue:43408", + "issue:43452", + "issue:43475", + "issue:43493", + "issue:43784", + "issue:43824", + "issue:43937", + "issue:43976", + "issue:44038", + "issue:44222", + "issue:44261", + "issue:44295", + "issue:44297", + "issue:44351", + "issue:44360", + "issue:44373", + "issue:44485", + "issue:44488", + "issue:44514", + "issue:44521", + "issue:44625", + "issue:44779", + "issue:44829", + "issue:44938", + "issue:44945", + "issue:44964", + "issue:44977", + "issue:45072", + "issue:45200", + "issue:45216", + "issue:45446", + "issue:45478", + "issue:45600", + "issue:45753" + ], + "soft_pairs": [ + "issue:43122|issue:44514", + "issue:43976|issue:44945", + "issue:39401|issue:44488", + "issue:43784|issue:45600", + "issue:44261|issue:44373", + "issue:43408|issue:43475", + "issue:42907|issue:43937", + "issue:39401|issue:44521", + "issue:43116|issue:45200", + "issue:44625|issue:44829", + "issue:42907|issue:44964", + "issue:44360|issue:44485", + "issue:44038|issue:44297", + "issue:28218|issue:30990", + "issue:44038|issue:44295", + "issue:43122|issue:44779", + "issue:42907|issue:43452", + "issue:36296|issue:43116", + "issue:44038|issue:44222", + "issue:43824|issue:44351", + "issue:45216|issue:45478", + "issue:42757|issue:44351", + "issue:43784|issue:45446", + "issue:43278|issue:44945", + "issue:22355|issue:29159", + "issue:43122|issue:44977", + "issue:42175|issue:44938", + "issue:43493|issue:45753", + "issue:42673|issue:45072" + ], + "prepared_review_unit_hash": "992c2b8cd9aa81120e9c54b8b8ec04b9391b15c5d66fa0c264532c80ba2bd6ea", + "component_budget": { + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 29, + "serialized_chars": 35357, + "estimated_input_tokens": 8840, + "estimated_eval_tokens": 17936 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "Mostly unrelated issues across tokenization, model loading, and backend regressions. The Qwen3.5 save/load reports are thematically close, but #45216 and #45478 are not strong enough evidence of the same underlying bug to merge as duplicates.", + "confidence": 0.63, + "canonical_issue_reason": "Issue #45216 is the broadest and clearest representative of the Qwen3.5 5.4.0 save_pretrained serialization regression, but it should not absorb the separate from_pretrained error in #45478 without stronger body-level evidence of a shared root cause.", + "canonical_pr_reason": null, + "best_issue_reason": "#45216 is the best anchor for the Qwen3.5 save/load regression theme because it describes the core checkpoint corruption symptom most generally.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43122", + "right": "issue:44514", + "accept": false, + "reason": "Both involve tokenizer/chat-template behavior, but one is a tokenizer regression and the other is a batched apply_chat_template crash; different symptoms and code paths." + }, + { + "left": "issue:43976", + "right": "issue:44945", + "accept": false, + "reason": "A Python version compatibility issue and an incorrect output under pipeline parallelism are unrelated bugs." + }, + { + "left": "issue:39401", + "right": "issue:44488", + "accept": false, + "reason": "Wrong offset_mapping for Qwen3 tokenizer and failure to load a specific model are different problems; only the model family overlaps." + }, + { + "left": "issue:43784", + "right": "issue:45600", + "accept": false, + "reason": "Both are import/load failures, but one is a missing nn symbol in sentence-transformers integration and the other is stale auto_mappings references." + }, + { + "left": "issue:44261", + "right": "issue:44373", + "accept": false, + "reason": "A numerical precision/config issue in MLA and a wrong docstring for position_ids are unrelated." + }, + { + "left": "issue:43408", + "right": "issue:43475", + "accept": false, + "reason": "Both are SAM 3 related, but one is a model-type warning and the other is a missing attribute on a vision encoder output." + }, + { + "left": "issue:42907", + "right": "issue:43937", + "accept": false, + "reason": "Saving dequantized models and invalid GenerationConfig are distinct code paths and symptoms." + }, + { + "left": "issue:39401", + "right": "issue:44521", + "accept": false, + "reason": "Both touch tokenizer/multimodal processing, but wrong offset mapping and all-zero assistant masks are separate bugs." + }, + { + "left": "issue:43116", + "right": "issue:45200", + "accept": false, + "reason": "A classification example returning empty results and a Gemma 4 token-type-id requirement are not the same underlying issue." + }, + { + "left": "issue:44625", + "right": "issue:44829", + "accept": false, + "reason": "num_labels propagation and flash_attention_3 degenerate training are unrelated model configuration/performance bugs." + }, + { + "left": "issue:42907", + "right": "issue:44964", + "accept": false, + "reason": "Saving dequantized models and loading a multimodal Phi-4 model are different problems." + }, + { + "left": "issue:44360", + "right": "issue:44485", + "accept": false, + "reason": "DSA indexer activation logic and GLM-5 RoPE implementation are unrelated." + }, + { + "left": "issue:44038", + "right": "issue:44297", + "accept": false, + "reason": "A Qwen3-VL-Moe loading issue and tokenizer_class mismatch on save_pretrained are different regressions." + }, + { + "left": "issue:28218", + "right": "issue:30990", + "accept": false, + "reason": "Tokenizer adding an extra space and Sentence Transformers getting stuck loading do not describe the same bug." + }, + { + "left": "issue:44038", + "right": "issue:44295", + "accept": false, + "reason": "Qwen3-VL-Moe loading and reading position_ids after buffer registration are unrelated." + }, + { + "left": "issue:43122", + "right": "issue:44779", + "accept": false, + "reason": "Both are tokenizer regressions, but one is version-difference behavior and the other is Deepseek incorrect results; different models and failure modes." + }, + { + "left": "issue:42907", + "right": "issue:43452", + "accept": false, + "reason": "Saving dequantized models and gguf_file breaking AutoTokenizer/AutoModel loading are distinct issues." + }, + { + "left": "issue:36296", + "right": "issue:43116", + "accept": false, + "reason": "Tensor parallel training and a multi-label classification example returning empty results are unrelated." + }, + { + "left": "issue:44038", + "right": "issue:44222", + "accept": false, + "reason": "Qwen3-VL-Moe loading and FP8 save_pretrained for moe are both MoE-related but not the same code-path failure." + }, + { + "left": "issue:43824", + "right": "issue:44351", + "accept": false, + "reason": "Both are import errors, but they involve different missing exports and different fixes." + }, + { + "left": "issue:45216", + "right": "issue:45478", + "accept": false, + "reason": "The titles point to a related Qwen3.5 save/load theme, but one is a save_pretrained checkpoint corruption report and the other is a from_pretrained load error for Qwen3.5 MoE; that's not enough to prove the same bug." + }, + { + "left": "issue:42757", + "right": "issue:44351", + "accept": false, + "reason": "Missing is_offline_mode from huggingface_hub and missing HybridCache from transformers are separate import/export issues." + }, + { + "left": "issue:43784", + "right": "issue:45446", + "accept": false, + "reason": "Importing sentence-transformers with latest transformers and a PyTorch version check for AuxRequest are unrelated." + }, + { + "left": "issue:43278", + "right": "issue:44945", + "accept": false, + "reason": "Embedding dtype changing between training and evaluate is not the same as incorrect output under pipeline parallelism." + }, + { + "left": "issue:22355", + "right": "issue:29159", + "accept": false, + "reason": "No module named transformers.onnx and slow/fast tokenizer inconsistency are different historical bugs." + }, + { + "left": "issue:43122", + "right": "issue:44977", + "accept": false, + "reason": "Tokenizer version-difference behavior and Qwen3.5 flash-attention generation problems are unrelated." + }, + { + "left": "issue:42175", + "right": "issue:44938", + "accept": false, + "reason": "Missing TensorFlow backend in an install extra and Python 3.14 loading failure are different packaging/runtime issues." + }, + { + "left": "issue:43493", + "right": "issue:45753", + "accept": false, + "reason": "SigLIP2 HF/JAX discrepancy and Qwen3_5 infinite loop on a specific image are separate model bugs." + }, + { + "left": "issue:42673", + "right": "issue:45072", + "accept": false, + "reason": "VRAM leak in Qwen3ForCausalLM multithreaded loading and dtype mismatches in bfloat16 inference are unrelated." + } + ] + }, + "evaluator_result": { + "accept": false, + "feedback": "The overall summary is fine, but the accepted edge between issue:45216 and issue:45478 is too speculative for the packet. The titles only show a Qwen3.5 / 5.4.0 save_pretrained checkpoint corruption report and a separate from_pretrained load error for Qwen3.5 MoE; that is a plausible theme match, but not enough evidence to call them the same bugfix. A more conservative verdict would keep this pair as non-duplicate unless the bodies show a shared serialization root cause." + }, + "evaluator_used": true, + "retried": true, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 17, + "review_unit_index": 42, + "review_unit_count": 50, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-428", + "nodes": [ + "issue:22355", + "issue:29127", + "issue:33453", + "issue:38617", + "issue:39401", + "issue:42222", + "issue:42371", + "issue:42673", + "issue:43064", + "issue:43066", + "issue:43116", + "issue:43441", + "issue:43450", + "issue:43493", + "issue:43611", + "issue:43673", + "issue:43825", + "issue:43827", + "issue:43874", + "issue:43906", + "issue:43937", + "issue:43976", + "issue:43986", + "issue:43992", + "issue:44038", + "issue:44075", + "issue:44117", + "issue:44291", + "issue:44380", + "issue:44466", + "issue:44485", + "issue:44704", + "issue:44779", + "issue:44936", + "issue:44977", + "issue:45071", + "issue:45081", + "issue:45083", + "issue:45278", + "issue:45357", + "issue:45405", + "issue:45446", + "issue:45478", + "issue:45563", + "issue:45588", + "issue:46002", + "issue:46077" + ], + "soft_pairs": [ + "issue:43992|issue:44704", + "issue:43066|issue:44779", + "issue:38617|issue:45278", + "issue:22355|issue:29127", + "issue:43611|issue:46002", + "issue:43937|issue:44485", + "issue:43976|issue:45446", + "issue:42222|issue:43441", + "issue:43116|issue:43493", + "issue:42371|issue:43874", + "issue:44075|issue:44936", + "issue:43064|issue:43673", + "issue:45357|issue:45478", + "issue:44038|issue:45478", + "issue:43116|issue:43825", + "issue:45083|issue:45446", + "issue:43937|issue:45563", + "issue:44380|issue:46077", + "issue:33453|issue:39401", + "issue:44117|issue:45588", + "issue:43827|issue:45446", + "issue:39401|issue:44779", + "issue:42673|issue:43906", + "issue:43937|issue:44977", + "issue:44291|issue:45071", + "issue:43450|issue:43986", + "issue:44466|issue:46002", + "issue:45405|issue:45446", + "issue:44117|issue:45081" + ], + "prepared_review_unit_hash": "9262a5943d5f2df4cf9b9ca845cc446d74839acbbc5c6c595f35b0e2af74a07b", + "component_budget": { + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 29, + "serialized_chars": 35667, + "estimated_input_tokens": 8917, + "estimated_eval_tokens": 18090 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "These issues are largely unrelated regressions across different models, APIs, and failure modes. None of the soft pairs look like the same underlying bug or change, so I would not merge any of them as duplicates.", + "confidence": 0.93, + "canonical_issue_reason": "issue:45278 is the broadest representative item in the set: it describes multiple import errors after upgrading to 5.5.0, which makes it the most central umbrella issue among these otherwise distinct reports.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45278 is the best single issue to anchor the cluster because it captures a broad upgrade regression and is more representative than the model-specific or narrowly scoped reports.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43992", + "right": "issue:44704", + "accept": false, + "reason": "Different code paths: UMT5Encoder weight loading vs AutoProcessor forwarding kwargs to cached_file." + }, + { + "left": "issue:43066", + "right": "issue:44779", + "accept": false, + "reason": "Both tokenizer-related, but one is wrong decoder type and the other is incorrect Deepseek tokenization results; not the same bug." + }, + { + "left": "issue:38617", + "right": "issue:45278", + "accept": false, + "reason": "One is a specific missing import, the other is a broad set of import errors after upgrade; related theme but not the same underlying defect." + }, + { + "left": "issue:22355", + "right": "issue:29127", + "accept": false, + "reason": "Completely different problems: missing transformers.onnx module vs a confusing layoutlmv3 error message." + }, + { + "left": "issue:43611", + "right": "issue:46002", + "accept": false, + "reason": "Loading with base_model_prefix broke, but the other report is NaNs in classification heads after checkout/init; unrelated failure modes." + }, + { + "left": "issue:43937", + "right": "issue:44485", + "accept": false, + "reason": "Both mention GLM-5, but one is invalid GenerationConfig and the other is RoPE implementation discussion; different bugs." + }, + { + "left": "issue:43976", + "right": "issue:45446", + "accept": false, + "reason": "Python version compatibility regression vs an incorrect PyTorch version check in flex_attention; unrelated." + }, + { + "left": "issue:42222", + "right": "issue:43441", + "accept": false, + "reason": "VitPose model breakage and Ministral FlashAttention failure are different model-specific issues." + }, + { + "left": "issue:43116", + "right": "issue:43493", + "accept": false, + "reason": "Example-script multi-label classification behavior is unrelated to SigLIP2 implementation discrepancy." + }, + { + "left": "issue:42371", + "right": "issue:43874", + "accept": false, + "reason": "TF32 settings and a missing image-processor method are unrelated." + }, + { + "left": "issue:44075", + "right": "issue:44936", + "accept": false, + "reason": "Optimizer argument handling and trainer.evaluate() after train are different trainer/optimizer issues." + }, + { + "left": "issue:43064", + "right": "issue:43673", + "accept": false, + "reason": "Distributed optimizer-state corruption and missing GenerationMixin cache are unrelated code paths." + }, + { + "left": "issue:45357", + "right": "issue:45478", + "accept": false, + "reason": "Both concern Qwen3.5, but save_pretrained key serialization and from_pretrained loading errors are different bugs." + }, + { + "left": "issue:44038", + "right": "issue:45478", + "accept": false, + "reason": "Same general model family, but Qwen3-VL-Moe vs Qwen3.5 MoE from_pretrained are not the same failure." + }, + { + "left": "issue:43116", + "right": "issue:43825", + "accept": false, + "reason": "Classification example output bug vs pipeline translation-task messaging are unrelated." + }, + { + "left": "issue:45083", + "right": "issue:45446", + "accept": false, + "reason": "A feature-extraction helper issue and a PyTorch version check bug are unrelated." + }, + { + "left": "issue:43937", + "right": "issue:45563", + "accept": false, + "reason": "GenerationConfig invalidity and a stale warning for num_return_sequences are different issues." + }, + { + "left": "issue:44380", + "right": "issue:46077", + "accept": false, + "reason": "Attention-scaling ignored in GPT2 backend vs Gemma4 bidirectional mask construction are different model/backend bugs." + }, + { + "left": "issue:33453", + "right": "issue:39401", + "accept": false, + "reason": "Both tokenizer-related regressions, but generic loading regression vs Qwen3 offset_mapping bug are not the same defect." + }, + { + "left": "issue:44117", + "right": "issue:45588", + "accept": false, + "reason": "Tokenizer mapping None handling and flash_attention crashing on s_aux=None are unrelated." + }, + { + "left": "issue:43827", + "right": "issue:45446", + "accept": false, + "reason": "Documentation references to pipeline() vs a runtime version check bug have no shared underlying code path." + }, + { + "left": "issue:39401", + "right": "issue:44779", + "accept": false, + "reason": "Different model-specific tokenizer regressions: Qwen3 offset mapping vs Deepseek incorrect results." + }, + { + "left": "issue:42673", + "right": "issue:43906", + "accept": false, + "reason": "VRAM leak in Qwen3ForCausalLM threads is not the same as a reproduction of issue 38071." + }, + { + "left": "issue:43937", + "right": "issue:44977", + "accept": false, + "reason": "Different GLM/Qwen generation failures with different symptoms and likely different causes." + }, + { + "left": "issue:44291", + "right": "issue:45071", + "accept": false, + "reason": "init_empty_weights argument incompatibility is unrelated to PretrainedConfig type-checking breakage." + }, + { + "left": "issue:43450", + "right": "issue:43986", + "accept": false, + "reason": "Batched video processor shape bug vs missing torchvision crash when loading via AutoProcessor are distinct issues." + }, + { + "left": "issue:44466", + "right": "issue:46002", + "accept": false, + "reason": "Serialization of tied lm_head weights and NaNs in classification heads are unrelated." + }, + { + "left": "issue:45405", + "right": "issue:45446", + "accept": false, + "reason": "A PEFT minimum-version bump issue is unrelated to the flex_attention import version check." + }, + { + "left": "issue:44117", + "right": "issue:45081", + "accept": false, + "reason": "Tokenizer mapping fallback handling and a Mistral regex patch crash are different tokenizer bugs." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 17, + "review_unit_index": 43, + "review_unit_count": 50, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-428", + "nodes": [ + "issue:38617", + "issue:39401", + "issue:41628", + "issue:42175", + "issue:42222", + "issue:42371", + "issue:42490", + "issue:42617", + "issue:42673", + "issue:42831", + "issue:43012", + "issue:43066", + "issue:43097", + "issue:43122", + "issue:43408", + "issue:43421", + "issue:43441", + "issue:43577", + "issue:43582", + "issue:43673", + "issue:43742", + "issue:43874", + "issue:43937", + "issue:44075", + "issue:44220", + "issue:44261", + "issue:44263", + "issue:44360", + "issue:44410", + "issue:44483", + "issue:44484", + "issue:44485", + "issue:44493", + "issue:44496", + "issue:44829", + "issue:44933", + "issue:45071", + "issue:45216", + "issue:45278", + "issue:45310", + "issue:45341", + "issue:45464", + "issue:45478", + "issue:45542", + "issue:45561", + "issue:45656", + "issue:46172", + "issue:46178" + ], + "soft_pairs": [ + "issue:45341|issue:45561", + "issue:42222|issue:46178", + "issue:38617|issue:41628", + "issue:44410|issue:45310", + "issue:39401|issue:43421", + "issue:44829|issue:45216", + "issue:42617|issue:45341", + "issue:44484|issue:45310", + "issue:42673|issue:43582", + "issue:42490|issue:42673", + "issue:43408|issue:43742", + "issue:42371|issue:44261", + "issue:44933|issue:45278", + "issue:44483|issue:45464", + "issue:38617|issue:44220", + "issue:43937|issue:44496", + "issue:42222|issue:45542", + "issue:43066|issue:43122", + "issue:42175|issue:42222", + "issue:42371|issue:42831", + "issue:44075|issue:45656", + "issue:43673|issue:45341", + "issue:44410|issue:45478", + "issue:43012|issue:43577", + "issue:44484|issue:45478", + "issue:43097|issue:44493", + "issue:41628|issue:45071", + "issue:43874|issue:44261", + "issue:44263|issue:44485", + "issue:45071|issue:45542", + "issue:44263|issue:46172", + "issue:44360|issue:46172", + "issue:43012|issue:43408", + "issue:43441|issue:44829" + ], + "prepared_review_unit_hash": "573ac12730a34a75c005a551453139f4b133089737cba965ec9930cc7ce31a15", + "component_budget": { + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 34, + "serialized_chars": 37402, + "estimated_input_tokens": 9351, + "estimated_eval_tokens": 18958 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is heterogeneous: most items are distinct import, tokenizer, model-loading, backend, or training bugs rather than true duplicates. None of the soft pairs look like the same concrete issue, so all are rejected.", + "confidence": 0.8, + "canonical_issue_reason": "issue:45278 is the broadest umbrella in the set ('Many import errors after update from 4.57.0 to 5.5.0') and is the closest representative of the import-breakage subset, though the cluster overall is not a single duplicate group.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45278 is the most representative high-level regression report and the best anchor if one issue must be chosen from this mixed set.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45341", + "right": "issue:45561", + "accept": false, + "reason": "Both mention testing_utils, but one is a vague bug report and the other is a specific xdist race on captured_info.txt; different failures." + }, + { + "left": "issue:42222", + "right": "issue:46178", + "accept": false, + "reason": "Different subsystems: vitpose model breakage vs backend-mapping error for tensorflow_text." + }, + { + "left": "issue:38617", + "right": "issue:41628", + "accept": false, + "reason": "Both are import errors, but for different missing symbols and code paths." + }, + { + "left": "issue:44410", + "right": "issue:45310", + "accept": false, + "reason": "Different Qwen issues: missing layer projections vs a Qwen3.5 MoE from_pretrained error." + }, + { + "left": "issue:39401", + "right": "issue:43421", + "accept": false, + "reason": "Both are tokenizer-related, but wrong offset_mapping and runtime post-processor updates are different problems." + }, + { + "left": "issue:44829", + "right": "issue:45216", + "accept": false, + "reason": "FlashAttention-3 degenerate training and Qwen3.5 save_pretrained checkpoint corruption are unrelated regressions." + }, + { + "left": "issue:42617", + "right": "issue:45341", + "accept": false, + "reason": "A script failure and a testing_utils bug are unrelated." + }, + { + "left": "issue:44484", + "right": "issue:45310", + "accept": false, + "reason": "Shard-size behavior in save_pretrained is unrelated to the Qwen3.5 from_pretrained regression." + }, + { + "left": "issue:42673", + "right": "issue:43582", + "accept": false, + "reason": "VRAM leakage in multi-threaded loading and an Apple Silicon TypeError are distinct issues." + }, + { + "left": "issue:42490", + "right": "issue:42673", + "accept": false, + "reason": "Trailing-slash path warning vs VRAM leak under multiple dataloader threads are different bugs." + }, + { + "left": "issue:43408", + "right": "issue:43742", + "accept": false, + "reason": "Model-type mismatch warning for sam3_tracker is unrelated to the MobileLLM key error." + }, + { + "left": "issue:42371", + "right": "issue:44261", + "accept": false, + "reason": "TF32 settings warning and MLA rms_norm_eps precision error are different problems." + }, + { + "left": "issue:44933", + "right": "issue:45278", + "accept": false, + "reason": "Both are import-related, but missing image_utils import and broad 5.5.0 import errors are not the same concrete failure." + }, + { + "left": "issue:44483", + "right": "issue:45464", + "accept": false, + "reason": "Different API failures: generic chat/completions rejection vs Qwen3.5 streaming inference failure." + }, + { + "left": "issue:38617", + "right": "issue:44220", + "accept": false, + "reason": "ImportError for layer_type_validation is unrelated to _torch_extract_fbank_features()." + }, + { + "left": "issue:43937", + "right": "issue:44496", + "accept": false, + "reason": "Invalid GenerationConfig for GLM-5 is different from unrecognized model config.json." + }, + { + "left": "issue:42222", + "right": "issue:45542", + "accept": false, + "reason": "vitpose model breakage and missing TensorFlow backend are unrelated." + }, + { + "left": "issue:43066", + "right": "issue:43122", + "accept": false, + "reason": "Tokenizer decoder type mismatch and changed tokenization output are distinct regressions." + }, + { + "left": "issue:42175", + "right": "issue:42222", + "accept": false, + "reason": "TensorFlow extra/backend packaging issue vs vitpose model breakage." + }, + { + "left": "issue:42371", + "right": "issue:42831", + "accept": false, + "reason": "TF32 API warning is unrelated to FineGrainedFP8 accuracy." + }, + { + "left": "issue:44075", + "right": "issue:45656", + "accept": false, + "reason": "Optimizer SGD args ignored and optimizer step called twice under deepspeed are different optimizer bugs." + }, + { + "left": "issue:43673", + "right": "issue:45341", + "accept": false, + "reason": "Generation cache missing in v5 RC is unrelated to the testing_utils bug." + }, + { + "left": "issue:44410", + "right": "issue:45478", + "accept": false, + "reason": "Missing projections and Qwen3.5 from_pretrained error are different code-path failures." + }, + { + "left": "issue:43012", + "right": "issue:43577", + "accept": false, + "reason": "bfloat16 precision warning and BLIP2 dtype propagation bug are different dtype issues." + }, + { + "left": "issue:44484", + "right": "issue:45478", + "accept": false, + "reason": "Shard-size question and Qwen3.5 from_pretrained regression are unrelated." + }, + { + "left": "issue:43097", + "right": "issue:44493", + "accept": false, + "reason": "Removed tie_embeddings_and_encoder_decoder API and unexpected key with position id are separate regressions." + }, + { + "left": "issue:41628", + "right": "issue:45071", + "accept": false, + "reason": "AutoImageProcessor ImportError and PreTrainedConfig type-checking breakage are unrelated." + }, + { + "left": "issue:43874", + "right": "issue:44261", + "accept": false, + "reason": "Missing image patch-count helper and MLA eps precision bug are different issues." + }, + { + "left": "issue:44263", + "right": "issue:44485", + "accept": false, + "reason": "torch.split return handling in an indexer and GLM-5 RoPE implementation are unrelated." + }, + { + "left": "issue:45071", + "right": "issue:45542", + "accept": false, + "reason": "Type-checking breakage is unrelated to the missing TensorFlow backend error." + }, + { + "left": "issue:44263", + "right": "issue:46172", + "accept": false, + "reason": "DSA indexer split handling and deepseek-v4 transpose/dtype issue are unrelated." + }, + { + "left": "issue:44360", + "right": "issue:46172", + "accept": false, + "reason": "DSA indexer lacking ReLU and deepseek-v4 comb transpose bug are different problems." + }, + { + "left": "issue:43012", + "right": "issue:43408", + "accept": false, + "reason": "bfloat16 warning and sam3 model-type mismatch are different model-loading problems." + }, + { + "left": "issue:43441", + "right": "issue:44829", + "accept": false, + "reason": "FlashAttention failure in Ministral-3 and flash_attention_3 training degeneration are unrelated." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 17, + "review_unit_index": 44, + "review_unit_count": 50, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-428", + "nodes": [ + "issue:38617", + "issue:42757", + "issue:43064", + "issue:43066", + "issue:43097", + "issue:43208", + "issue:43335", + "issue:43408", + "issue:43452", + "issue:43502", + "issue:43531", + "issue:43550", + "issue:43673", + "issue:43824", + "issue:43825", + "issue:43873", + "issue:43874", + "issue:43906", + "issue:43931", + "issue:43986", + "issue:43992", + "issue:44038", + "issue:44062", + "issue:44117", + "issue:44261", + "issue:44263", + "issue:44315", + "issue:44351", + "issue:44368", + "issue:44410", + "issue:44484", + "issue:44485", + "issue:44655", + "issue:44704", + "issue:44938", + "issue:44987", + "issue:45003", + "issue:45071", + "issue:45103", + "issue:45341", + "issue:45372", + "issue:45375", + "issue:45440", + "issue:45588", + "issue:46050", + "issue:46123", + "issue:46172", + "issue:46178" + ], + "soft_pairs": [ + "issue:43931|issue:44410", + "issue:43931|issue:44038", + "issue:43335|issue:45071", + "issue:43673|issue:46123", + "issue:43097|issue:43452", + "issue:45003|issue:45341", + "issue:43986|issue:45341", + "issue:44117|issue:44987", + "issue:43408|issue:44368", + "issue:43986|issue:44484", + "issue:43874|issue:44263", + "issue:43986|issue:45372", + "issue:45440|issue:46172", + "issue:43531|issue:45341", + "issue:38617|issue:42757", + "issue:44261|issue:44485", + "issue:43064|issue:44315", + "issue:43097|issue:46050", + "issue:43097|issue:45588", + "issue:44062|issue:44117", + "issue:44410|issue:45375", + "issue:38617|issue:44351", + "issue:44038|issue:44410", + "issue:43502|issue:43992", + "issue:44704|issue:45103", + "issue:43097|issue:43873", + "issue:43986|issue:44117", + "issue:43824|issue:43986", + "issue:43066|issue:43906", + "issue:44938|issue:46178", + "issue:43208|issue:43550", + "issue:43992|issue:44484", + "issue:43825|issue:44655" + ], + "prepared_review_unit_hash": "dccbd5bd37c7d5b432acbef53af0a81bca7ef2d960739c78fbd39d4bd2d5fc0c", + "component_budget": { + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 33, + "serialized_chars": 37511, + "estimated_input_tokens": 9378, + "estimated_eval_tokens": 19012 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is a grab bag of mostly unrelated Transformers regression reports, many centered on v5 migration breakages. None of the soft pairs look like the same underlying bug or change, so all are rejected as duplicates.", + "confidence": 0.91, + "canonical_issue_reason": "Issue 43097 is the broadest, clearest v5 regression report in the set and reads like a representative umbrella complaint about a breaking API removal; the other issues are separate bugs in different code paths.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 43097 is the best single representative of the cluster\u2019s general v5-breaking-change theme, but it should not be treated as a duplicate target for the other reports.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43931", + "right": "issue:44410", + "accept": false, + "reason": "Both are Qwen-related loading/model-structure bugs, but they describe different failures: weight-shape mismatch vs missing projections in a different architecture." + }, + { + "left": "issue:43931", + "right": "issue:44038", + "accept": false, + "reason": "Same broad model family, but the reported failures are distinct Qwen-VL issues with different symptoms and likely different code paths." + }, + { + "left": "issue:43335", + "right": "issue:45071", + "accept": false, + "reason": "SwitchTransformers sparse-layer construction and PreTrainedConfig type checking are unrelated bugs." + }, + { + "left": "issue:43673", + "right": "issue:46123", + "accept": false, + "reason": "Chunked prefill cache handling and MaskGenerationPipeline final-batch handling are different generation/pipeline defects." + }, + { + "left": "issue:43097", + "right": "issue:43452", + "accept": false, + "reason": "Both are v5 breakage reports, but one is about a removed config helper and the other about tokenizer/model loading for gguf_file; not the same bug." + }, + { + "left": "issue:45003", + "right": "issue:45341", + "accept": false, + "reason": "Unsafe sys.modules access and a testing_utils bug are unrelated." + }, + { + "left": "issue:43986", + "right": "issue:45341", + "accept": false, + "reason": "Video-model loading without torchvision and a testing utility issue do not share the same failure path." + }, + { + "left": "issue:44117", + "right": "issue:44987", + "accept": false, + "reason": "TOKENIZER_MAPPING_NAMES returning None and a model-loading failure for physical-intelligence/fast are not the same underlying defect." + }, + { + "left": "issue:43408", + "right": "issue:44368", + "accept": false, + "reason": "These are different warning/regression reports involving different model families and different config fields." + }, + { + "left": "issue:43986", + "right": "issue:44484", + "accept": false, + "reason": "AutoProcessor/torchvision loading crash is unrelated to save_pretrained shard-size behavior." + }, + { + "left": "issue:43874", + "right": "issue:44263", + "accept": false, + "reason": "Different GLM bugs in different components: image processor missing a method versus torch.split return handling." + }, + { + "left": "issue:43986", + "right": "issue:45372", + "accept": false, + "reason": "Missing torchvision and missing ReasoningEffort are separate import/dependency problems." + }, + { + "left": "issue:45440", + "right": "issue:46172", + "accept": false, + "reason": "Both are DeepSeek-family issues, but one is an implementation divergence and the other is a specific tensor transpose expression; not clearly the same bug." + }, + { + "left": "issue:43531", + "right": "issue:45341", + "accept": false, + "reason": "Qwen3-MoE sliding_window behavior and a testing_utils bug are unrelated." + }, + { + "left": "issue:38617", + "right": "issue:42757", + "accept": false, + "reason": "Both are import errors, but they involve different missing symbols from different modules and different breakages." + }, + { + "left": "issue:44261", + "right": "issue:44485", + "accept": false, + "reason": "MLA layer-norm epsilon precision and GLM RoPE implementation are different numerical/modeling issues." + }, + { + "left": "issue:43064", + "right": "issue:44315", + "accept": false, + "reason": "Trainer/FSDP2 optimizer-state corruption and Liger Kernel not being applied are separate training-path bugs." + }, + { + "left": "issue:43097", + "right": "issue:46050", + "accept": false, + "reason": "A removed tie_embeddings_and_encoder_decoder API and a quantized-model save/load regression are different failure modes." + }, + { + "left": "issue:43097", + "right": "issue:45588", + "accept": false, + "reason": "A v5 API removal and a flash-attention sinkless-model crash do not describe the same code-path problem." + }, + { + "left": "issue:44062", + "right": "issue:44117", + "accept": false, + "reason": "AddedToken keyword duplication and tokenizer mapping None are tokenizer-related but distinct defects." + }, + { + "left": "issue:44410", + "right": "issue:45375", + "accept": false, + "reason": "Different Qwen config/model bugs: missing projections versus a missing vision config field." + }, + { + "left": "issue:38617", + "right": "issue:44351", + "accept": false, + "reason": "Different missing imports from different namespaces; these are not the same underlying regression." + }, + { + "left": "issue:44038", + "right": "issue:44410", + "accept": false, + "reason": "Both involve Qwen models, but one is a VL-MoE bug and the other is a missing-projections issue in a different model variant." + }, + { + "left": "issue:43502", + "right": "issue:43992", + "accept": false, + "reason": "Local_files_only network leakage and UMT5Encoder weight loading are separate from_pretrained issues with different causes." + }, + { + "left": "issue:44704", + "right": "issue:45103", + "accept": false, + "reason": "AutoProcessor kwargs propagation and auto_docstring annotation handling are unrelated." + }, + { + "left": "issue:43097", + "right": "issue:43873", + "accept": false, + "reason": "A removed v5 API and quantization offloading behavior are different classes of bugs." + }, + { + "left": "issue:43986", + "right": "issue:44117", + "accept": false, + "reason": "Video-model loading without torchvision and tokenizer mapping assumptions are separate loading paths." + }, + { + "left": "issue:43824", + "right": "issue:43986", + "accept": false, + "reason": "Different processor/model import failures for different model families." + }, + { + "left": "issue:43066", + "right": "issue:43906", + "accept": false, + "reason": "Wrong tokenizer decoder type and an isolated reproduction issue are not the same bug." + }, + { + "left": "issue:44938", + "right": "issue:46178", + "accept": false, + "reason": "Python 3.14 import/load failure and an unsupported tensorflow_text backend error are unrelated environment problems." + }, + { + "left": "issue:43208", + "right": "issue:43550", + "accept": false, + "reason": "xLSTM training blockers and Bamba torch.compile/SDPA failure are different model/backend bugs." + }, + { + "left": "issue:43992", + "right": "issue:44484", + "accept": false, + "reason": "Missing embed_tokens.weight and default max_shard_size are unrelated save/load behaviors." + }, + { + "left": "issue:43825", + "right": "issue:44655", + "accept": false, + "reason": "Pipeline translation error messaging and pipeline save_pretrained support are different pipeline features." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 17, + "review_unit_index": 45, + "review_unit_count": 50, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-428", + "nodes": [ + "issue:36683", + "issue:38617", + "issue:42490", + "issue:42757", + "issue:42898", + "issue:42907", + "issue:42994", + "issue:43012", + "issue:43064", + "issue:43066", + "issue:43208", + "issue:43299", + "issue:43352", + "issue:43475", + "issue:43502", + "issue:43526", + "issue:43550", + "issue:43582", + "issue:43618", + "issue:43756", + "issue:43761", + "issue:43792", + "issue:43824", + "issue:43867", + "issue:43901", + "issue:43976", + "issue:44038", + "issue:44162", + "issue:44222", + "issue:44230", + "issue:44368", + "issue:44380", + "issue:44483", + "issue:44485", + "issue:44514", + "issue:44655", + "issue:44683", + "issue:45276", + "issue:45290", + "issue:45381", + "issue:45478", + "issue:45479", + "issue:45698", + "issue:45753", + "issue:46077", + "issue:46123", + "issue:46132", + "issue:46172" + ], + "soft_pairs": [ + "issue:43012|issue:43550", + "issue:43012|issue:44368", + "issue:42907|issue:45753", + "issue:42907|issue:44655", + "issue:43352|issue:44380", + "issue:44162|issue:46172", + "issue:43352|issue:43976", + "issue:43792|issue:46123", + "issue:44380|issue:45479", + "issue:44380|issue:46132", + "issue:43064|issue:44380", + "issue:44483|issue:44514", + "issue:44222|issue:45478", + "issue:36683|issue:43582", + "issue:43012|issue:44683", + "issue:43208|issue:43526", + "issue:42490|issue:43012", + "issue:43618|issue:43761", + "issue:42757|issue:43824", + "issue:45698|issue:46172", + "issue:43618|issue:44380", + "issue:38617|issue:43502", + "issue:44038|issue:44230", + "issue:44483|issue:45290", + "issue:44483|issue:45381", + "issue:43756|issue:44485", + "issue:43299|issue:43867", + "issue:36683|issue:43475", + "issue:45276|issue:46077", + "issue:43901|issue:44655", + "issue:42994|issue:44655", + "issue:42898|issue:43066", + "issue:43299|issue:45753" + ], + "prepared_review_unit_hash": "e67c0f3c353fb1e5b67f8b538fe2638ad28ae3f7bfffc7ab2a237dfa5afb3179", + "component_budget": { + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 33, + "serialized_chars": 37291, + "estimated_input_tokens": 9323, + "estimated_eval_tokens": 18902 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "These items are not a true duplicate cluster; they span unrelated regressions across model loading, saving, compilation, tokenization, and API behavior. No soft edge looks safe to merge, so the best outcome is to keep them separate and use one broad issue only as a loose cluster anchor.", + "confidence": 0.37, + "canonical_issue_reason": "issue:44483 is the broadest, highest-impact regression in the set and the most plausible cluster anchor by title and severity; the rest are mostly model-specific or orthogonal failures.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:44483 is the best single representative issue if one must choose one, but it should not be treated as a duplicate magnet for the other items.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43012", + "right": "issue:43550", + "accept": false, + "reason": "Different failures: a PyTorch precision warning during compile versus a Bamba SDPA torch.compile error." + }, + { + "left": "issue:43012", + "right": "issue:44368", + "accept": false, + "reason": "Unrelated problems: bfloat16 compile warning versus a config warning about tie_word_embeddings in Qwen3.5 fine-tuning." + }, + { + "left": "issue:42907", + "right": "issue:45753", + "accept": false, + "reason": "Both involve model saving/loading, but one is dequantized Ministral/Devstral save failure and the other is an image-specific infinite-loop inference bug." + }, + { + "left": "issue:42907", + "right": "issue:44655", + "accept": false, + "reason": "Saving a dequantized model is not the same as saving Pipeline objects; different code paths and failure modes." + }, + { + "left": "issue:43352", + "right": "issue:44380", + "accept": false, + "reason": "Flash Attention support error for NemotronH versus GPT2 attention scaling being ignored under SDPA/FlashAttention are distinct bugs." + }, + { + "left": "issue:44162", + "right": "issue:46172", + "accept": false, + "reason": "ESM2 breakage is a model-loading/runtime issue, while deepseek-v4 hits a specific tensor transpose bug." + }, + { + "left": "issue:43352", + "right": "issue:43976", + "accept": false, + "reason": "Nemotron Flash Attention support and Python version compatibility are unrelated." + }, + { + "left": "issue:43792", + "right": "issue:46123", + "accept": false, + "reason": "Whisper inference failure is unrelated to MaskGenerationPipeline dropping the final partial batch." + }, + { + "left": "issue:44380", + "right": "issue:45479", + "accept": false, + "reason": "Attention scaling being ignored is different from a degenerate zero-loss bug in sequence classification." + }, + { + "left": "issue:44380", + "right": "issue:46132", + "accept": false, + "reason": "GPT2 backend-specific attention scaling is unrelated to AttentionInterface.register changing callback behavior." + }, + { + "left": "issue:43064", + "right": "issue:44380", + "accept": false, + "reason": "Trainer/FSDP2 optimizer-state corruption is a distributed training bug, not an attention-backend config regression." + }, + { + "left": "issue:44483", + "right": "issue:44514", + "accept": false, + "reason": "Chat/completions request rejection is unrelated to Qwen2_5_VLProcessor crashing on batched chat templates." + }, + { + "left": "issue:44222", + "right": "issue:45478", + "accept": false, + "reason": "Both mention FP8/quantization and MoE, but one is save_pretrained failure and the other is from_pretrained loading error." + }, + { + "left": "issue:36683", + "right": "issue:43582", + "accept": false, + "reason": "Gemma3Config missing vocab_size and Apple Silicon caching_allocator_warmup TypeError are unrelated." + }, + { + "left": "issue:43012", + "right": "issue:44683", + "accept": false, + "reason": "A warning during bfloat16 compile is not the same as flex_attention failing on torch >= 2.9." + }, + { + "left": "issue:43208", + "right": "issue:43526", + "accept": false, + "reason": "xLSTM training bugs and BeitImageProcessorFast reduce_labels truncation are different subsystems and bugs." + }, + { + "left": "issue:42490", + "right": "issue:43012", + "accept": false, + "reason": "Path-trailing-slash import warning differs completely from a compile-time PyTorch precision warning." + }, + { + "left": "issue:43618", + "right": "issue:43761", + "accept": false, + "reason": "CLIP attentions being unassigned and CLIPVisionModel returning hidden_states=None are separate v5 regressions." + }, + { + "left": "issue:42757", + "right": "issue:43824", + "accept": false, + "reason": "Missing is_offline_mode and missing Qwen2_5_VLForConditionalGeneration are different import errors from different symbols." + }, + { + "left": "issue:45698", + "right": "issue:46172", + "accept": false, + "reason": "A custom-module reload bug after save_pretrained is unrelated to a low-level tensor transpose bug in deepseek-v4." + }, + { + "left": "issue:43618", + "right": "issue:44380", + "accept": false, + "reason": "CLIP output attentions regression and GPT2 attention scaling behavior are distinct model/backend issues." + }, + { + "left": "issue:38617", + "right": "issue:43502", + "accept": false, + "reason": "Missing layer_type_validation import and local_files_only still making API requests are unrelated import/network bugs." + }, + { + "left": "issue:44038", + "right": "issue:44230", + "accept": false, + "reason": "Both involve Qwen3-VL-Moe, but one is a generic transformers loading bug and the other is FP8 inference support." + }, + { + "left": "issue:44483", + "right": "issue:45290", + "accept": false, + "reason": "Chat/completions request handling is unrelated to apply_chat_template crashing on assistant tool-call messages." + }, + { + "left": "issue:44483", + "right": "issue:45381", + "accept": false, + "reason": "API request acceptance and Qwen2.5-VL video vision_position_ids are different failure modes in different code paths." + }, + { + "left": "issue:43756", + "right": "issue:44485", + "accept": false, + "reason": "Smollm3 RoPE layer dropping and GLM-5 RoPE implementation discussion are not the same concrete bug." + }, + { + "left": "issue:43299", + "right": "issue:43867", + "accept": false, + "reason": "Qwen3VL-Moe loading breakage and state_dict sorting load errors are different model-loading problems." + }, + { + "left": "issue:36683", + "right": "issue:43475", + "accept": false, + "reason": "Gemma3Config vocab_size AttributeError and SAM 3 Video missing fpn_position_embeddings are unrelated missing-attribute bugs." + }, + { + "left": "issue:45276", + "right": "issue:46077", + "accept": false, + "reason": "Gemma4 token embedding resize behavior and bidirectional attention mask construction are different bugs in different parts of the model." + }, + { + "left": "issue:43901", + "right": "issue:44655", + "accept": false, + "reason": "Deprecated docs wording for return_all_scores is unrelated to pipeline save_pretrained failure." + }, + { + "left": "issue:42994", + "right": "issue:44655", + "accept": false, + "reason": "Quantized model saving failure and pipeline serialization failure are different save_pretrained bugs." + }, + { + "left": "issue:42898", + "right": "issue:43066", + "accept": false, + "reason": "Tokenizer clean_up_tokenization_spaces behavior changes and wrong decoder type are unrelated v5 issues." + }, + { + "left": "issue:43299", + "right": "issue:45753", + "accept": false, + "reason": "Qwen3.5-VL/Moe loading failure and a specific-image infinite loop are different bugs with different triggers." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 17, + "review_unit_index": 46, + "review_unit_count": 50, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-428", + "nodes": [ + "issue:28282", + "issue:33453", + "issue:36010", + "issue:36296", + "issue:36683", + "issue:41669", + "issue:42490", + "issue:42757", + "issue:42890", + "issue:43010", + "issue:43064", + "issue:43352", + "issue:43525", + "issue:43575", + "issue:43611", + "issue:43618", + "issue:43867", + "issue:43931", + "issue:43937", + "issue:43976", + "issue:43992", + "issue:44075", + "issue:44077", + "issue:44242", + "issue:44246", + "issue:44355", + "issue:44360", + "issue:44380", + "issue:44484", + "issue:44492", + "issue:44625", + "issue:44655", + "issue:44855", + "issue:44857", + "issue:44938", + "issue:45071", + "issue:45216", + "issue:45278", + "issue:45310", + "issue:45357", + "issue:45375", + "issue:45440", + "issue:45478", + "issue:45561", + "issue:45593", + "issue:45850", + "issue:46077", + "issue:46082" + ], + "soft_pairs": [ + "issue:36010|issue:42757", + "issue:44077|issue:44380", + "issue:36683|issue:44077", + "issue:43352|issue:43937", + "issue:43611|issue:44655", + "issue:43867|issue:45216", + "issue:44625|issue:46077", + "issue:43976|issue:44938", + "issue:46077|issue:46082", + "issue:43352|issue:45593", + "issue:36683|issue:43525", + "issue:43618|issue:45216", + "issue:36683|issue:44242", + "issue:43352|issue:43931", + "issue:43064|issue:44492", + "issue:43010|issue:45278", + "issue:43992|issue:45310", + "issue:44355|issue:44855", + "issue:43867|issue:45375", + "issue:41669|issue:44246", + "issue:36683|issue:43064", + "issue:44938|issue:45071", + "issue:42890|issue:45561", + "issue:43010|issue:45850", + "issue:43867|issue:45357", + "issue:42490|issue:44075", + "issue:43010|issue:43976", + "issue:43618|issue:44077", + "issue:43992|issue:45478", + "issue:36296|issue:43575", + "issue:28282|issue:33453", + "issue:43867|issue:44484", + "issue:43618|issue:44857", + "issue:44360|issue:45440" + ], + "prepared_review_unit_hash": "c90995dda5f7afcae37de3aa51afdba66abafbd636ca643fc106dcf81c6114df", + "component_budget": { + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 34, + "serialized_chars": 37386, + "estimated_input_tokens": 9347, + "estimated_eval_tokens": 18950 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is very heterogeneous: it mixes unrelated import errors, model/config regressions, save/load bugs, training issues, and test/performance complaints. None of the soft edges look like true duplicates of the same underlying bug, so I would not merge any of them.", + "confidence": 0.93, + "canonical_issue_reason": "If forced to pick one issue, issue 36010 is the cleanest concrete bug report: a specific import failure with a clear missing symbol traceback. It is only a weak canonical choice because the surrounding items are mostly unrelated.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 36010 is the best representative issue for triage clarity, but not because it matches the rest of the cluster; it is simply the most specific and actionable report among a mixed set.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:36010", + "right": "issue:42757", + "accept": false, + "reason": "Both are import-name errors, but they involve different missing symbols from different packages and likely different fixes." + }, + { + "left": "issue:44077", + "right": "issue:44380", + "accept": false, + "reason": "One is about optimizer arguments not being used; the other is about attention scaling being ignored in SDPA/FlashAttention. Different code paths." + }, + { + "left": "issue:36683", + "right": "issue:44077", + "accept": false, + "reason": "Gemma3 config missing vocab_size is unrelated to SGD optimizer args not being used." + }, + { + "left": "issue:43352", + "right": "issue:43937", + "accept": false, + "reason": "FlashAttention unsupported for NemotronHForCausalLM is unrelated to GLM-5 generation config validation." + }, + { + "left": "issue:43611", + "right": "issue:44655", + "accept": false, + "reason": "base_model_prefix loading regression and Pipeline save_pretrained failure are different save/load behaviors with different failure modes." + }, + { + "left": "issue:43867", + "right": "issue:45216", + "accept": false, + "reason": "State-dict ordering load error is not the same bug as Qwen3.5 checkpoint saving regression." + }, + { + "left": "issue:44625", + "right": "issue:46077", + "accept": false, + "reason": "num_labels propagation in Qwen3.5 is unrelated to Gemma4 attention mask construction." + }, + { + "left": "issue:43976", + "right": "issue:44938", + "accept": false, + "reason": "Both mention Python versions, but one is a 5.1.0 compatibility issue and the other is a Python 3.14 import failure with a different cause." + }, + { + "left": "issue:46077", + "right": "issue:46082", + "accept": false, + "reason": "Gemma4 bidirectional attention masking and LlamaConfig head_dim validation are separate configuration bugs." + }, + { + "left": "issue:43352", + "right": "issue:45593", + "accept": false, + "reason": "Unsupported FlashAttention for Nemotron and D-FINE auxiliary-loss behavior are unrelated model issues." + }, + { + "left": "issue:36683", + "right": "issue:43525", + "accept": false, + "reason": "Missing vocab_size in Gemma3Config and missing pad_token_id in Llama4Config are similar symptom types but different configs and fields." + }, + { + "left": "issue:43618", + "right": "issue:45216", + "accept": false, + "reason": "CLIPOutput attentions assignment and Qwen3.5 save_pretrained regression are distinct subsystems." + }, + { + "left": "issue:36683", + "right": "issue:44242", + "accept": false, + "reason": "A missing Gemma3 config attribute is not the same as load balancing loss omitted when output_router_logits=False." + }, + { + "left": "issue:43352", + "right": "issue:43931", + "accept": false, + "reason": "Nemotron FlashAttention support and Qwen3-VL weight-shape mismatch are different model-loading problems." + }, + { + "left": "issue:43064", + "right": "issue:44492", + "accept": false, + "reason": "FSDP2 optimizer-state corruption is unrelated to a typo in cache strategy docs/code." + }, + { + "left": "issue:43010", + "right": "issue:45278", + "accept": false, + "reason": "Cache update no_grad behavior and broad import errors after upgrade are not the same issue." + }, + { + "left": "issue:43992", + "right": "issue:45310", + "accept": false, + "reason": "UMT5Encoder missing embed_tokens.weight is unrelated to Qwen3.5 MoE from_pretrained errors." + }, + { + "left": "issue:44355", + "right": "issue:44855", + "accept": false, + "reason": "Both are import-related, but one concerns compiled Python files and the other is a Python 3.13 IndentationError from a torch.jit.script decorator/comment parsing issue." + }, + { + "left": "issue:43867", + "right": "issue:45375", + "accept": false, + "reason": "State-dict sorting load error and missing deepstack_visual_indexes are different Qwen3.5 failure modes." + }, + { + "left": "issue:41669", + "right": "issue:44246", + "accept": false, + "reason": "Both touch import time, but one is about removing import * usage and the other is about slow imports; not the same bug." + }, + { + "left": "issue:36683", + "right": "issue:43064", + "accept": false, + "reason": "Gemma3 config missing vocab_size and optimizer-state corruption under FSDP2 are unrelated." + }, + { + "left": "issue:44938", + "right": "issue:45071", + "accept": false, + "reason": "Python 3.14 load failure and PretrainedConfig type-checking breakage are different regression reports." + }, + { + "left": "issue:42890", + "right": "issue:45561", + "accept": false, + "reason": "Missing set_seed in SamHQ tests and xdist race on captured_info.txt are separate test-flakiness issues." + }, + { + "left": "issue:43010", + "right": "issue:45850", + "accept": false, + "reason": "A specific cache/no_grad bug is not the same as a broad 'too many models broken' complaint." + }, + { + "left": "issue:43867", + "right": "issue:45357", + "accept": false, + "reason": "Sorted state-dict loading error and incorrect visual encoder keys in save_pretrained are different save/load regressions." + }, + { + "left": "issue:42490", + "right": "issue:44075", + "accept": false, + "reason": "Local-path slash import warning is unrelated to SGD args not being used." + }, + { + "left": "issue:43010", + "right": "issue:43976", + "accept": false, + "reason": "Cache update decoration and Python-version compatibility are unrelated." + }, + { + "left": "issue:43618", + "right": "issue:44077", + "accept": false, + "reason": "CLIPOutput attentions missing and optimizer argument handling are different code paths." + }, + { + "left": "issue:43992", + "right": "issue:45478", + "accept": false, + "reason": "Missing embed_tokens.weight in UMT5Encoder is unrelated to Qwen3.5 MoE from_pretrained errors." + }, + { + "left": "issue:36296", + "right": "issue:43575", + "accept": false, + "reason": "Tensor-parallel training bug and Qwen2-57B-A14B-Instruct OOM are related only at a very broad level; they are different failures." + }, + { + "left": "issue:28282", + "right": "issue:33453", + "accept": false, + "reason": "AutoModel missing PyTorch and tokenizer-loading regression are unrelated import/loading issues." + }, + { + "left": "issue:43867", + "right": "issue:44484", + "accept": false, + "reason": "Model loading error from sorted state_dict and a question about max_shard_size are not duplicates." + }, + { + "left": "issue:43618", + "right": "issue:44857", + "accept": false, + "reason": "Missing CLIPOutput attentions and an AMP/CUDA crash in LwDetrImageLoss are unrelated bugs." + }, + { + "left": "issue:44360", + "right": "issue:45440", + "accept": false, + "reason": "DSA indexer ReLU discussion and DeepseekV3MoE implementation divergence do not describe the same underlying problem." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 17, + "review_unit_index": 47, + "review_unit_count": 50, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-428", + "nodes": [ + "issue:30333", + "issue:30990", + "issue:31795", + "issue:33453", + "issue:34689", + "issue:36296", + "issue:38617", + "issue:41628", + "issue:42548", + "issue:42757", + "issue:42831", + "issue:42890", + "issue:43010", + "issue:43065", + "issue:43232", + "issue:43299", + "issue:43502", + "issue:43576", + "issue:43618", + "issue:43723", + "issue:43824", + "issue:43827", + "issue:43874", + "issue:44062", + "issue:44112", + "issue:44222", + "issue:44230", + "issue:44315", + "issue:44351", + "issue:44355", + "issue:44360", + "issue:44387", + "issue:44485", + "issue:44556", + "issue:44655", + "issue:44749", + "issue:44961", + "issue:44964", + "issue:44977", + "issue:45071", + "issue:45125", + "issue:45341", + "issue:45375", + "issue:45468", + "issue:45478", + "issue:45600", + "issue:46132", + "issue:46178" + ], + "soft_pairs": [ + "issue:43618|issue:44977", + "issue:42890|issue:44964", + "issue:43502|issue:44062", + "issue:42890|issue:44112", + "issue:42890|issue:45600", + "issue:43576|issue:46132", + "issue:43723|issue:44556", + "issue:44387|issue:44556", + "issue:42890|issue:45341", + "issue:44222|issue:45125", + "issue:43010|issue:43065", + "issue:42548|issue:44351", + "issue:44355|issue:44749", + "issue:43502|issue:44485", + "issue:45125|issue:45478", + "issue:30333|issue:33453", + "issue:43010|issue:43232", + "issue:45125|issue:45375", + "issue:30990|issue:31795", + "issue:43502|issue:44222", + "issue:43299|issue:44556", + "issue:42831|issue:43874", + "issue:43502|issue:44961", + "issue:44315|issue:44360", + "issue:43827|issue:44961", + "issue:30333|issue:36296", + "issue:44355|issue:45071", + "issue:44355|issue:44655", + "issue:33453|issue:34689", + "issue:42757|issue:43576", + "issue:44222|issue:44230", + "issue:43576|issue:45071", + "issue:33453|issue:36296", + "issue:43576|issue:43824", + "issue:31795|issue:38617", + "issue:44961|issue:45468", + "issue:41628|issue:42548", + "issue:38617|issue:43576", + "issue:44355|issue:46178" + ], + "prepared_review_unit_hash": "6996115dfa2504d97f548376f92dc036288768318c2f4004e795050628e5cd43", + "component_budget": { + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 39, + "serialized_chars": 38792, + "estimated_input_tokens": 9698, + "estimated_eval_tokens": 19652 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is a loose mix of Transformers v5 regression reports, import/load failures, and a few unrelated docs/test/model-specific bugs. I do not see any strong duplicate pairs; all soft edges look like separate underlying issues.", + "confidence": 0.42, + "canonical_issue_reason": "Issue 44556 is the broadest, most representative loading-regression report in the set: a v4.57-to-v5.x checkpoint reload breakage. It fits the common compatibility-break theme better than the narrower import, model-specific, test, or docs issues.", + "canonical_pr_reason": null, + "best_issue_reason": "44556 is the best single issue to anchor this cluster because it captures the general v5 upgrade/loading incompatibility pattern without being tied to one niche model or auxiliary symptom.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43618", + "right": "issue:44977", + "accept": false, + "reason": "Different model-level bugs: CLIP attentions assignment vs Qwen3.5 flash-attention generation." + }, + { + "left": "issue:42890", + "right": "issue:44964", + "accept": false, + "reason": "Both involve loading, but one is a SAM test flake and the other is a Phi-4 multimodal model-load failure; different code paths." + }, + { + "left": "issue:43502", + "right": "issue:44062", + "accept": false, + "reason": "Unrelated failures: local_files_only still making requests vs AddedToken keyword duplication." + }, + { + "left": "issue:42890", + "right": "issue:44112", + "accept": false, + "reason": "A seeding-related SAM CI flake is not the same as a GraniteSpeech stale device override failure." + }, + { + "left": "issue:42890", + "right": "issue:45600", + "accept": false, + "reason": "Different areas: SAM integration test flakiness vs broken auto_mappings references." + }, + { + "left": "issue:43576", + "right": "issue:46132", + "accept": false, + "reason": "Different subsystems: v5 env CLI behavior vs AttentionInterface.register changing callbacks." + }, + { + "left": "issue:43723", + "right": "issue:44556", + "accept": false, + "reason": "Both are loading regressions, but one is tokenizer loading and the other is checkpoint reload compatibility; not the same bug." + }, + { + "left": "issue:44387", + "right": "issue:44556", + "accept": false, + "reason": "Int4 memory/OOM regression is a different issue from checkpoint reload failure." + }, + { + "left": "issue:42890", + "right": "issue:45341", + "accept": false, + "reason": "Both touch tests, but this is SAM test seeding flakiness vs a generic testing_utils bug." + }, + { + "left": "issue:44222", + "right": "issue:45125", + "accept": false, + "reason": "Serialization bug in FP8 save_pretrained is not the same as missing tensor-parallel plan metadata." + }, + { + "left": "issue:43010", + "right": "issue:43065", + "accept": false, + "reason": "No common underlying bug: no_grad decoration for cache updates vs dummy Conv2d in Sam3PixelDecoder." + }, + { + "left": "issue:42548", + "right": "issue:44351", + "accept": false, + "reason": "Different missing imports: PreTrainedModel vs HybridCache." + }, + { + "left": "issue:44355", + "right": "issue:44749", + "accept": false, + "reason": "Compiled-file errors and a Chinese report about preprocessing slowdown are unrelated." + }, + { + "left": "issue:43502", + "right": "issue:44485", + "accept": false, + "reason": "Network requests despite local_files_only is unrelated to GLM-5 RoPE implementation." + }, + { + "left": "issue:45125", + "right": "issue:45478", + "accept": false, + "reason": "Tensor-parallel plan metadata and Qwen3.5 Moe loading failure are related only by model family, not the same concrete bug." + }, + { + "left": "issue:30333", + "right": "issue:33453", + "accept": false, + "reason": "MLFlow failure reporting and tokenizer loading regression are unrelated." + }, + { + "left": "issue:43010", + "right": "issue:43232", + "accept": false, + "reason": "Both generation-adjacent, but one is a no_grad decoration request and the other is a sync_gpus generation state bug." + }, + { + "left": "issue:45125", + "right": "issue:45375", + "accept": false, + "reason": "Missing _tp_plan and missing deepstack_visual_indexes are separate model config issues." + }, + { + "left": "issue:30990", + "right": "issue:31795", + "accept": false, + "reason": "A loading hang in Sentence Transformers is not the same as a forward-docs clarification issue." + }, + { + "left": "issue:43502", + "right": "issue:44222", + "accept": false, + "reason": "local_files_only network leakage and FP8 save_pretrained moe are different bugs." + }, + { + "left": "issue:43299", + "right": "issue:44556", + "accept": false, + "reason": "Both are version-related loading regressions, but they affect different model families and failure modes." + }, + { + "left": "issue:42831", + "right": "issue:43874", + "accept": false, + "reason": "FP8 accuracy regression is unrelated to the GLM46V image-patch counting AttributeError." + }, + { + "left": "issue:43502", + "right": "issue:44961", + "accept": false, + "reason": "local_files_only behavior and the racoon issue are unrelated." + }, + { + "left": "issue:44315", + "right": "issue:44360", + "accept": false, + "reason": "Liger Kernel application under model_init is unrelated to the DSA indexer ReLU discussion." + }, + { + "left": "issue:43827", + "right": "issue:44961", + "accept": false, + "reason": "Docs still referencing pipeline() after removals is unrelated to racoon." + }, + { + "left": "issue:30333", + "right": "issue:36296", + "accept": false, + "reason": "MLFlow reporting on failed training jobs is unrelated to tensor-parallel training." + }, + { + "left": "issue:44355", + "right": "issue:45071", + "accept": false, + "reason": "Compiled Python file errors are not the same as PretrainedConfig type-checking breakage." + }, + { + "left": "issue:44355", + "right": "issue:44655", + "accept": false, + "reason": "Compiled-file errors and Pipeline save_pretrained failure are different problems." + }, + { + "left": "issue:33453", + "right": "issue:34689", + "accept": false, + "reason": "Tokenizer loading regression and Llama 3.2 90B Vision model loading breakage are different concrete failures." + }, + { + "left": "issue:42757", + "right": "issue:43576", + "accept": false, + "reason": "Missing is_offline_mode import and broken transformers env command are unrelated." + }, + { + "left": "issue:44222", + "right": "issue:44230", + "accept": false, + "reason": "FP8 save_pretrained moe vs FP8 inference support for Qwen3.5 VL are distinct issues." + }, + { + "left": "issue:43576", + "right": "issue:45071", + "accept": false, + "reason": "Broken env CLI and PretrainedConfig type checking are unrelated." + }, + { + "left": "issue:33453", + "right": "issue:36296", + "accept": false, + "reason": "Tokenizer loading regression and tensor-parallel training bug are different issues." + }, + { + "left": "issue:43576", + "right": "issue:43824", + "accept": false, + "reason": "CLI env command issues are unrelated to the Qwen2.5 VL import error." + }, + { + "left": "issue:31795", + "right": "issue:38617", + "accept": false, + "reason": "Documentation confusion and an ImportError for layer_type_validation are unrelated." + }, + { + "left": "issue:44961", + "right": "issue:45468", + "accept": false, + "reason": "racoon is not related to Gemma-4 audio positional encoding." + }, + { + "left": "issue:41628", + "right": "issue:42548", + "accept": false, + "reason": "AutoImageProcessor import failure and PreTrainedModel import failure are different symbols and fixes." + }, + { + "left": "issue:38617", + "right": "issue:43576", + "accept": false, + "reason": "ImportError for layer_type_validation is unrelated to the transformers env command regression." + }, + { + "left": "issue:44355", + "right": "issue:46178", + "accept": false, + "reason": "Compiled-file errors are unrelated to the tensorflow_text backend mapping ValueError." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 17, + "review_unit_index": 48, + "review_unit_count": 50, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-428", + "nodes": [ + "issue:36296", + "issue:41669", + "issue:42371", + "issue:42890", + "issue:43352", + "issue:43408", + "issue:43525", + "issue:43526", + "issue:43572", + "issue:43576", + "issue:43673", + "issue:43827", + "issue:43901", + "issue:43976", + "issue:44038", + "issue:44077", + "issue:44230", + "issue:44246", + "issue:44273", + "issue:44295", + "issue:44297", + "issue:44315", + "issue:44351", + "issue:44393", + "issue:44485", + "issue:44492", + "issue:44556", + "issue:44623", + "issue:44704", + "issue:44861", + "issue:44908", + "issue:44945", + "issue:44961", + "issue:45003", + "issue:45071", + "issue:45092", + "issue:45125", + "issue:45230", + "issue:45278", + "issue:45341", + "issue:45362", + "issue:45478", + "issue:45542", + "issue:45753", + "issue:45850", + "issue:46095", + "issue:46143", + "issue:46144" + ], + "soft_pairs": [ + "issue:46143|issue:46144", + "issue:44908|issue:44961", + "issue:44485|issue:44961", + "issue:43673|issue:44492", + "issue:36296|issue:43901", + "issue:36296|issue:43526", + "issue:45362|issue:45850", + "issue:44623|issue:45230", + "issue:44273|issue:45542", + "issue:44038|issue:45125", + "issue:44273|issue:45003", + "issue:44393|issue:45071", + "issue:44230|issue:46095", + "issue:44393|issue:45753", + "issue:41669|issue:43673", + "issue:43525|issue:45850", + "issue:43576|issue:44556", + "issue:44556|issue:45092", + "issue:43572|issue:45850", + "issue:44492|issue:45850", + "issue:42371|issue:44492", + "issue:45278|issue:45850", + "issue:44246|issue:44273", + "issue:42890|issue:44908", + "issue:44704|issue:44908", + "issue:43408|issue:44315", + "issue:43976|issue:44393", + "issue:43352|issue:44393", + "issue:44297|issue:45230", + "issue:43576|issue:44908", + "issue:44295|issue:45230", + "issue:44908|issue:45071", + "issue:41669|issue:44351", + "issue:41669|issue:45341", + "issue:44492|issue:44945", + "issue:43976|issue:45850", + "issue:44393|issue:45478", + "issue:44077|issue:44861", + "issue:43827|issue:44908", + "issue:41669|issue:44908" + ], + "prepared_review_unit_hash": "aa1aee864624b8b3559eba34a87729c8a81a1ff9a19a098123ba0a032d89d730", + "component_budget": { + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 40, + "serialized_chars": 38998, + "estimated_input_tokens": 9750, + "estimated_eval_tokens": 19756 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The cluster is noisy and mostly contains unrelated issues; the only clearly related subtheme is import/lazy-loading performance regressions, but the soft edges are too weak to merge. I did not find any PRs in the input.", + "confidence": 0.58, + "canonical_issue_reason": "issue:44246 is the broadest, most representative report of the import/performance subtheme and has the clearest user-facing framing; it reads like the umbrella issue for the cluster.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:44246 is the best single representative issue because it captures the main recurring symptom in this set: slow or problematic imports around lazy loading.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:46143", + "right": "issue:46144", + "accept": false, + "reason": "Same RoFormer area, but one is kwargs propagation and the other is attention-interface usage; related but not the same bug." + }, + { + "left": "issue:44908", + "right": "issue:44961", + "accept": false, + "reason": "Scheduler kwargs vs an unrelated placeholder issue title; no shared underlying defect." + }, + { + "left": "issue:44485", + "right": "issue:44961", + "accept": false, + "reason": "RoPE implementation discussion is unrelated to the placeholder issue." + }, + { + "left": "issue:43673", + "right": "issue:44492", + "accept": false, + "reason": "Chunked prefill cache regression and a cache-strategy typo are different cache problems." + }, + { + "left": "issue:36296", + "right": "issue:43901", + "accept": false, + "reason": "Tensor parallel training and TextClassificationPipeline docs are unrelated." + }, + { + "left": "issue:36296", + "right": "issue:43526", + "accept": false, + "reason": "Tensor parallel training and BeitImageProcessorFast label reduction are unrelated defects." + }, + { + "left": "issue:45362", + "right": "issue:45850", + "accept": false, + "reason": "Both mention model breakage, but one is a specific Qwen3.5 chat crash and the other is a broad complaint." + }, + { + "left": "issue:44623", + "right": "issue:45230", + "accept": false, + "reason": "Processor save_pretrained missing files is not the same as a generic bug report." + }, + { + "left": "issue:44273", + "right": "issue:45542", + "accept": false, + "reason": "Lazy loading issues and a TensorFlow backend import error are different failure modes." + }, + { + "left": "issue:44038", + "right": "issue:45125", + "accept": false, + "reason": "Qwen3-VL-Moe support bug and missing _tp_plan for tensor parallelism are distinct model issues." + }, + { + "left": "issue:44273", + "right": "issue:45003", + "accept": false, + "reason": "Both touch import mechanics, but sys.modules access and lazy loading are not clearly the same bug." + }, + { + "left": "issue:44393", + "right": "issue:45071", + "accept": false, + "reason": "Qwen3-VL 2D bbox errors and PretrainedConfig type checking breakage are unrelated." + }, + { + "left": "issue:44230", + "right": "issue:46095", + "accept": false, + "reason": "FP8 inference support and a weight conversion-script question are not the same change or bug." + }, + { + "left": "issue:44393", + "right": "issue:45753", + "accept": false, + "reason": "Both involve Qwen3-VL models, but one is bbox output and the other is an infinite-loop image case." + }, + { + "left": "issue:41669", + "right": "issue:43673", + "accept": false, + "reason": "Import-star performance regression and chunked-prefill cache regression are different code-path problems." + }, + { + "left": "issue:43525", + "right": "issue:45850", + "accept": false, + "reason": "A missing pad_token_id attribute and a broad v5 breakage complaint are not the same issue." + }, + { + "left": "issue:43576", + "right": "issue:44556", + "accept": false, + "reason": "Broken env command and checkpoint reload incompatibility are different upgrade regressions." + }, + { + "left": "issue:44556", + "right": "issue:45092", + "accept": false, + "reason": "Both are upgrade-related, but one is generic reload failure and the other is InternVL2 meta-init incompatibility." + }, + { + "left": "issue:43572", + "right": "issue:45850", + "accept": false, + "reason": "Missing pad_token_idx in StableLmConfig is unrelated to the broad breakage report." + }, + { + "left": "issue:44492", + "right": "issue:45850", + "accept": false, + "reason": "Cache-strategy typo is a narrow docs/code typo, not the same as a broad compatibility complaint." + }, + { + "left": "issue:42371", + "right": "issue:44492", + "accept": false, + "reason": "TF32 settings guidance and cache-strategy typo are unrelated." + }, + { + "left": "issue:45278", + "right": "issue:45850", + "accept": false, + "reason": "Many import errors after upgrade and a vague 'breaks too many models' complaint are too broad to call the same bug." + }, + { + "left": "issue:44246", + "right": "issue:44273", + "accept": false, + "reason": "Both are import/lazy-loading themed, but one is slow import performance and the other is a broader lazy-loading failure; not clearly the same underlying defect." + }, + { + "left": "issue:42890", + "right": "issue:44908", + "accept": false, + "reason": "Missing set_seed in SAM HQ tests and scheduler kwargs are unrelated." + }, + { + "left": "issue:44704", + "right": "issue:44908", + "accept": false, + "reason": "AutoProcessor kwargs passing and scheduler kwargs are different APIs and bugs." + }, + { + "left": "issue:43408", + "right": "issue:44315", + "accept": false, + "reason": "SAM3 model-type warning and Liger Kernel application via model_init are unrelated." + }, + { + "left": "issue:43976", + "right": "issue:44393", + "accept": false, + "reason": "Python-version compatibility and Qwen3-VL bbox errors are unrelated." + }, + { + "left": "issue:43352", + "right": "issue:44393", + "accept": false, + "reason": "Flash Attention support for NemotronH and Qwen3-VL bbox errors are unrelated." + }, + { + "left": "issue:44297", + "right": "issue:45230", + "accept": false, + "reason": "Tokenizer class mismatch on save_pretrained is not the same as a generic bug report." + }, + { + "left": "issue:43576", + "right": "issue:44908", + "accept": false, + "reason": "Broken env command and scheduler kwargs are unrelated." + }, + { + "left": "issue:44295", + "right": "issue:45230", + "accept": false, + "reason": "position_ids buffer access error is not the same as a generic bug report." + }, + { + "left": "issue:44908", + "right": "issue:45071", + "accept": false, + "reason": "Scheduler kwargs handling and PretrainedConfig type checking are unrelated." + }, + { + "left": "issue:41669", + "right": "issue:44351", + "accept": false, + "reason": "Import-star performance and a HybridCache import error are not the same concrete bug." + }, + { + "left": "issue:41669", + "right": "issue:45341", + "accept": false, + "reason": "Import performance regression and a testing_utils typo are unrelated." + }, + { + "left": "issue:44492", + "right": "issue:44945", + "accept": false, + "reason": "Cache strategy typo and incorrect pipeline-parallel output are different bugs." + }, + { + "left": "issue:43976", + "right": "issue:45850", + "accept": false, + "reason": "Python-version incompatibility and broad model breakage are too different to merge." + }, + { + "left": "issue:44393", + "right": "issue:45478", + "accept": false, + "reason": "Different Qwen3.5 model failures with different symptoms and likely different code paths." + }, + { + "left": "issue:44077", + "right": "issue:44861", + "accept": false, + "reason": "Optional post_init policy and tied-weight-key crash are unrelated." + }, + { + "left": "issue:43827", + "right": "issue:44908", + "accept": false, + "reason": "Docs still mentioning pipeline() and scheduler kwargs are unrelated." + }, + { + "left": "issue:41669", + "right": "issue:44908", + "accept": false, + "reason": "Import-star slowdown and inverse_sqrt scheduler kwargs are unrelated." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 17, + "review_unit_index": 49, + "review_unit_count": 50, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-428", + "nodes": [ + "issue:30333", + "issue:31515", + "issue:31795", + "issue:33453", + "issue:34689", + "issue:36296", + "issue:37428", + "issue:38617", + "issue:41669", + "issue:42371", + "issue:42913", + "issue:42936", + "issue:43066", + "issue:43298", + "issue:43502", + "issue:43519", + "issue:43525", + "issue:43572", + "issue:43576", + "issue:43606", + "issue:43618", + "issue:43704", + "issue:43717", + "issue:44230", + "issue:44261", + "issue:44263", + "issue:44273", + "issue:44315", + "issue:44360", + "issue:44485", + "issue:44492", + "issue:44623", + "issue:44655", + "issue:44829", + "issue:44861", + "issue:44908", + "issue:44998", + "issue:45092", + "issue:45230", + "issue:45290", + "issue:45310", + "issue:45341", + "issue:45468", + "issue:45850", + "issue:46018", + "issue:46095", + "issue:46144", + "issue:46172" + ], + "soft_pairs": [ + "issue:31795|issue:33453", + "issue:44315|issue:44829", + "issue:43576|issue:44273", + "issue:44360|issue:45468", + "issue:44655|issue:44861", + "issue:45230|issue:45310", + "issue:44273|issue:44908", + "issue:46018|issue:46172", + "issue:43519|issue:44485", + "issue:43519|issue:44908", + "issue:41669|issue:43704", + "issue:31795|issue:34689", + "issue:31795|issue:37428", + "issue:44315|issue:45092", + "issue:43704|issue:44485", + "issue:44485|issue:45468", + "issue:43298|issue:46095", + "issue:42371|issue:43704", + "issue:43502|issue:43519", + "issue:43519|issue:44623", + "issue:43519|issue:43704", + "issue:43606|issue:45468", + "issue:44263|issue:45468", + "issue:44261|issue:45468", + "issue:42913|issue:43066", + "issue:44908|issue:44998", + "issue:44492|issue:44998", + "issue:44861|issue:45290", + "issue:43618|issue:44998", + "issue:44998|issue:45850", + "issue:44998|issue:45341", + "issue:43525|issue:44861", + "issue:43572|issue:44861", + "issue:31515|issue:38617", + "issue:44230|issue:44623", + "issue:30333|issue:31515", + "issue:31515|issue:36296", + "issue:31515|issue:31795", + "issue:44230|issue:45310", + "issue:31515|issue:33453", + "issue:43298|issue:44230", + "issue:43717|issue:46144", + "issue:42936|issue:46018" + ], + "prepared_review_unit_hash": "f72ddff805dccf4641ee58a32d7a817595d4739fc79e571f5f1a3438292d9ddd", + "component_budget": { + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 43, + "serialized_chars": 39838, + "estimated_input_tokens": 9960, + "estimated_eval_tokens": 20176 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is dominated by Transformer v5 regression reports, especially model/tokenizer loading and configuration/API breakages. The broad umbrella issue about v5 breaking many models is the best representative; none of the soft pairs look like the same concrete bug/change.", + "confidence": 0.77, + "canonical_issue_reason": "issue:45850 is the broadest and most cluster-representative report: it explicitly frames a v5-wide breakage affecting many models, which matches the surrounding loading/config/import regression issues.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45850 best captures the overall theme of the cluster as a general v5 regression umbrella rather than a narrow model-specific bug.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:31795", + "right": "issue:33453", + "accept": false, + "reason": "Both touch model/tokenizer docs vs tokenizer loading regressions, but one is a documentation confusion issue and the other is a runtime loading regression; not the same underlying bug." + }, + { + "left": "issue:44315", + "right": "issue:44829", + "accept": false, + "reason": "Liger kernel application via model_init is unrelated to flash_attention_3 causing degenerate training; different code paths and failure modes." + }, + { + "left": "issue:43576", + "right": "issue:44273", + "accept": false, + "reason": "Broken `transformers env` CLI and lazy loading malfunction are separate features with no shared concrete bug." + }, + { + "left": "issue:44360", + "right": "issue:45468", + "accept": false, + "reason": "DSA indexer ReLU concern and Gemma-4 audio positional encoding are different model internals; no shared fix target." + }, + { + "left": "issue:44655", + "right": "issue:44861", + "accept": false, + "reason": "Pipeline `save_pretrained` failure and `_get_tied_weight_keys` crashing on list input are separate save/load helper bugs." + }, + { + "left": "issue:45230", + "right": "issue:45310", + "accept": false, + "reason": "Generic bug report and Qwen3.5 MoE `from_pretrained` error are too vague vs specific to be the same issue." + }, + { + "left": "issue:44273", + "right": "issue:44908", + "accept": false, + "reason": "Lazy loading failure and inverse_sqrt scheduler kwargs being ignored are unrelated subsystems." + }, + { + "left": "issue:46018", + "right": "issue:46172", + "accept": false, + "reason": "DeepSeek-V4 gating logic and a tensor transpose bug in `comb.to(dtype)` are different implementation defects." + }, + { + "left": "issue:43519", + "right": "issue:44485", + "accept": false, + "reason": "Qwen3VL timestamp calculation and GLM-5 RoPE implementation are separate model-specific math bugs." + }, + { + "left": "issue:43519", + "right": "issue:44908", + "accept": false, + "reason": "Processor timestamp math and scheduler kwargs handling are unrelated." + }, + { + "left": "issue:41669", + "right": "issue:43704", + "accept": false, + "reason": "Import-star performance issue and Qwen3ForCausalLM VRAM leak are different problems despite both being runtime regressions." + }, + { + "left": "issue:31795", + "right": "issue:34689", + "accept": false, + "reason": "Documentation confusion around inputs/past_key_values is not the same as a Llama 3.2 model loading regression." + }, + { + "left": "issue:31795", + "right": "issue:37428", + "accept": false, + "reason": "Forward-argument docs and a flash-attention import error are unrelated." + }, + { + "left": "issue:44315", + "right": "issue:45092", + "accept": false, + "reason": "Applying Liger Kernel in model_init and meta-initialization incompatibility for old InternVL2 checkpoints are different integration bugs." + }, + { + "left": "issue:43704", + "right": "issue:44485", + "accept": false, + "reason": "VRAM leakage across dataloader threads is unrelated to RoPE implementation correctness." + }, + { + "left": "issue:44485", + "right": "issue:45468", + "accept": false, + "reason": "GLM-5 RoPE math and Gemma-4 audio positional encoding are different model-specific position-encoding issues." + }, + { + "left": "issue:43298", + "right": "issue:46095", + "accept": false, + "reason": "Continuous batching not supporting audio models and a DeepSeek-V4 weight conversion question are not the same bug." + }, + { + "left": "issue:42371", + "right": "issue:43704", + "accept": false, + "reason": "TF32 API guidance and a VRAM leak do not share the same concrete code path." + }, + { + "left": "issue:43502", + "right": "issue:43519", + "accept": false, + "reason": "Accidental API requests despite `local_files_only=True` and incorrect Qwen3VL timestamp computation are unrelated." + }, + { + "left": "issue:43519", + "right": "issue:44623", + "accept": false, + "reason": "Timestamp calculation bug and missing files from `processor.save_pretrained(...)` are separate processor issues." + }, + { + "left": "issue:43519", + "right": "issue:43704", + "accept": false, + "reason": "Qwen3VL timestamp math and Qwen3ForCausalLM thread-related VRAM leak are different bugs." + }, + { + "left": "issue:43606", + "right": "issue:45468", + "accept": false, + "reason": "CPU offload device mismatch for bark-small and Gemma-4 audio positional encoding are both audio-related but not the same failure." + }, + { + "left": "issue:44263", + "right": "issue:45468", + "accept": false, + "reason": "`torch.split` return handling in GlmMoeDsaIndexer is unrelated to Gemma-4 audio positional encoding." + }, + { + "left": "issue:44261", + "right": "issue:45468", + "accept": false, + "reason": "MLA layernorm epsilon precision issue and Gemma-4 audio positional encoding are distinct model math bugs." + }, + { + "left": "issue:42913", + "right": "issue:43066", + "accept": false, + "reason": "Tokenizer behavior change from v4 to v5 and wrong tokenizer decoder type in v5 are both tokenizer-related but not clearly the same concrete defect." + }, + { + "left": "issue:44908", + "right": "issue:44998", + "accept": false, + "reason": "Scheduler kwargs ignored and a non-technical issue ('Unemployment') are obviously unrelated." + }, + { + "left": "issue:44492", + "right": "issue:44998", + "accept": false, + "reason": "Cache strategies typo and a non-technical issue are unrelated." + }, + { + "left": "issue:44861", + "right": "issue:45290", + "accept": false, + "reason": "Tied-weight key handling and chat template crashes with tool calls are different code paths." + }, + { + "left": "issue:43618", + "right": "issue:44998", + "accept": false, + "reason": "CLIPOutput attentions regression and a non-technical issue are unrelated." + }, + { + "left": "issue:44998", + "right": "issue:45850", + "accept": false, + "reason": "A non-technical issue cannot be merged with a broad v5 regression tracker." + }, + { + "left": "issue:44998", + "right": "issue:45341", + "accept": false, + "reason": "A non-technical issue and a small testing_utils bug are unrelated." + }, + { + "left": "issue:43525", + "right": "issue:44861", + "accept": false, + "reason": "Missing `pad_token_id` on `Llama4Config` and tied-weight key crash are different attribute/utility problems." + }, + { + "left": "issue:43572", + "right": "issue:44861", + "accept": false, + "reason": "Missing `pad_token_idx` in `StableLmConfig` is a config-API regression, not the tied-weight list handling crash." + }, + { + "left": "issue:31515", + "right": "issue:38617", + "accept": false, + "reason": "Slow checkpoint loading and missing `layer_type_validation` import are both loading-related in a broad sense, but not the same bug or fix." + }, + { + "left": "issue:44230", + "right": "issue:44623", + "accept": false, + "reason": "FP8 support request for Qwen3-VL/MoE and missing files on processor save are unrelated." + }, + { + "left": "issue:30333", + "right": "issue:31515", + "accept": false, + "reason": "MLFlow reporting on failed training jobs and slow checkpoint loading are completely different concerns." + }, + { + "left": "issue:31515", + "right": "issue:36296", + "accept": false, + "reason": "Slow `from_pretrained` checkpoint loading and tensor parallel training bugs are different subsystems and failure modes." + }, + { + "left": "issue:31515", + "right": "issue:31795", + "accept": false, + "reason": "Checkpoint load slowness and forward doc confusion are not the same issue." + }, + { + "left": "issue:44230", + "right": "issue:45310", + "accept": false, + "reason": "FP8 support request and Qwen3.5 MoE `from_pretrained` error are not the same bug." + }, + { + "left": "issue:31515", + "right": "issue:33453", + "accept": false, + "reason": "Slow checkpoint loading and tokenizer loading regression are both load-time issues, but they are not the same concrete defect." + }, + { + "left": "issue:43298", + "right": "issue:44230", + "accept": false, + "reason": "Continuous batching audio-model support and FP8 inference support are different feature gaps." + }, + { + "left": "issue:43717", + "right": "issue:46144", + "accept": false, + "reason": "Mamba init-weights usage and RoFormer attention interface usage are unrelated architecture implementation issues." + }, + { + "left": "issue:42936", + "right": "issue:46018", + "accept": false, + "reason": "Mask2Former ignore_value misuse and DeepSeek-V4 gating logic are distinct model bugs." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 17, + "review_unit_index": 50, + "review_unit_count": 50, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-428", + "nodes": [ + "issue:37934", + "issue:42936", + "issue:43240", + "issue:43668", + "issue:43813", + "issue:44031", + "issue:46144" + ], + "soft_pairs": [ + "issue:43668|issue:44031", + "issue:42936|issue:43240", + "issue:43240|issue:46144", + "issue:42936|issue:44031", + "issue:42936|issue:46144", + "issue:42936|issue:43813", + "issue:43240|issue:43668", + "issue:43668|issue:43813", + "issue:37934|issue:43668", + "issue:42936|issue:43668" + ], + "prepared_review_unit_hash": "ac5bc3e5a5b27972796756213c96bec1d52767ac4a255a2677055e4dd84054da", + "component_budget": { + "node_count": 428, + "item_count": 428, + "soft_pair_count": 1582, + "serialized_chars": 818962, + "estimated_input_tokens": 204741, + "estimated_eval_tokens": 409738 + }, + "budget": { + "node_count": 7, + "item_count": 7, + "soft_pair_count": 10, + "serialized_chars": 6938, + "estimated_input_tokens": 1735, + "estimated_eval_tokens": 3726 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "These issues are only loosely similar at the wording level; they cover unrelated bugs across norms, loss plumbing, configuration typing, typo fixes, tokenizer warnings, and model attention implementation. I would not merge any of them as duplicates.", + "confidence": 0.97, + "canonical_issue_reason": "No clear canonical issue: the cluster is a set of distinct bug reports affecting different models and code paths, with no shared targets or concrete overlap.", + "canonical_pr_reason": null, + "best_issue_reason": "None of the issues stands out as the obvious duplicate hub; if forced, issue 43240 is the most broadly phrased, but it is still a distinct loss-passing bug rather than a representative superset.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43668", + "right": "issue:44031", + "accept": false, + "reason": "One is a ModernBERT type-hint bug; the other is a tokenizer regex warning regression. Different subsystems and failure modes." + }, + { + "left": "issue:42936", + "right": "issue:43240", + "accept": false, + "reason": "Mask2Former ignore_value handling and loss-kwargs forwarding are unrelated implementation bugs, even though both touch training code." + }, + { + "left": "issue:43240", + "right": "issue:46144", + "accept": false, + "reason": "Loss kwargs propagation and RoFormer attention interface usage affect different model paths and do not describe the same change." + }, + { + "left": "issue:42936", + "right": "issue:44031", + "accept": false, + "reason": "Mask2Former ignore_value and tokenizer regex warnings are unrelated; no shared code path or fix." + }, + { + "left": "issue:42936", + "right": "issue:46144", + "accept": false, + "reason": "A segmentation loss-parameter bug is not the same as a RoFormer attention-interface bug." + }, + { + "left": "issue:42936", + "right": "issue:43813", + "accept": false, + "reason": "Ignore_value usage and a typo in a quantization attribute are different classes of issues with no concrete overlap." + }, + { + "left": "issue:43240", + "right": "issue:43668", + "accept": false, + "reason": "Passing kwargs to loss calculation is unrelated to a ModernBERT config annotation problem." + }, + { + "left": "issue:43668", + "right": "issue:43813", + "accept": false, + "reason": "Both are small fixes, but they target different files and bugs: type hint vs typo." + }, + { + "left": "issue:37934", + "right": "issue:43668", + "accept": false, + "reason": "Llama4TextL2Norm/RMS norm question is unrelated to ModernBERT norm_eps typing." + }, + { + "left": "issue:42936", + "right": "issue:43668", + "accept": false, + "reason": "Different models and different bug types; no same underlying defect." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 2, + "component_count": 17, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10519-3", + "nodes": [ + "issue:10519", + "issue:11005", + "issue:11307" + ], + "soft_pairs": [ + "issue:10519|issue:11307", + "issue:10519|issue:11005", + "issue:11005|issue:11307" + ], + "prepared_review_unit_hash": "de98ab6a0b469a88cefd82b7f8dbe4a49599f62e96fd510d40ae0cf65a31b8ab", + "component_budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 3, + "serialized_chars": 2875, + "estimated_input_tokens": 719, + "estimated_eval_tokens": 1694 + }, + "budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 3, + "serialized_chars": 2875, + "estimated_input_tokens": 719, + "estimated_eval_tokens": 1694 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These three issues are about different feature requests in different areas: token truncation behavior, learning-rate scheduling, and Wav2Vec2 word boundary offsets. I do not see any duplicates or a single coherent canonical issue here.", + "confidence": 0.98, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:10519", + "right": "issue:11307", + "accept": false, + "reason": "One is a tokenizer truncation option request; the other asks for word start/end timestamps in Wav2Vec2. Different API areas and different underlying problems." + }, + { + "left": "issue:10519", + "right": "issue:11005", + "accept": false, + "reason": "Token truncation behavior and ReduceLROnPlateau-like scheduler functionality are unrelated feature requests with no shared code-path or concrete bug." + }, + { + "left": "issue:11005", + "right": "issue:11307", + "accept": false, + "reason": "A learning-rate scheduler feature request is unrelated to extracting time offsets from Wav2Vec2 word boundaries." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 3, + "component_count": 17, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-13244-3", + "nodes": [ + "issue:13244", + "issue:15980", + "issue:16618" + ], + "soft_pairs": [ + "issue:13244|issue:16618", + "issue:15980|issue:16618" + ], + "prepared_review_unit_hash": "5f276f9695cd09466d7695025896d5ce7e5e6173ce00f8622c4a3c02ca35051e", + "component_budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 2, + "serialized_chars": 2479, + "estimated_input_tokens": 620, + "estimated_eval_tokens": 1496 + }, + "budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 2, + "serialized_chars": 2479, + "estimated_input_tokens": 620, + "estimated_eval_tokens": 1496 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "The cluster is not a true duplicate set. Issue 13244 is about Tapas tokenization matching TensorFlow code, while 15980 is about a bad login-related error for private model downloads, and 16618 is a separate model-loading failure for bert-base-uncased. The soft-similarity signals are too broad to treat any pair as the same underlying bug.", + "confidence": 0.91, + "canonical_issue_reason": "No single canonical issue is strongly supported; if one must be chosen, #16618 is the broadest model-loading report, but it does not match the others closely enough to serve as a true duplicate anchor.", + "canonical_pr_reason": null, + "best_issue_reason": "#16618 is the most central and broadly phrased model-loading issue in the set, but it still appears distinct from both the Tapas tokenization bug and the private-model login/error-message issue.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:13244", + "right": "issue:16618", + "accept": false, + "reason": "Tapas tokenization vs. bert-base-uncased loading are different code paths and different failure modes; only a vague NLP/model similarity is shared." + }, + { + "left": "issue:15980", + "right": "issue:16618", + "accept": false, + "reason": "Both concern model download/load failures, but one is specifically about private models without login and the other is a public model load failure; not the same concrete bug." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 4, + "component_count": 17, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-26350-2", + "nodes": [ + "issue:26350", + "issue:33260" + ], + "soft_pairs": [ + "issue:26350|issue:33260" + ], + "prepared_review_unit_hash": "ab1a4f331af7c6d6534f6bdcd7a616d0221a85181649e4f63ca332c1dcb19f4c", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1589, + "estimated_input_tokens": 398, + "estimated_eval_tokens": 1052 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1589, + "estimated_input_tokens": 398, + "estimated_eval_tokens": 1052 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are two separate community contribution tracking issues for different features: Flash Attention 2 support vs GGUF support. They share a broad theme of expanding architecture support, but they do not describe the same underlying change or bug.", + "confidence": 0.98, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": "Issue 26350 is the stronger representative issue overall because it has substantially higher discussion activity and appears more central within this cluster, but it is not a duplicate of 33260.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:26350", + "right": "issue:33260", + "accept": false, + "reason": "Reject: the issues target different features and code paths. Flash Attention 2 support and GGUF support are unrelated contributions beyond both being architecture-expansion requests." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 5, + "component_count": 17, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-27088-6", + "nodes": [ + "issue:27088", + "issue:28936", + "issue:38435", + "issue:43834", + "issue:44365", + "issue:44370" + ], + "soft_pairs": [ + "issue:27088|issue:44370", + "issue:27088|issue:44365", + "issue:27088|issue:43834", + "issue:27088|issue:38435", + "issue:27088|issue:28936", + "issue:28936|issue:43834", + "issue:28936|issue:44365", + "issue:28936|issue:38435", + "issue:28936|issue:44370", + "issue:38435|issue:44370", + "issue:38435|issue:44365", + "issue:38435|issue:43834", + "issue:43834|issue:44370", + "issue:43834|issue:44365", + "issue:44365|issue:44370" + ], + "prepared_review_unit_hash": "bc60e30a09a96aef241cce34c5e31df766a9e0cf2372eae5fc8422a29172b3e7", + "component_budget": { + "node_count": 6, + "item_count": 6, + "soft_pair_count": 15, + "serialized_chars": 8794, + "estimated_input_tokens": 2199, + "estimated_eval_tokens": 4654 + }, + "budget": { + "node_count": 6, + "item_count": 6, + "soft_pair_count": 15, + "serialized_chars": 8794, + "estimated_input_tokens": 2199, + "estimated_eval_tokens": 4654 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are all documentation translation requests, but the explicitly named issues are for different languages, and the generic placeholder-title issues do not provide enough evidence to treat them as the same underlying request. No PRs are present.", + "confidence": 0.86, + "canonical_issue_reason": "No single canonical duplicate stands out: the named issues are language-specific requests, and the placeholder-title issues are too underspecified to safely collapse into one canonical issue.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 43834 is the closest representative only because it is the broad template-style translation issue, but it still should not subsume the language-specific requests.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:27088", + "right": "issue:44370", + "accept": false, + "reason": "Turkish docs translation vs generic placeholder translation; no evidence they are the same request." + }, + { + "left": "issue:27088", + "right": "issue:44365", + "accept": false, + "reason": "Turkish docs translation vs generic placeholder translation; different underlying request is more likely." + }, + { + "left": "issue:27088", + "right": "issue:43834", + "accept": false, + "reason": "Named Turkish translation request vs underspecified template issue; not enough to merge." + }, + { + "left": "issue:27088", + "right": "issue:38435", + "accept": false, + "reason": "Different target languages: Turkish vs Romanian." + }, + { + "left": "issue:27088", + "right": "issue:28936", + "accept": false, + "reason": "Different target languages: Turkish vs Spanish." + }, + { + "left": "issue:28936", + "right": "issue:43834", + "accept": false, + "reason": "Spanish translation request vs generic placeholder issue; no concrete same-bug evidence." + }, + { + "left": "issue:28936", + "right": "issue:44365", + "accept": false, + "reason": "Spanish translation request vs generic placeholder issue; insufficient evidence of a single underlying issue." + }, + { + "left": "issue:28936", + "right": "issue:38435", + "accept": false, + "reason": "Different target languages: Spanish vs Romanian." + }, + { + "left": "issue:28936", + "right": "issue:44370", + "accept": false, + "reason": "Spanish translation request vs generic placeholder issue; do not merge on broad similarity alone." + }, + { + "left": "issue:38435", + "right": "issue:44370", + "accept": false, + "reason": "Romanian translation request vs generic placeholder issue; not enough evidence they are the same request." + }, + { + "left": "issue:38435", + "right": "issue:44365", + "accept": false, + "reason": "Romanian translation request vs generic placeholder issue; no explicit link or shared target." + }, + { + "left": "issue:38435", + "right": "issue:43834", + "accept": false, + "reason": "Romanian translation request vs underspecified template issue; too little evidence to merge." + }, + { + "left": "issue:43834", + "right": "issue:44370", + "accept": false, + "reason": "Both are template-style titles, but the bodies/targets are not shown; strict triage should not assume they are duplicates." + }, + { + "left": "issue:43834", + "right": "issue:44365", + "accept": false, + "reason": "Identical placeholder title is not enough to prove the same underlying translation request." + }, + { + "left": "issue:44365", + "right": "issue:44370", + "accept": false, + "reason": "Both are generic placeholder translation issues, but there is no concrete evidence they target the same language or work item." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 6, + "component_count": 17, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-36817-16", + "nodes": [ + "issue:36817", + "issue:36879", + "issue:37015", + "issue:41115", + "issue:41211", + "issue:42503", + "issue:42738", + "issue:43671", + "issue:43837", + "issue:43909", + "issue:44121", + "issue:44205", + "issue:45027", + "issue:45177", + "issue:45322", + "issue:46133" + ], + "soft_pairs": [ + "issue:36817|issue:41211", + "issue:36817|issue:43671", + "issue:36817|issue:36879", + "issue:36879|issue:41211", + "issue:36879|issue:43909", + "issue:36879|issue:44121", + "issue:36879|issue:43837", + "issue:36879|issue:43671", + "issue:37015|issue:43837", + "issue:37015|issue:44121", + "issue:37015|issue:43909", + "issue:36879|issue:37015", + "issue:37015|issue:42503", + "issue:41115|issue:43837", + "issue:41115|issue:44121", + "issue:41115|issue:42503", + "issue:41115|issue:45027", + "issue:41211|issue:43671", + "issue:42503|issue:43837", + "issue:42503|issue:43671", + "issue:42503|issue:42738", + "issue:42738|issue:45027", + "issue:43671|issue:43837", + "issue:43671|issue:45027", + "issue:43671|issue:45177", + "issue:43671|issue:44121", + "issue:43671|issue:43909", + "issue:43837|issue:43909", + "issue:43837|issue:45322", + "issue:43837|issue:44121", + "issue:43909|issue:44205", + "issue:44121|issue:45322", + "issue:43837|issue:45027", + "issue:37015|issue:45027", + "issue:43837|issue:45177", + "issue:45322|issue:46133" + ], + "prepared_review_unit_hash": "4e38cc7d50b7e00681e72aebc9f2c16725af0212e2f2882789aa0f22d9e6203c", + "component_budget": { + "node_count": 16, + "item_count": 16, + "soft_pair_count": 36, + "serialized_chars": 21657, + "estimated_input_tokens": 5415, + "estimated_eval_tokens": 11086 + }, + "budget": { + "node_count": 16, + "item_count": 16, + "soft_pair_count": 36, + "serialized_chars": 21657, + "estimated_input_tokens": 5415, + "estimated_eval_tokens": 11086 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "This cluster is a loose topical group of unrelated model-addition requests. They all resemble \u201cplease add support for model X,\u201d but they are not duplicates of the same underlying issue. No PRs are present.", + "confidence": 0.95, + "canonical_issue_reason": "issue:43671 is the broadest and most central representative of the cluster\u2019s generic model-support requests, but it is not a true duplicate of the others.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43671 is the best single representative because it is an open, discussion-backed model support request and sits near the center of the similarity graph.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:36817", + "right": "issue:41211", + "accept": false, + "reason": "Different requested models (EuroBert vs DEIMv2); only the generic add-support framing overlaps." + }, + { + "left": "issue:36817", + "right": "issue:43671", + "accept": false, + "reason": "EuroBert config support and Qwen3-TTS support are unrelated model additions." + }, + { + "left": "issue:36817", + "right": "issue:36879", + "accept": false, + "reason": "EuroBert vs RF-DETR are different architectures and different feature requests." + }, + { + "left": "issue:36879", + "right": "issue:41211", + "accept": false, + "reason": "RF-DETR and DEIMv2 are distinct model integrations, not the same bug/change." + }, + { + "left": "issue:36879", + "right": "issue:43909", + "accept": false, + "reason": "RF-DETR detection support and LFM2.5 Audio support are different domains." + }, + { + "left": "issue:36879", + "right": "issue:44121", + "accept": false, + "reason": "RF-DETR and OpenAI weight-sparse transformer are unrelated model requests." + }, + { + "left": "issue:36879", + "right": "issue:43837", + "accept": false, + "reason": "RF-DETR and Qwen3-ASR are different models with different code paths." + }, + { + "left": "issue:36879", + "right": "issue:43671", + "accept": false, + "reason": "RF-DETR and Qwen3-TTS are unrelated additions despite both being model support requests." + }, + { + "left": "issue:37015", + "right": "issue:43837", + "accept": false, + "reason": "NeoBERT and Qwen3-ASR are unrelated model additions." + }, + { + "left": "issue:37015", + "right": "issue:44121", + "accept": false, + "reason": "NeoBERT support and weight-sparse transformer support do not describe the same change." + }, + { + "left": "issue:37015", + "right": "issue:43909", + "accept": false, + "reason": "NeoBERT and LFM2.5 Audio are distinct model requests." + }, + { + "left": "issue:36879", + "right": "issue:37015", + "accept": false, + "reason": "RF-DETR and NeoBERT are different model integrations." + }, + { + "left": "issue:37015", + "right": "issue:42503", + "accept": false, + "reason": "NeoBERT and ModernVBERT are different BERT-family requests, not the same model." + }, + { + "left": "issue:41115", + "right": "issue:43837", + "accept": false, + "reason": "MiniCPM3 architecture support and Qwen3-ASR support are unrelated." + }, + { + "left": "issue:41115", + "right": "issue:44121", + "accept": false, + "reason": "MiniCPM3 and the weight-sparse transformer request target different code paths." + }, + { + "left": "issue:41115", + "right": "issue:42503", + "accept": false, + "reason": "MiniCPM3 and ModernVBERT are separate model additions." + }, + { + "left": "issue:41115", + "right": "issue:45027", + "accept": false, + "reason": "MiniCPM3 and Voxtral TTS are different model families and tasks." + }, + { + "left": "issue:41211", + "right": "issue:43671", + "accept": false, + "reason": "DEIMv2 and Qwen3-TTS are unrelated model requests." + }, + { + "left": "issue:42503", + "right": "issue:43837", + "accept": false, + "reason": "ModernVBERT and Qwen3-ASR are different model additions." + }, + { + "left": "issue:42503", + "right": "issue:43671", + "accept": false, + "reason": "ModernVBERT and Qwen3-TTS are different models with different support needs." + }, + { + "left": "issue:42503", + "right": "issue:42738", + "accept": false, + "reason": "ModernVBERT and BERT-like models with RoPE are related only broadly, not the same concrete change." + }, + { + "left": "issue:42738", + "right": "issue:45027", + "accept": false, + "reason": "RoPE support for BERT-like models and Voxtral TTS support are unrelated changes." + }, + { + "left": "issue:43671", + "right": "issue:43837", + "accept": false, + "reason": "Qwen3-TTS and Qwen3-ASR are related product lines but different modalities and implementations." + }, + { + "left": "issue:43671", + "right": "issue:45027", + "accept": false, + "reason": "Qwen3-TTS and Voxtral TTS are distinct TTS model support requests." + }, + { + "left": "issue:43671", + "right": "issue:45177", + "accept": false, + "reason": "Qwen3-TTS and DeepSeek-OCR2 are different model additions." + }, + { + "left": "issue:43671", + "right": "issue:44121", + "accept": false, + "reason": "Qwen3-TTS support and weight-sparse transformer support are not the same underlying change." + }, + { + "left": "issue:43671", + "right": "issue:43909", + "accept": false, + "reason": "Qwen3-TTS and LFM2.5 Audio are different audio-model requests." + }, + { + "left": "issue:43837", + "right": "issue:43909", + "accept": false, + "reason": "Qwen3-ASR and LFM2.5 Audio are separate audio model integrations." + }, + { + "left": "issue:43837", + "right": "issue:45322", + "accept": false, + "reason": "Qwen3-ASR and EUPE are unrelated model requests." + }, + { + "left": "issue:43837", + "right": "issue:44121", + "accept": false, + "reason": "Qwen3-ASR support and weight-sparse transformer support are different changes." + }, + { + "left": "issue:43909", + "right": "issue:44205", + "accept": false, + "reason": "LFM2.5 Audio and SAM3-LiteText are different model families and tasks." + }, + { + "left": "issue:44121", + "right": "issue:45322", + "accept": false, + "reason": "Weight-sparse transformer support and EUPE support are unrelated model requests." + }, + { + "left": "issue:43837", + "right": "issue:45027", + "accept": false, + "reason": "Qwen3-ASR and Voxtral TTS are different speech-model additions." + }, + { + "left": "issue:37015", + "right": "issue:45027", + "accept": false, + "reason": "NeoBERT and Voxtral TTS are unrelated model requests." + }, + { + "left": "issue:43837", + "right": "issue:45177", + "accept": false, + "reason": "Qwen3-ASR and DeepSeek-OCR2 are different model additions." + }, + { + "left": "issue:45322", + "right": "issue:46133", + "accept": false, + "reason": "EUPE and TIPSv2 are distinct vision-language model requests." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 7, + "component_count": 17, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-41084-2", + "nodes": [ + "issue:41084", + "issue:43037" + ], + "soft_pairs": [ + "issue:41084|issue:43037" + ], + "prepared_review_unit_hash": "37fca706d0e8b45ecbafe199df80c911dc3460adad84ad37fb0efcb1f4c24703", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1480, + "estimated_input_tokens": 370, + "estimated_eval_tokens": 996 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1480, + "estimated_input_tokens": 370, + "estimated_eval_tokens": 996 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "The two issues are unrelated: one is a block decoding setting change, the other is a request for DeepSeek v3.2 support. They should not be treated as duplicates or one cluster representative.", + "confidence": 0.98, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:41084", + "right": "issue:43037", + "accept": false, + "reason": "Different underlying requests: 'Set Block Decoding' is a decoding/configuration feature, while 'DeepSeek v3.2 support' is model support. No shared code-path or concrete bug overlap is evident." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 8, + "component_count": 17, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-43598-2", + "nodes": [ + "issue:43598", + "issue:43600" + ], + "soft_pairs": [ + "issue:43598|issue:43600" + ], + "prepared_review_unit_hash": "7a2f54d8b939df308495286e3bd634788e7888f6f2d004c7ba890ac0cf813ddd", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1583, + "estimated_input_tokens": 396, + "estimated_eval_tokens": 1048 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1583, + "estimated_input_tokens": 396, + "estimated_eval_tokens": 1048 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "The two issues are loosely related through Trainer/custom-trainer behavior, but they describe different problems: #43598 is about `remove_unused_column` customization, while #43600 is about FLOPS estimation for custom trainers. They should not be treated as duplicates.", + "confidence": 0.96, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": "No issue is a clear canonical duplicate representative because the two reports target different Trainer code paths; if forced, #43598 is the more self-contained feature request, but it does not subsume #43600.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43598", + "right": "issue:43600", + "accept": false, + "reason": "Different underlying concerns: one changes column-removal behavior for customizability, the other changes FLOPS estimation for custom trainers. Shared Trainer context is too broad to merge them." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 9, + "component_count": 17, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-43684-2", + "nodes": [ + "issue:43684", + "issue:44405" + ], + "soft_pairs": [ + "issue:43684|issue:44405" + ], + "prepared_review_unit_hash": "fdfd7cfed96e347f416dae52fae993f9a95da4f7b861acd420c98dca7a9a6a59", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1612, + "estimated_input_tokens": 403, + "estimated_eval_tokens": 1062 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1612, + "estimated_input_tokens": 403, + "estimated_eval_tokens": 1062 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are separate feature requests for different model families and different auto-model heads, so they should not be clustered as duplicates.", + "confidence": 0.95, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": "No single best canonical issue: #43684 asks for Qwen3-Omni registration in AutoModel/AutoModelForConditionalGeneration, while #44405 asks for AutoModelForSequenceClassification support for Qwen3.5 (Qwen3_5Config). They target different code paths and tasks.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43684", + "right": "issue:44405", + "accept": false, + "reason": "Different model variants and different auto-class heads; this is not the same underlying bug or change." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 10, + "component_count": 17, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-43835-4", + "nodes": [ + "issue:43835", + "issue:43845", + "issue:43846", + "issue:43859" + ], + "soft_pairs": [ + "issue:43835|issue:43845", + "issue:43835|issue:43846", + "issue:43835|issue:43859", + "issue:43845|issue:43846", + "issue:43845|issue:43859", + "issue:43846|issue:43859" + ], + "prepared_review_unit_hash": "0019635b2842297879147a3e434944583c548a51a1067d55309375c16fdce949", + "component_budget": { + "node_count": 4, + "item_count": 4, + "soft_pair_count": 6, + "serialized_chars": 4261, + "estimated_input_tokens": 1066, + "estimated_eval_tokens": 2388 + }, + "budget": { + "node_count": 4, + "item_count": 4, + "soft_pair_count": 6, + "serialized_chars": 4261, + "estimated_input_tokens": 1066, + "estimated_eval_tokens": 2388 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "The cluster is too underspecified to confidently treat these as duplicates. The three \"huggingface\" issues may be related to each other by title, but there is no body or target evidence showing the same underlying bug, and \"Fett\" looks unrelated.", + "confidence": 0.31, + "canonical_issue_reason": "Issue 43845 is the earliest of the repeated \"huggingface\" titles and is the most plausible anchor if this cluster were later confirmed to be duplicates.", + "canonical_pr_reason": null, + "best_issue_reason": "43845 is the best issue candidate only because it is the earliest repeated title; however, the evidence is too weak to call it a true canonical duplicate target.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43835", + "right": "issue:43845", + "accept": false, + "reason": "Different titles and no body/target evidence tying \"Fett\" to the \"huggingface\" reports." + }, + { + "left": "issue:43835", + "right": "issue:43846", + "accept": false, + "reason": "Different titles and no evidence of the same underlying problem." + }, + { + "left": "issue:43835", + "right": "issue:43859", + "accept": false, + "reason": "Different titles and no evidence these describe the same bug." + }, + { + "left": "issue:43845", + "right": "issue:43846", + "accept": false, + "reason": "Same short title, but there is no issue-body or target detail confirming the same underlying report." + }, + { + "left": "issue:43845", + "right": "issue:43859", + "accept": false, + "reason": "Matching title alone is not enough here; the issues are too underspecified to merge as duplicates." + }, + { + "left": "issue:43846", + "right": "issue:43859", + "accept": false, + "reason": "Matching title alone is insufficient to confirm the same bug." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 11, + "component_count": 17, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-44371-5", + "nodes": [ + "issue:44371", + "issue:44418", + "issue:44572", + "issue:44573", + "issue:44937" + ], + "soft_pairs": [ + "issue:44371|issue:44572", + "issue:44371|issue:44573", + "issue:44371|issue:44418", + "issue:44371|issue:44937", + "issue:44418|issue:44572", + "issue:44418|issue:44573", + "issue:44572|issue:44573", + "issue:44572|issue:44937", + "issue:44573|issue:44937" + ], + "prepared_review_unit_hash": "984722bb2ca964ae12be9c7c15283bde548d490db181ca44dadf228f9221149c", + "component_budget": { + "node_count": 5, + "item_count": 5, + "soft_pair_count": 9, + "serialized_chars": 5788, + "estimated_input_tokens": 1447, + "estimated_eval_tokens": 3150 + }, + "budget": { + "node_count": 5, + "item_count": 5, + "soft_pair_count": 9, + "serialized_chars": 5788, + "estimated_input_tokens": 1447, + "estimated_eval_tokens": 3150 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "All items are low-signal and look like unrelated spam/placeholder issues rather than duplicates of one concrete bug or change. No soft pair should be merged.", + "confidence": 0.96, + "canonical_issue_reason": "issue:44418 is the only item with a somewhat descriptive title, but it still lacks concrete problem details and does not match the others as the same underlying report.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:44418 is the most informative representative by title alone, but it is still not a strong canonical candidate for a duplicate cluster.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44371", + "right": "issue:44572", + "accept": false, + "reason": "Both are placeholder '' issues with no substantive bug description; no evidence they report the same problem." + }, + { + "left": "issue:44371", + "right": "issue:44573", + "accept": false, + "reason": "Same as above: generic spam-like placeholders, not a concrete shared issue." + }, + { + "left": "issue:44371", + "right": "issue:44418", + "accept": false, + "reason": "A spam placeholder and a vague documentation suggestion are not enough to establish the same underlying issue." + }, + { + "left": "issue:44371", + "right": "issue:44937", + "accept": false, + "reason": "The titles are unrelated and there is no shared code-path or explicit linkage." + }, + { + "left": "issue:44418", + "right": "issue:44572", + "accept": false, + "reason": "Different vague topics with no concrete symptoms, targets, or evidence of duplication." + }, + { + "left": "issue:44418", + "right": "issue:44573", + "accept": false, + "reason": "The documentation suggestion does not look like the same report as the spam placeholder issue." + }, + { + "left": "issue:44572", + "right": "issue:44573", + "accept": false, + "reason": "Both are spam-like placeholders, but there is no basis to conclude they describe the same underlying bug or change." + }, + { + "left": "issue:44572", + "right": "issue:44937", + "accept": false, + "reason": "Unrelated titles and no shared issue details indicate different topics." + }, + { + "left": "issue:44573", + "right": "issue:44937", + "accept": false, + "reason": "No concrete overlap beyond being low-information closed issues." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 12, + "component_count": 17, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-45145-2", + "nodes": [ + "issue:45145", + "issue:45151" + ], + "soft_pairs": [ + "issue:45145|issue:45151" + ], + "prepared_review_unit_hash": "95571c546c479d44e639775f5e05db2caaffbfbaa98a0f38754315bf85ed912f", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1646, + "estimated_input_tokens": 412, + "estimated_eval_tokens": 1080 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1646, + "estimated_input_tokens": 412, + "estimated_eval_tokens": 1080 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "The two closed issues are highly similar and likely duplicate the same report: identical titles, very similar body lengths, close timestamps, and a soft-similarity signal. #45145 is the better canonical issue because it was created first and already has an inbound reference.", + "confidence": 0.87, + "canonical_issue_reason": "#45145 is the best canonical anchor: it appeared first and already has an inbound reference, while the paired issue is highly similar.", + "canonical_pr_reason": null, + "best_issue_reason": "#45145 is the strongest representative of the cluster due to its earlier creation time and existing inbound reference, with no evidence that it is a different report.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45145", + "right": "issue:45151", + "accept": true, + "reason": "The pair looks like the same underlying issue: identical titles, very close body sizes/timestamps, and a similarity signal. The evidence is suggestive rather than definitive, but sufficient to treat them as duplicates." + } + ] + }, + "evaluator_result": { + "accept": false, + "feedback": "The pair looks plausibly duplicate, but the analyst overstates the evidence by calling it a definite \"true duplicate pair.\" The packet only supports identical titles, similar body length, close timestamps, and a soft-similarity signal; there are no explicit links or stronger deterministic indicators. The summary is mostly grounded, but the soft-edge verdict should be more conservative." + }, + "evaluator_used": true, + "retried": true, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 13, + "component_count": 17, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "cache", + "cache_hit": true, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-45295-2", + "nodes": [ + "issue:45295", + "issue:45373" + ], + "soft_pairs": [], + "prepared_review_unit_hash": "f5fd95fada2f530ab0999e343c17c16266cde0c7182bb22509972ca44827cd49", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 0, + "serialized_chars": 1286, + "estimated_input_tokens": 322, + "estimated_eval_tokens": 900 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 0, + "serialized_chars": 1286, + "estimated_input_tokens": 322, + "estimated_eval_tokens": 900 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These two issues describe the same missing Gemma 4 sequence-classification support; the later, more specific open issue is the better cluster representative.", + "confidence": 0.94, + "canonical_issue_reason": "issue #45373 is the clearest canonical tracker: it is open, explicitly names the missing class `Gemma4ForSequenceClassification`, and states the module-level gap relative to Gemma 2/3.", + "canonical_pr_reason": null, + "best_issue_reason": "issue #45373 best captures the concrete request and likely remaining work, while #45295 is a shorter earlier duplicate with less detail and is already closed.", + "best_pr_reason": null, + "soft_edge_verdicts": [] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 14, + "component_count": 17, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-45672-2", + "nodes": [ + "issue:45672", + "issue:45674" + ], + "soft_pairs": [ + "issue:45672|issue:45674" + ], + "prepared_review_unit_hash": "4d665438b390f53ff0576751f55adfa4c00a49f651492b172f1ded1b3a6aa3ad", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1668, + "estimated_input_tokens": 417, + "estimated_eval_tokens": 1090 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1668, + "estimated_input_tokens": 417, + "estimated_eval_tokens": 1090 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Two open issues report different 4-bit inference failures in different parts of the stack: one in Gemma4 audio module dtype handling, the other in BitsAndBytesConfig module-exclusion logic. They are related by quantized inference, but not the same underlying bug.", + "confidence": 0.92, + "canonical_issue_reason": "Issue #45674 is the better representative because it describes a clear, config-driven regression with a concrete failure mode and reproduction path.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue #45674 is the strongest single issue in the cluster: it is specific, actionable, and centers on a concrete 4-bit inference regression.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45672", + "right": "issue:45674", + "accept": false, + "reason": "Different failures and code paths: #45672 is a Gemma4 audio-module uint8/torch.finfo TypeError, while #45674 is a BitsAndBytesConfig lm_int8_skip_modules regression that clears the default lm_head exclusion and triggers an AssertionError. They should not be merged as duplicates." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 15, + "component_count": 17, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-45762-2", + "nodes": [ + "issue:45762", + "issue:45763" + ], + "soft_pairs": [ + "issue:45762|issue:45763" + ], + "prepared_review_unit_hash": "8e9a399efd21ef390332339b6077c3df032430369ead22e53e32172cc706f7f3", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1477, + "estimated_input_tokens": 370, + "estimated_eval_tokens": 996 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1477, + "estimated_input_tokens": 370, + "estimated_eval_tokens": 996 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "The two closed issues look unrelated: both are bare personal-name/date entries with no body, no links, and no shared bug or feature context. This is not a duplicate cluster.", + "confidence": 0.96, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45762", + "right": "issue:45763", + "accept": false, + "reason": "Titles are different personal-name/date strings and there is no issue body or linked context showing the same underlying problem." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 16, + "component_count": 17, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-45764-2", + "nodes": [ + "issue:45764", + "issue:45765" + ], + "soft_pairs": [ + "issue:45764|issue:45765" + ], + "prepared_review_unit_hash": "d486228ace9611605118b53c09e6d7daa7b5bbfffc48772f0b28c8fd96ac8230", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1481, + "estimated_input_tokens": 371, + "estimated_eval_tokens": 998 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1481, + "estimated_input_tokens": 371, + "estimated_eval_tokens": 998 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "The two closed issues are unrelated personal-name/date entries with no description, references, or shared technical context. There is no evidence they describe the same bug, request, or change.", + "confidence": 0.99, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45764", + "right": "issue:45765", + "accept": false, + "reason": "The titles are different names and dates, and both issues lack body text or linking evidence; they do not appear to be the same underlying report." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 17, + "component_count": 17, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-45869-5", + "nodes": [ + "issue:45869", + "issue:45870", + "issue:45871", + "issue:45872", + "issue:45873" + ], + "soft_pairs": [ + "issue:45869|issue:45870", + "issue:45869|issue:45871", + "issue:45870|issue:45871", + "issue:45870|issue:45872", + "issue:45870|issue:45873", + "issue:45871|issue:45872", + "issue:45871|issue:45873", + "issue:45872|issue:45873" + ], + "prepared_review_unit_hash": "fe5b78ddf4fbb5ce8d6ec645192f544204d79042420d1536212cef440886aee4", + "component_budget": { + "node_count": 5, + "item_count": 5, + "soft_pair_count": 8, + "serialized_chars": 5726, + "estimated_input_tokens": 1432, + "estimated_eval_tokens": 3120 + }, + "budget": { + "node_count": 5, + "item_count": 5, + "soft_pair_count": 8, + "serialized_chars": 5726, + "estimated_input_tokens": 1432, + "estimated_eval_tokens": 3120 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Two high-confidence duplicate candidates appear: 45870/45871 and 45872/45873. 45869 looks like a separate dataset-promo issue. The evidence supports duplicate pairing mainly from identical titles and identical body lengths, but the packet does not expose body text, so the certainty should stay slightly qualified.", + "confidence": 0.89, + "canonical_issue_reason": "Issue 45870 is a reasonable canonical issue for the first duplicate pair because it is the earlier of the two identical-title reports and has slightly more discussion activity.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 45870 is the best overall issue representative for this cluster because it is the earliest clear duplicate candidate and has the most activity among the example-prediction pair.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45869", + "right": "issue:45870", + "accept": false, + "reason": "Different titles and different apparent topic: dataset promo versus example prediction thread." + }, + { + "left": "issue:45869", + "right": "issue:45871", + "accept": false, + "reason": "Different titles and different apparent topic: dataset promo versus example prediction thread." + }, + { + "left": "issue:45870", + "right": "issue:45871", + "accept": true, + "reason": "Same title, same body length, and close timestamps; strong duplicate signal." + }, + { + "left": "issue:45870", + "right": "issue:45872", + "accept": false, + "reason": "Related financial/DataForge wording, but the titles describe different requests." + }, + { + "left": "issue:45870", + "right": "issue:45873", + "accept": false, + "reason": "Related financial/DataForge wording, but the titles describe different requests." + }, + { + "left": "issue:45871", + "right": "issue:45872", + "accept": false, + "reason": "Different titles and different request shape: example prediction versus scraping-service topic." + }, + { + "left": "issue:45871", + "right": "issue:45873", + "accept": false, + "reason": "Different titles and different request shape: example prediction versus scraping-service topic." + }, + { + "left": "issue:45872", + "right": "issue:45873", + "accept": true, + "reason": "Same title, same body length, and close timestamps; strong duplicate signal." + } + ] + }, + "evaluator_result": { + "accept": false, + "feedback": "The grouping is directionally plausible, but several claims overstate the evidence available in the packet. The packet only provides titles, body lengths, and timestamps; it does not expose body text, so statements like \u201cexact duplicate,\u201d \u201cmatching body length/content pattern,\u201d and \u201cexact same title\u201d are only partly grounded. In particular, calling 45870/45871 and 45872/45873 definite duplicates is stronger than the evidence supports; \u2018likely duplicate\u2019 or \u2018high-confidence duplicate candidate\u2019 would be safer. The summary also says \u2018Two clear duplicate pairs appear,\u2019 which is more certain than the packet justifies. The 45869 separation is conservative and grounded." + }, + "evaluator_used": true, + "retried": true, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 1, + "component_count": 27, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-39895-13", + "nodes": [ + "pull_request:39895", + "pull_request:41855", + "pull_request:42668", + "pull_request:43251", + "pull_request:43254", + "pull_request:43424", + "pull_request:43823", + "pull_request:43995", + "pull_request:44827", + "pull_request:45170", + "pull_request:45213", + "pull_request:45655", + "pull_request:46048" + ], + "soft_pairs": [ + "pull_request:39895|pull_request:43424", + "pull_request:41855|pull_request:43823", + "pull_request:41855|pull_request:43424", + "pull_request:42668|pull_request:43823", + "pull_request:43251|pull_request:43823", + "pull_request:43251|pull_request:44827", + "pull_request:43424|pull_request:44827", + "pull_request:43424|pull_request:43823", + "pull_request:43823|pull_request:45655", + "pull_request:43823|pull_request:43995", + "pull_request:43995|pull_request:45213", + "pull_request:43995|pull_request:44827", + "pull_request:44827|pull_request:45170", + "pull_request:44827|pull_request:46048", + "pull_request:44827|pull_request:45655", + "pull_request:44827|pull_request:45213", + "pull_request:43823|pull_request:45213", + "pull_request:43251|pull_request:43254" + ], + "prepared_review_unit_hash": "12ddc956583868684b2c7dcc90306d3dbae282d1d3d967612f01c080cd619b0d", + "component_budget": { + "node_count": 13, + "item_count": 13, + "soft_pair_count": 18, + "serialized_chars": 34775, + "estimated_input_tokens": 8694, + "estimated_eval_tokens": 17644 + }, + "budget": { + "node_count": 13, + "item_count": 13, + "soft_pair_count": 18, + "serialized_chars": 34775, + "estimated_input_tokens": 8694, + "estimated_eval_tokens": 17644 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Most items are unrelated PRs across different models/features. The only clear duplicate pair is the two `fixed_cross_entropy` PRs, which patch the same function for the same issue.", + "confidence": 0.96, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR #43251 is the best canonical PR for the duplicate pair because it directly targets issue #43240 and presents the more complete, issue-linked fix for `fixed_cross_entropy`.", + "best_issue_reason": null, + "best_pr_reason": "PR #43251 is the strongest representative change in this set: it has an explicit issue target, a focused code-path fix, and matches the concrete bug described in the diff. ", + "soft_edge_verdicts": [ + { + "left": "pull_request:39895", + "right": "pull_request:43424", + "accept": false, + "reason": "Different features: new VideoPrism model package/docs vs an Executorch dynamic-shapes test. No shared code-path or mergeable overlap." + }, + { + "left": "pull_request:41855", + "right": "pull_request:43823", + "accept": false, + "reason": "Unrelated tokenizer API additions for Mistral common backend vs adding the MobileLLM model." + }, + { + "left": "pull_request:41855", + "right": "pull_request:43424", + "accept": false, + "reason": "Tokenizer method addition vs Executorch export test; same broad repo area only, not the same bug/change." + }, + { + "left": "pull_request:42668", + "right": "pull_request:43823", + "accept": false, + "reason": "Processor auto-loading robustness changes vs a new model implementation; too different to be duplicates." + }, + { + "left": "pull_request:43251", + "right": "pull_request:43823", + "accept": false, + "reason": "Loss-function kwargs fix vs adding MobileLLM; no shared underlying bug." + }, + { + "left": "pull_request:43251", + "right": "pull_request:44827", + "accept": false, + "reason": "`cross_entropy` kwargs support fix is unrelated to Mistral4 test/model adjustments." + }, + { + "left": "pull_request:43424", + "right": "pull_request:44827", + "accept": false, + "reason": "Executorch dynamic-shapes export test and Mistral4 test fixes address different concerns." + }, + { + "left": "pull_request:43424", + "right": "pull_request:43823", + "accept": false, + "reason": "Dynamic-shapes export coverage vs MobileLLM model addition are unrelated." + }, + { + "left": "pull_request:43823", + "right": "pull_request:45655", + "accept": false, + "reason": "New MobileLLM model vs config-resolution order fix; different code paths and intents." + }, + { + "left": "pull_request:43823", + "right": "pull_request:43995", + "accept": false, + "reason": "MobileLLM model addition vs Falcon output-collection refactor; not the same change." + }, + { + "left": "pull_request:43995", + "right": "pull_request:45213", + "accept": false, + "reason": "Falcon forward-signature refactor vs a model-creation skill draft; no overlap in the concrete bug/change." + }, + { + "left": "pull_request:43995", + "right": "pull_request:44827", + "accept": false, + "reason": "Falcon interface refactor and Mistral4 fixes are distinct model-specific changes." + }, + { + "left": "pull_request:44827", + "right": "pull_request:45170", + "accept": false, + "reason": "Mistral4 test/model fixes vs CLIP-like `layrnorm` rename correction; separate bugs." + }, + { + "left": "pull_request:44827", + "right": "pull_request:46048", + "accept": false, + "reason": "Mistral4 fixes vs adding a TDT loss kernel; unrelated subsystems." + }, + { + "left": "pull_request:44827", + "right": "pull_request:45655", + "accept": false, + "reason": "Mistral4 test/model fixes are unrelated to class config resolution order." + }, + { + "left": "pull_request:44827", + "right": "pull_request:45213", + "accept": false, + "reason": "Model/test fixes vs a draft model-creation skill; not a duplicate change." + }, + { + "left": "pull_request:43823", + "right": "pull_request:45213", + "accept": false, + "reason": "New MobileLLM model addition vs model-creation skill draft; one is implementation, the other is tooling/docs." + }, + { + "left": "pull_request:43251", + "right": "pull_request:43254", + "accept": true, + "reason": "Both PRs modify `fixed_cross_entropy` in `loss_utils.py` to pass through the same supported `cross_entropy` kwargs for the same issue #43240; they are the same underlying fix." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet and stays conservative. The only accepted pair, PR #43251 and PR #43254, is well supported: both change `src/transformers/loss/loss_utils.py`, both target issue #43240, and both pass supported `cross_entropy` kwargs through the same helper. The other soft-edge verdicts are appropriately negative and do not overstate duplicate evidence." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 2, + "component_count": 27, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-40546-2", + "nodes": [ + "pull_request:40546", + "pull_request:42781" + ], + "soft_pairs": [ + "pull_request:40546|pull_request:42781" + ], + "prepared_review_unit_hash": "c8fac57a1efdc76eb99f3247e320e4de735b1669ab9873136dc753462f6519cb", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 7225, + "estimated_input_tokens": 1807, + "estimated_eval_tokens": 3870 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 7225, + "estimated_input_tokens": 1807, + "estimated_eval_tokens": 3870 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are related VibeVoice PRs, but they are not duplicates: one adds the base VibeVoice implementation, while the other adds a separate realtime/acoustic-tokenizer extension on top of that family.", + "confidence": 0.89, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR #40546 is the better canonical representative because it is the original non-draft VibeVoice implementation and covers the core model integration; #42781 is a later, narrower realtime-focused extension with additional tokenizer/generation work.", + "best_issue_reason": null, + "best_pr_reason": "PR #40546 is the strongest overall representative of the cluster since it is broader, more mature, and closer to the primary VibeVoice landing; #42781 is a distinct follow-on change rather than the same PR.", + "soft_edge_verdicts": [ + { + "left": "pull_request:40546", + "right": "pull_request:42781", + "accept": false, + "reason": "They share the VibeVoice subsystem, but #42781 is a separate realtime/acoustic-tokenizer addition rather than the same concrete code-path fix or the same mergeable change as #40546." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 3, + "component_count": 27, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-40755-2", + "nodes": [ + "pull_request:40755", + "pull_request:41224" + ], + "soft_pairs": [ + "pull_request:40755|pull_request:41224" + ], + "prepared_review_unit_hash": "d9b138efd317c52cd54378377030619618781764c8a96cefa0be603ff7f02336", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4675, + "estimated_input_tokens": 1169, + "estimated_eval_tokens": 2594 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4675, + "estimated_input_tokens": 1169, + "estimated_eval_tokens": 2594 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These two PRs are unrelated: one adds TimesFM covariate forecasting, the other adds DINOv3 image-classification support. They do not fix the same code path or represent one mergeable change.", + "confidence": 0.99, + "canonical_issue_reason": null, + "canonical_pr_reason": "No canonical PR for the cluster; the PRs implement distinct features in different model families and are not duplicates.", + "best_issue_reason": null, + "best_pr_reason": "No single best PR as a cluster representative because the items are heterogeneous and should not be grouped as one change.", + "soft_edge_verdicts": [ + { + "left": "pull_request:40755", + "right": "pull_request:41224", + "accept": false, + "reason": "Different models, different features, and different files/code paths; they are not the same underlying bug or change." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 4, + "component_count": 27, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "cache", + "cache_hit": true, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-41356-2", + "nodes": [ + "pull_request:41356", + "pull_request:44339" + ], + "soft_pairs": [ + "pull_request:41356|pull_request:44339" + ], + "prepared_review_unit_hash": "aead1308c167b0bae6ff11bf9358340baa5e76f816618383418b505a10a9cd01", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 6050, + "estimated_input_tokens": 1513, + "estimated_eval_tokens": 3282 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 6050, + "estimated_input_tokens": 1513, + "estimated_eval_tokens": 3282 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Two DEIMv2 pull requests target the same model family, but they are not the same change: PR 41356 is an early/basic model + image processor add, while PR 44339 is a much broader follow-up that adds loss code, auto-mappings, conversion tooling, modular files, and more integration work.", + "confidence": 0.91, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 44339 is the more complete and representative DEIMv2 implementation; PR 41356 looks like an earlier partial version of the same model bring-up.", + "best_issue_reason": null, + "best_pr_reason": "PR 44339 is the strongest single artifact for the cluster because it subsumes the basic model work and includes the broader integration needed for a full Transformers addition.", + "soft_edge_verdicts": [ + { + "left": "pull_request:41356", + "right": "pull_request:44339", + "accept": false, + "reason": "Same model area, but not the same concrete change: 44339 expands the implementation substantially and adds new code paths beyond the basic PR 41356. They look like related iterations, not mergeable duplicates." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 5, + "component_count": 27, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-41797-3", + "nodes": [ + "pull_request:41797", + "pull_request:45186", + "pull_request:45630" + ], + "soft_pairs": [ + "pull_request:41797|pull_request:45186", + "pull_request:45186|pull_request:45630" + ], + "prepared_review_unit_hash": "682fdc1f74e4375464101dcc610536d03253489e616cf8791c6b0e4f6779ec6d", + "component_budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 2, + "serialized_chars": 10435, + "estimated_input_tokens": 2609, + "estimated_eval_tokens": 5474 + }, + "budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 2, + "serialized_chars": 10435, + "estimated_input_tokens": 2609, + "estimated_eval_tokens": 5474 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are three separate model-addition PRs (DeepSeek OCR, Isaac, Kimi2-6). They share common integration scaffolding, but each targets a different model and codepath, so they are not duplicates.", + "confidence": 0.98, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": "Kimi2-6 is the broadest and most mature-looking integration here, with the highest review activity and the largest set of model/auto-registration changes, but it is still a distinct PR rather than a duplicate of the others.", + "soft_edge_verdicts": [ + { + "left": "pull_request:41797", + "right": "pull_request:45186", + "accept": false, + "reason": "Different models and implementations: DeepSeek OCR vs Isaac. The overlap is only in generic model-registration/docs boilerplate, not the same underlying change." + }, + { + "left": "pull_request:45186", + "right": "pull_request:45630", + "accept": false, + "reason": "Different model families and concrete integrations: Isaac vs Kimi2-6. Shared auto-mapping/docs scaffolding is too broad to treat as the same PR." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 6, + "component_count": 27, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-42975-2", + "nodes": [ + "pull_request:42975", + "pull_request:42976" + ], + "soft_pairs": [ + "pull_request:42975|pull_request:42976" + ], + "prepared_review_unit_hash": "08e51c95551a9519460377214bcdc8027bc0d4af16c3ae606be57ab1ad7fe9d9", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 5935, + "estimated_input_tokens": 1484, + "estimated_eval_tokens": 3224 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 5935, + "estimated_input_tokens": 1484, + "estimated_eval_tokens": 3224 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Both items are broad GitHub Actions upgrade PRs, but they target different action families and different concerns: Node 24 compatibility vs latest-version bumps for docker/build-push-action. They are related in scope yet not the same concrete change, so they should not be deduplicated.", + "confidence": 0.95, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 42975 is the better representative because its title and edits are more specifically about the compatibility-driven workflow action upgrade theme, which is the closer cluster anchor here.", + "best_issue_reason": null, + "best_pr_reason": "PR 42975 is the strongest single representative of the cluster because it touches many workflow actions in a clearly defined compatibility upgrade, while 42976 is a separate version-bump effort focused on docker build actions.", + "soft_edge_verdicts": [ + { + "left": "pull_request:42975", + "right": "pull_request:42976", + "accept": false, + "reason": "Related workflow maintenance, but they update different actions for different purposes and would not naturally be merged into one concrete fix." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 7, + "component_count": 27, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-42978-5", + "nodes": [ + "pull_request:42978", + "pull_request:43448", + "pull_request:43451", + "pull_request:45490", + "pull_request:46162" + ], + "soft_pairs": [ + "pull_request:42978|pull_request:43448", + "pull_request:42978|pull_request:43451", + "pull_request:42978|pull_request:45490", + "pull_request:43448|pull_request:43451", + "pull_request:43448|pull_request:45490", + "pull_request:43451|pull_request:45490", + "pull_request:45490|pull_request:46162" + ], + "prepared_review_unit_hash": "9a01b9fe7c7b4db8bb3d75736c2724a48e77859d31b1b6f644b71f629161cb22", + "component_budget": { + "node_count": 5, + "item_count": 5, + "soft_pair_count": 7, + "serialized_chars": 16554, + "estimated_input_tokens": 4139, + "estimated_eval_tokens": 8534 + }, + "budget": { + "node_count": 5, + "item_count": 5, + "soft_pair_count": 7, + "serialized_chars": 16554, + "estimated_input_tokens": 4139, + "estimated_eval_tokens": 8534 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These PRs are all separate additions or test-only changes across different models and code paths; the shared files are mostly boilerplate model-registration/docs touchpoints, not evidence of duplicate underlying work.", + "confidence": 0.97, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 43451 is the most substantial and actively discussed item in the cluster, but it is not a duplicate of the others; it is just the best representative if one must be chosen.", + "best_issue_reason": null, + "best_pr_reason": "PR 43451 has the broadest scope and highest review activity, making it the strongest standalone representative of the cluster, though not a duplicate target.", + "soft_edge_verdicts": [ + { + "left": "pull_request:42978", + "right": "pull_request:43448", + "accept": false, + "reason": "Different model additions: ViT NEPA vs Molmo. Overlap is limited to shared registration/docs scaffolding, not the same code-path change." + }, + { + "left": "pull_request:42978", + "right": "pull_request:43451", + "accept": false, + "reason": "Different model families and implementations: ViT NEPA vs Molmo2. Shared auto/docs edits are incidental boilerplate." + }, + { + "left": "pull_request:42978", + "right": "pull_request:45490", + "accept": false, + "reason": "Completely different domains: a vision model addition vs a time-series model addition. No same bug or mergeable single change." + }, + { + "left": "pull_request:43448", + "right": "pull_request:43451", + "accept": false, + "reason": "Molmo and Molmo2 are related names but are distinct model releases with separate configs, processors, and tests; not the same underlying change." + }, + { + "left": "pull_request:43448", + "right": "pull_request:45490", + "accept": false, + "reason": "Molmo model addition vs CTSM model addition; only broad model-registry files overlap, which is not enough to treat them as duplicates." + }, + { + "left": "pull_request:43451", + "right": "pull_request:45490", + "accept": false, + "reason": "Molmo2 and CTSM are unrelated model additions touching different architectures and files; shared auto-registration edits are superficial." + }, + { + "left": "pull_request:45490", + "right": "pull_request:46162", + "accept": false, + "reason": "CTSM model integration vs test annotation changes for Florence2/LFM2-VL. The latter is test maintenance, not the same code-path fix or feature." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 8, + "component_count": 27, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-43291-4", + "nodes": [ + "pull_request:43291", + "pull_request:43665", + "pull_request:44815", + "pull_request:46102" + ], + "soft_pairs": [ + "pull_request:43291|pull_request:43665", + "pull_request:43291|pull_request:44815", + "pull_request:43665|pull_request:46102", + "pull_request:43665|pull_request:44815" + ], + "prepared_review_unit_hash": "4155d50ded71537fe1e003f7367ed45759fc4545aef5582129c953827f68724f", + "component_budget": { + "node_count": 4, + "item_count": 4, + "soft_pair_count": 4, + "serialized_chars": 9391, + "estimated_input_tokens": 2348, + "estimated_eval_tokens": 4952 + }, + "budget": { + "node_count": 4, + "item_count": 4, + "soft_pair_count": 4, + "serialized_chars": 9391, + "estimated_input_tokens": 2348, + "estimated_eval_tokens": 4952 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are not duplicates: the PRs target unrelated areas (Whisper decoding, CLIP/ViT test edits, FP8 dequantization, and FSDP tied-weights handling). The only commonality is generic \"fix\" wording and test touchpoints in some diffs.", + "confidence": 0.96, + "canonical_issue_reason": null, + "canonical_pr_reason": "pull_request:46102 is the clearest focused bug fix with a concrete code path (FSDP tied weights / model loading), but it is not a duplicate of the others.", + "best_issue_reason": null, + "best_pr_reason": "pull_request:46102 is the strongest standalone PR: specific title, cohesive FSDP scope, and a focused fix path compared with the more generic or test-heavy alternatives.", + "soft_edge_verdicts": [ + { + "left": "pull_request:43291", + "right": "pull_request:43665", + "accept": false, + "reason": "Different subsystems and problems: Whisper tokenizer decoding vs CLIP/ViT test edits. Superficial \"fix\" similarity only." + }, + { + "left": "pull_request:43291", + "right": "pull_request:44815", + "accept": false, + "reason": "Whisper timestamp decoding and FP8 dequant/model-loading changes are unrelated code paths; no same underlying bug." + }, + { + "left": "pull_request:43665", + "right": "pull_request:46102", + "accept": false, + "reason": "Test-file edits for vision models are not the same issue as FSDP tied-weights/model-loading changes." + }, + { + "left": "pull_request:43665", + "right": "pull_request:44815", + "accept": false, + "reason": "These touch unrelated areas (vision tests vs FP8 quantization/loading). Shared generic wording is insufficient." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 9, + "component_count": 27, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-43488-5", + "nodes": [ + "pull_request:43488", + "pull_request:43532", + "pull_request:45932", + "pull_request:46100", + "pull_request:46114" + ], + "soft_pairs": [ + "pull_request:43488|pull_request:46114", + "pull_request:43488|pull_request:43532", + "pull_request:43532|pull_request:46114", + "pull_request:45932|pull_request:46100", + "pull_request:46100|pull_request:46114" + ], + "prepared_review_unit_hash": "7877d7166add60e71b11c9f8b09302c913da14e5c60f60271fa9f28e0c9320be", + "component_budget": { + "node_count": 5, + "item_count": 5, + "soft_pair_count": 5, + "serialized_chars": 11341, + "estimated_input_tokens": 2836, + "estimated_eval_tokens": 5928 + }, + "budget": { + "node_count": 5, + "item_count": 5, + "soft_pair_count": 5, + "serialized_chars": 11341, + "estimated_input_tokens": 2836, + "estimated_eval_tokens": 5928 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "This cluster is mostly synthetic or test-only PRs. PR 45932 is the only substantive implementation change and is the best representative. The one true duplicate pair is 46100 and 46114, which have identical ViT test edits; the other soft-similarity pairs do not look like the same underlying change.", + "confidence": 0.82, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 45932 is the best canonical PR because it contains the real multi-file implementation/config/CI change, while the others are either synthetic test PRs or unrelated workflow/test edits.", + "best_issue_reason": null, + "best_pr_reason": "PR 45932 is the clearest representative of the cluster\u2019s substantive change set and is more useful as the cluster anchor than the synthetic duplicates.", + "soft_edge_verdicts": [ + { + "left": "pull_request:43488", + "right": "pull_request:46114", + "accept": false, + "reason": "Both touch ViT tests, but 43488 also changes setup/dependency files and has a different purpose; the diffs are not the same underlying change." + }, + { + "left": "pull_request:43488", + "right": "pull_request:43532", + "accept": false, + "reason": "Shared test-file overlap is too weak: 43488 is setup/version/dependency-oriented, while 43532 is workflow/testing-utils oriented." + }, + { + "left": "pull_request:43532", + "right": "pull_request:46114", + "accept": false, + "reason": "43532 changes workflows and testing utilities, whereas 46114 is a forked ViT test-migration PR; they are not the same code change." + }, + { + "left": "pull_request:45932", + "right": "pull_request:46100", + "accept": false, + "reason": "These are different changes: 45932 updates model/config/CI code across many files, while 46100 only injects failing assertions into tests/models/vit/test_modeling_vit.py." + }, + { + "left": "pull_request:46100", + "right": "pull_request:46114", + "accept": true, + "reason": "These appear to be the same synthetic ViT test PR: same file, same inserted assertions, and matching diff preview, so they could plausibly be merged as one PR." + } + ] + }, + "evaluator_result": { + "accept": false, + "feedback": "The summary is not fully grounded in the packet. Most notably, the packet marks pull_request:45932 | pull_request:46100 as deterministic_accept=true, but the analyst says that pair is not the same underlying change and instead calls 46100 | 46114 the only true duplicate pair. That directly conflicts with the provided pair evidence. Also, the claim that 46100 and 46114 are the same synthetic PR is stronger than the packet supports, since their evidence is only soft_similarity, not deterministic acceptance. The representative choice of 45932 is fine, but the soft-edge verdicts and duplicate-pair summary need to be more conservative." + }, + "evaluator_used": true, + "retried": true, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 10, + "component_count": 27, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-43612-2", + "nodes": [ + "pull_request:43612", + "pull_request:43613" + ], + "soft_pairs": [ + "pull_request:43612|pull_request:43613" + ], + "prepared_review_unit_hash": "b99469c6ee2b9bb25e897c96d8d68749c688fae68bad626cfd307cd1df2a7d67", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 6083, + "estimated_input_tokens": 1521, + "estimated_eval_tokens": 3298 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 6083, + "estimated_input_tokens": 1521, + "estimated_eval_tokens": 3298 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Two feature PRs with overlapping scaffolding, but they implement different pipelines and different underlying segmentation tasks; they should not be merged as duplicates.", + "confidence": 0.94, + "canonical_issue_reason": null, + "canonical_pr_reason": "No canonical duplicate PR: 43612 adds Promptable Concept Segmentation, while 43613 adds Promptable Visual Segmentation; they target different code paths and model integrations.", + "best_issue_reason": null, + "best_pr_reason": "PR 43613 is the broader of the two feature additions and touches more model/test integration, so it is the better representative if one must be chosen, though it is still not a duplicate of 43612.", + "soft_edge_verdicts": [ + { + "left": "pull_request:43612", + "right": "pull_request:43613", + "accept": false, + "reason": "Different pipelines/features (concept segmentation vs visual segmentation) with different model and test coverage; shared docs/boilerplate files are not enough to treat them as the same change." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 11, + "component_count": 27, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-43651-2", + "nodes": [ + "pull_request:43651", + "pull_request:43663" + ], + "soft_pairs": [ + "pull_request:43651|pull_request:43663" + ], + "prepared_review_unit_hash": "cdc471dec3cf0047c73971c6e64d50ff26d19b9750dadea2aade31cc0ada433d", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4293, + "estimated_input_tokens": 1074, + "estimated_eval_tokens": 2404 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4293, + "estimated_input_tokens": 1074, + "estimated_eval_tokens": 2404 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Two open Trainer PRs touch the same file but implement different extension points: one changes gradient-accumulation loss scaling, the other refactors signature-column filtering for custom trainers. They are not the same underlying change.", + "confidence": 0.95, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 43651 is the more concrete standalone fix: it changes loss scaling behavior in training_step and adds a supporting property. PR 43663 is a separate refactor for column-filtering extensibility.", + "best_issue_reason": null, + "best_pr_reason": "PR 43651 is the better representative because it has a narrowly defined behavior change and clearer bug-fix intent than the column-filtering refactor in 43663.", + "soft_edge_verdicts": [ + { + "left": "pull_request:43651", + "right": "pull_request:43663", + "accept": false, + "reason": "Different code paths and goals: 43651 adjusts gradient accumulation loss normalization, while 43663 adds an overrideable helper for signature-column filtering. Shared file alone is not enough to treat them as duplicates." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 12, + "component_count": 27, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-43656-2", + "nodes": [ + "pull_request:43656", + "pull_request:43842" + ], + "soft_pairs": [ + "pull_request:43656|pull_request:43842" + ], + "prepared_review_unit_hash": "5963cc1b90f71e31551fbf24df588b6326f9a971b0e7be3981afc138a5c72c8d", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 3664, + "estimated_input_tokens": 916, + "estimated_eval_tokens": 2088 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 3664, + "estimated_input_tokens": 916, + "estimated_eval_tokens": 2088 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Two open PRs touch the same file and the same general TypeAdapter annotation problem in `src/transformers/cli/serve.py`, but they use different fixes and appear to target different concrete failure modes, so they should not be collapsed as duplicates.", + "confidence": 0.78, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR #43656 is the stronger canonical representative: it has the explicit issue link, more discussion/review activity, and a broader runtime-safe typing cleanup in the CLI serve code path.", + "best_issue_reason": null, + "best_pr_reason": "PR #43842 is the more narrowly targeted fix for the specific pydantic-absent `TypeAdapter` NameError, but it is less established than #43656 and appears to be an alternate implementation rather than the same change.", + "soft_edge_verdicts": [ + { + "left": "pull_request:43656", + "right": "pull_request:43842", + "accept": false, + "reason": "Both are in `cli/serve.py` and mention `TypeAdapter`, but the diffs show different remedies (`from __future__ import annotations`/typing cleanup vs `TYPE_CHECKING` import plus quoted annotation) and different issue targets, so they do not look like the same concrete PR change." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 13, + "component_count": 27, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "cache", + "cache_hit": true, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-43779-2", + "nodes": [ + "pull_request:43779", + "pull_request:43816" + ], + "soft_pairs": [ + "pull_request:43779|pull_request:43816" + ], + "prepared_review_unit_hash": "e7c674adcc031aa8883109a5dfc678b4407fb20b10764302092e384d031ba83a", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4362, + "estimated_input_tokens": 1091, + "estimated_eval_tokens": 2438 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4362, + "estimated_input_tokens": 1091, + "estimated_eval_tokens": 2438 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Both pull requests appear to implement the same SwanLab integration change: adding run id/resume support in `integration_utils.py` for issue #43698, with one variant also expanding the docs. They are strong duplicate PR candidates.", + "confidence": 0.94, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR #43816 is the cleaner representative: it carries the same SwanLab id/resume support plus the accompanying documentation updates, making it the more complete version of the shared change.", + "best_issue_reason": null, + "best_pr_reason": "PR #43816 is the best global representative because it appears to include the full feature change and docs for SwanLab run id/resume support, while #43779 is the narrower implementation-focused variant.", + "soft_edge_verdicts": [ + { + "left": "pull_request:43779", + "right": "pull_request:43816", + "accept": true, + "reason": "Same underlying SwanLab callback enhancement for id/resume support in the same file and same issue target; the doc differences do not change the core code-path change, so these are duplicate PRs." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet: both PRs target issue #43698, touch the same file, and the previews indicate the same SwanLab id/resume support with #43816 adding docs. The duplicate/mergeable verdict is conservative enough given the shared target and overlapping code change." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 14, + "component_count": 27, + "review_unit_index": 1, + "review_unit_count": 2, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-43996-28", + "nodes": [ + "pull_request:43996", + "pull_request:43997", + "pull_request:44007", + "pull_request:44010", + "pull_request:44013", + "pull_request:44017", + "pull_request:44018", + "pull_request:44019", + "pull_request:44044", + "pull_request:44056", + "pull_request:44059", + "pull_request:44066", + "pull_request:44068", + "pull_request:44071", + "pull_request:44072", + "pull_request:44073", + "pull_request:44074", + "pull_request:44076", + "pull_request:44085", + "pull_request:44086", + "pull_request:44098", + "pull_request:44101", + "pull_request:44116", + "pull_request:44129", + "pull_request:44154", + "pull_request:44161", + "pull_request:44722" + ], + "soft_pairs": [ + "pull_request:43996|pull_request:44085", + "pull_request:43996|pull_request:44044", + "pull_request:44066|pull_request:44085", + "pull_request:44007|pull_request:44072", + "pull_request:44072|pull_request:44722", + "pull_request:44066|pull_request:44072", + "pull_request:44013|pull_request:44044", + "pull_request:44066|pull_request:44086", + "pull_request:44066|pull_request:44071", + "pull_request:44018|pull_request:44068", + "pull_request:44066|pull_request:44068", + "pull_request:44018|pull_request:44066", + "pull_request:44019|pull_request:44722", + "pull_request:44086|pull_request:44722", + "pull_request:44019|pull_request:44085", + "pull_request:44018|pull_request:44019", + "pull_request:44019|pull_request:44071", + "pull_request:44068|pull_request:44116", + "pull_request:44068|pull_request:44722", + "pull_request:44018|pull_request:44116", + "pull_request:44071|pull_request:44722", + "pull_request:44018|pull_request:44086", + "pull_request:44018|pull_request:44085", + "pull_request:44068|pull_request:44086", + "pull_request:44071|pull_request:44085", + "pull_request:44059|pull_request:44086", + "pull_request:44085|pull_request:44086", + "pull_request:44098|pull_request:44101", + "pull_request:44059|pull_request:44722", + "pull_request:44018|pull_request:44071", + "pull_request:44017|pull_request:44161", + "pull_request:44017|pull_request:44722", + "pull_request:44017|pull_request:44073", + "pull_request:44068|pull_request:44071", + "pull_request:43997|pull_request:44073", + "pull_request:44071|pull_request:44086", + "pull_request:44073|pull_request:44101", + "pull_request:44017|pull_request:44071", + "pull_request:44056|pull_request:44101", + "pull_request:44059|pull_request:44071", + "pull_request:44073|pull_request:44161", + "pull_request:44059|pull_request:44068", + "pull_request:44018|pull_request:44059", + "pull_request:44101|pull_request:44161", + "pull_request:43997|pull_request:44017", + "pull_request:43997|pull_request:44056", + "pull_request:44071|pull_request:44076", + "pull_request:44074|pull_request:44101", + "pull_request:44056|pull_request:44161", + "pull_request:43997|pull_request:44018", + "pull_request:43997|pull_request:44161", + "pull_request:44074|pull_request:44161", + "pull_request:44018|pull_request:44161", + "pull_request:43997|pull_request:44071", + "pull_request:44010|pull_request:44071", + "pull_request:44010|pull_request:44073", + "pull_request:44018|pull_request:44129", + "pull_request:43997|pull_request:44074", + "pull_request:44071|pull_request:44129", + "pull_request:43997|pull_request:44129", + "pull_request:44010|pull_request:44018", + "pull_request:43996|pull_request:44161", + "pull_request:43996|pull_request:44017", + "pull_request:44010|pull_request:44044", + "pull_request:44044|pull_request:44074", + "pull_request:44066|pull_request:44722", + "pull_request:44085|pull_request:44722", + "pull_request:44085|pull_request:44129", + "pull_request:43996|pull_request:44007", + "pull_request:43996|pull_request:44129", + "pull_request:44044|pull_request:44129", + "pull_request:44044|pull_request:44154" + ], + "prepared_review_unit_hash": "75773f077841fe924b1af82b6bb723107ba20ff9dcaef128f3441637aeaa8bdc", + "component_budget": { + "node_count": 28, + "item_count": 28, + "soft_pair_count": 78, + "serialized_chars": 84901, + "estimated_input_tokens": 21226, + "estimated_eval_tokens": 42708 + }, + "budget": { + "node_count": 27, + "item_count": 27, + "soft_pair_count": 72, + "serialized_chars": 75096, + "estimated_input_tokens": 18774, + "estimated_eval_tokens": 37804 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is dominated by model-specific output-tracing refactors. The only clear duplicate groups are the GPT-Neo pair and the GPT-J family (including the later GPT-J/CodeGen follow-up); the rest are same-theme but different model/code paths.", + "confidence": 0.84, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 44722 is the strongest representative of the GPT-J output-tracing refactor cluster: it is the latest and most complete decorator-based migration on the same code path, with the clearest overlap to the earlier GPT-J PRs.", + "best_issue_reason": null, + "best_pr_reason": "PR 44722 best represents the duplicate cluster because it most completely captures the shared GPT-J output-tracing change and also subsumes the related CodeGen work.", + "soft_edge_verdicts": [ + { + "left": "pull_request:43996", + "right": "pull_request:44085", + "accept": false, + "reason": "Different model families and code paths; same output-tracing theme is too broad." + }, + { + "left": "pull_request:43996", + "right": "pull_request:44044", + "accept": false, + "reason": "Unrelated models (FNet/CVT vs DeBERTa); not the same underlying change." + }, + { + "left": "pull_request:44066", + "right": "pull_request:44085", + "accept": true, + "reason": "Both are GPT-J output-tracing refactors on the same model code path; the later PR is a broader variant, but the underlying change matches." + }, + { + "left": "pull_request:44007", + "right": "pull_request:44072", + "accept": false, + "reason": "Different vision models and different implementations; only the refactor pattern is shared." + }, + { + "left": "pull_request:44072", + "right": "pull_request:44722", + "accept": false, + "reason": "EfficientNet and GPT-J/CodeGen are unrelated; same refactor style does not make them duplicates." + }, + { + "left": "pull_request:44066", + "right": "pull_request:44072", + "accept": false, + "reason": "GPT-J/CodeGen vs EfficientNet; different code paths and model families." + }, + { + "left": "pull_request:44013", + "right": "pull_request:44044", + "accept": false, + "reason": "MobileNetV2 and DeBERTa are unrelated models." + }, + { + "left": "pull_request:44066", + "right": "pull_request:44086", + "accept": false, + "reason": "GPT-J/CodeGen vs MGP-STR; not the same bug/change." + }, + { + "left": "pull_request:44066", + "right": "pull_request:44071", + "accept": false, + "reason": "GPT-J/CodeGen and MPT are different model implementations." + }, + { + "left": "pull_request:44018", + "right": "pull_request:44068", + "accept": true, + "reason": "Same GPT-Neo output-tracing migration on the same model family and code path." + }, + { + "left": "pull_request:44066", + "right": "pull_request:44068", + "accept": false, + "reason": "GPT-J/CodeGen and GPT-Neo are unrelated model code paths." + }, + { + "left": "pull_request:44018", + "right": "pull_request:44066", + "accept": false, + "reason": "Different model families (GPT-Neo vs GPT-J/CodeGen); same issue target is not enough." + }, + { + "left": "pull_request:44019", + "right": "pull_request:44722", + "accept": false, + "reason": "ResNet and GPT-J/CodeGen are unrelated; no shared concrete bug/change." + }, + { + "left": "pull_request:44086", + "right": "pull_request:44722", + "accept": false, + "reason": "MGP-STR and GPT-J/CodeGen are unrelated code paths." + }, + { + "left": "pull_request:44019", + "right": "pull_request:44085", + "accept": false, + "reason": "ResNet and GPT-J are unrelated model implementations." + }, + { + "left": "pull_request:44018", + "right": "pull_request:44019", + "accept": false, + "reason": "GPT-Neo and ResNet are unrelated; only the refactor theme overlaps." + }, + { + "left": "pull_request:44019", + "right": "pull_request:44071", + "accept": false, + "reason": "ResNet and MPT are unrelated models and code paths." + }, + { + "left": "pull_request:44068", + "right": "pull_request:44116", + "accept": false, + "reason": "GPT-Neo and Flaubert are unrelated model families." + }, + { + "left": "pull_request:44068", + "right": "pull_request:44722", + "accept": false, + "reason": "GPT-Neo and GPT-J/CodeGen are different model families and refactors." + }, + { + "left": "pull_request:44018", + "right": "pull_request:44116", + "accept": false, + "reason": "GPT-Neo and Flaubert are unrelated code paths." + }, + { + "left": "pull_request:44071", + "right": "pull_request:44722", + "accept": false, + "reason": "MPT and GPT-J/CodeGen are unrelated model implementations." + }, + { + "left": "pull_request:44018", + "right": "pull_request:44086", + "accept": false, + "reason": "GPT-Neo and MGP-STR are unrelated; same output-tracing pattern is too broad." + }, + { + "left": "pull_request:44018", + "right": "pull_request:44085", + "accept": false, + "reason": "GPT-Neo and GPT-J are different model code paths." + }, + { + "left": "pull_request:44068", + "right": "pull_request:44086", + "accept": false, + "reason": "GPT-Neo and MGP-STR are unrelated." + }, + { + "left": "pull_request:44071", + "right": "pull_request:44086", + "accept": false, + "reason": "MPT and MGP-STR are unrelated model refactors." + }, + { + "left": "pull_request:44059", + "right": "pull_request:44086", + "accept": false, + "reason": "GPT2 and MGP-STR are unrelated code paths." + }, + { + "left": "pull_request:44085", + "right": "pull_request:44086", + "accept": false, + "reason": "GPT-J and MGP-STR are unrelated models." + }, + { + "left": "pull_request:44098", + "right": "pull_request:44101", + "accept": false, + "reason": "ViLT and Flaubert/XLM are unrelated model families." + }, + { + "left": "pull_request:44059", + "right": "pull_request:44722", + "accept": false, + "reason": "GPT2 and GPT-J/CodeGen are different model implementations." + }, + { + "left": "pull_request:44018", + "right": "pull_request:44071", + "accept": false, + "reason": "GPT-Neo and MPT are unrelated code paths." + }, + { + "left": "pull_request:44059", + "right": "pull_request:44068", + "accept": false, + "reason": "GPT2 and GPT-Neo are different model families." + }, + { + "left": "pull_request:44071", + "right": "pull_request:44085", + "accept": false, + "reason": "MPT and GPT-J are unrelated models." + }, + { + "left": "pull_request:44059", + "right": "pull_request:44071", + "accept": false, + "reason": "GPT2 and MPT are unrelated code paths." + }, + { + "left": "pull_request:44073", + "right": "pull_request:44161", + "accept": false, + "reason": "VisualBERT and LongT5 are unrelated models." + }, + { + "left": "pull_request:44018", + "right": "pull_request:44059", + "accept": false, + "reason": "GPT-Neo and GPT2 are different model implementations." + }, + { + "left": "pull_request:44101", + "right": "pull_request:44161", + "accept": false, + "reason": "Flaubert/XLM and LongT5 are unrelated code paths." + }, + { + "left": "pull_request:43997", + "right": "pull_request:44017", + "accept": false, + "reason": "RegNet and SegFormer are different vision models." + }, + { + "left": "pull_request:43997", + "right": "pull_request:44056", + "accept": false, + "reason": "RegNet and MPNet are unrelated models." + }, + { + "left": "pull_request:44071", + "right": "pull_request:44076", + "accept": false, + "reason": "MPT and ImageGPT are unrelated model implementations." + }, + { + "left": "pull_request:44074", + "right": "pull_request:44101", + "accept": false, + "reason": "TextNet and Flaubert/XLM are unrelated code paths." + }, + { + "left": "pull_request:44056", + "right": "pull_request:44161", + "accept": false, + "reason": "MPNet and LongT5 are unrelated models." + }, + { + "left": "pull_request:43997", + "right": "pull_request:44018", + "accept": false, + "reason": "RegNet and GPT-Neo are unrelated model families." + }, + { + "left": "pull_request:43997", + "right": "pull_request:44161", + "accept": false, + "reason": "RegNet and LongT5 are unrelated code paths." + }, + { + "left": "pull_request:44074", + "right": "pull_request:44161", + "accept": false, + "reason": "TextNet and LongT5 are unrelated models." + }, + { + "left": "pull_request:44018", + "right": "pull_request:44161", + "accept": false, + "reason": "GPT-Neo and LongT5 are unrelated code paths." + }, + { + "left": "pull_request:43997", + "right": "pull_request:44071", + "accept": false, + "reason": "RegNet and MPT are unrelated models." + }, + { + "left": "pull_request:44010", + "right": "pull_request:44071", + "accept": false, + "reason": "SqueezeBert and MPT are unrelated model implementations." + }, + { + "left": "pull_request:44010", + "right": "pull_request:44073", + "accept": false, + "reason": "SqueezeBert and VisualBERT are different models." + }, + { + "left": "pull_request:44018", + "right": "pull_request:44044", + "accept": false, + "reason": "GPT-Neo and DeBERTa are unrelated code paths." + }, + { + "left": "pull_request:44044", + "right": "pull_request:44074", + "accept": false, + "reason": "DeBERTa and TextNet are unrelated models." + }, + { + "left": "pull_request:44066", + "right": "pull_request:44722", + "accept": true, + "reason": "Same GPT-J/CodeGen output-tracing refactor on overlapping code paths; the later PR looks like a broader rework of the same change." + }, + { + "left": "pull_request:44085", + "right": "pull_request:44722", + "accept": true, + "reason": "Both target the GPT-J output-tracing migration on the same underlying code path, with the later PR expanding the refactor." + }, + { + "left": "pull_request:44085", + "right": "pull_request:44129", + "accept": false, + "reason": "GPT-J and SpeechT5 are unrelated models." + }, + { + "left": "pull_request:44044", + "right": "pull_request:44129", + "accept": false, + "reason": "DeBERTa and SpeechT5 are unrelated code paths." + }, + { + "left": "pull_request:44044", + "right": "pull_request:44154", + "accept": false, + "reason": "DeBERTa and VITS are unrelated model implementations." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The cluster summary is conservative and grounded in the packet. The accepted soft edges are limited to the clearly mergeable GPT-Neo and GPT-J/CodeGen follow-up chains, and the rejected edges avoid overgeneralizing across unrelated model families/code paths." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 14, + "component_count": 27, + "review_unit_index": 2, + "review_unit_count": 2, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-43996-28", + "nodes": [ + "pull_request:43997", + "pull_request:44010", + "pull_request:44024", + "pull_request:44056", + "pull_request:44073", + "pull_request:44074", + "pull_request:44076", + "pull_request:44129" + ], + "soft_pairs": [ + "pull_request:44024|pull_request:44076", + "pull_request:44073|pull_request:44074", + "pull_request:44056|pull_request:44073", + "pull_request:44056|pull_request:44074", + "pull_request:43997|pull_request:44010", + "pull_request:44010|pull_request:44129" + ], + "prepared_review_unit_hash": "247ede06c9acf6503c18e57190c12d34cbd51f224c2527df5a6b22a969a1bc22", + "component_budget": { + "node_count": 28, + "item_count": 28, + "soft_pair_count": 78, + "serialized_chars": 84901, + "estimated_input_tokens": 21226, + "estimated_eval_tokens": 42708 + }, + "budget": { + "node_count": 8, + "item_count": 8, + "soft_pair_count": 6, + "serialized_chars": 17126, + "estimated_input_tokens": 4282, + "estimated_eval_tokens": 8820 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "These PRs share a broad theme of migrating model implementations to standardized output tracing/capture hooks, but they are model-specific refactors rather than duplicates of the same concrete code path. I would keep one representative PR for the cluster theme, but reject all soft pair merges.", + "confidence": 0.88, + "canonical_issue_reason": null, + "canonical_pr_reason": "44073 is a good canonical PR for the cluster because it clearly represents the recurring standardized-output refactor pattern (adding can_return_tuple/capture_outputs and model-specific output hooks) without being dominated by an unrelated issue target.", + "best_issue_reason": null, + "best_pr_reason": "44129 is the strongest standalone PR in the set because it has an explicit issue target and a concrete model-specific output-tracing refactor, making it the most actionable single change even though it is not a duplicate of the others.", + "soft_edge_verdicts": [ + { + "left": "pull_request:44024", + "right": "pull_request:44076", + "accept": false, + "reason": "Both are output-tracing refactors, but one is FocalNet and the other is ImageGPT; they touch different model code paths and are not the same fix." + }, + { + "left": "pull_request:44073", + "right": "pull_request:44074", + "accept": false, + "reason": "VisualBert vs TextNet are separate implementations; this is the same general migration style, not one mergeable duplicate PR." + }, + { + "left": "pull_request:44056", + "right": "pull_request:44073", + "accept": false, + "reason": "MPNet and VisualBert both adopt capture_outputs, but they are distinct model-specific changes with different hooks and forward paths." + }, + { + "left": "pull_request:44056", + "right": "pull_request:44074", + "accept": false, + "reason": "These are parallel refactors in different models, not the same underlying bug or change." + }, + { + "left": "pull_request:43997", + "right": "pull_request:44010", + "accept": false, + "reason": "RegNet and SqueezeBert are unrelated model code paths; similarity is only the shared output-tracing migration pattern." + }, + { + "left": "pull_request:44010", + "right": "pull_request:44129", + "accept": false, + "reason": "SqueezeBert and SpeechT5 are different model implementations, and the explicit issue target on SpeechT5 does not make these the same PR." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 15, + "component_count": 27, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-43998-14", + "nodes": [ + "pull_request:43998", + "pull_request:43999", + "pull_request:44000", + "pull_request:44001", + "pull_request:44002", + "pull_request:44003", + "pull_request:44004", + "pull_request:44025", + "pull_request:44026", + "pull_request:44027", + "pull_request:44028", + "pull_request:44029", + "pull_request:44030", + "pull_request:46106" + ], + "soft_pairs": [ + "pull_request:43998|pull_request:44002", + "pull_request:43998|pull_request:44000", + "pull_request:43998|pull_request:44001", + "pull_request:43998|pull_request:44003", + "pull_request:43998|pull_request:44029", + "pull_request:43999|pull_request:44001", + "pull_request:43999|pull_request:44002", + "pull_request:43999|pull_request:44025", + "pull_request:43999|pull_request:44026", + "pull_request:43999|pull_request:44027", + "pull_request:44000|pull_request:44001", + "pull_request:44000|pull_request:44002", + "pull_request:44000|pull_request:44003", + "pull_request:44000|pull_request:44025", + "pull_request:44001|pull_request:44002", + "pull_request:44001|pull_request:44003", + "pull_request:44001|pull_request:44025", + "pull_request:44002|pull_request:44029", + "pull_request:44002|pull_request:44003", + "pull_request:44003|pull_request:44028", + "pull_request:44003|pull_request:44004", + "pull_request:43998|pull_request:44004", + "pull_request:44001|pull_request:44004", + "pull_request:44002|pull_request:44004", + "pull_request:44000|pull_request:44004", + "pull_request:44025|pull_request:44026", + "pull_request:44025|pull_request:44027", + "pull_request:44025|pull_request:44028", + "pull_request:44025|pull_request:44029", + "pull_request:44025|pull_request:44030", + "pull_request:44026|pull_request:44029", + "pull_request:44026|pull_request:44027", + "pull_request:44026|pull_request:44028", + "pull_request:44026|pull_request:44030", + "pull_request:44027|pull_request:44028", + "pull_request:44027|pull_request:44029", + "pull_request:44027|pull_request:44030", + "pull_request:44028|pull_request:44029", + "pull_request:44028|pull_request:44030", + "pull_request:44029|pull_request:44030", + "pull_request:44027|pull_request:46106", + "pull_request:44026|pull_request:46106", + "pull_request:44001|pull_request:46106", + "pull_request:44002|pull_request:46106", + "pull_request:43998|pull_request:46106" + ], + "prepared_review_unit_hash": "6181569103897973301d948050265797e80e0ed8571c1a212309bf20cf1128a2", + "component_budget": { + "node_count": 14, + "item_count": 14, + "soft_pair_count": 45, + "serialized_chars": 43785, + "estimated_input_tokens": 10947, + "estimated_eval_tokens": 22150 + }, + "budget": { + "node_count": 14, + "item_count": 14, + "soft_pair_count": 45, + "serialized_chars": 43785, + "estimated_input_tokens": 10947, + "estimated_eval_tokens": 22150 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These PRs are mostly independent, model-specific output-tracing refactors that share a naming pattern but not the same concrete change. The only unrelated item is the docs toctree fix. I found no duplicate-worthy pairings here under a strict PR policy.", + "confidence": 0.9, + "canonical_issue_reason": null, + "canonical_pr_reason": "#44003 is the broadest and most representative output-tracing refactor in the set, touching two related model implementations and the shared output-capturing plumbing.", + "best_issue_reason": null, + "best_pr_reason": "#44003 is the strongest cluster representative: it is a substantive code change, spans multiple related files, and best matches the cluster\u2019s core theme of output-tracing refactoring.", + "soft_edge_verdicts": [ + { + "left": "pull_request:43998", + "right": "pull_request:44002", + "accept": false, + "reason": "Different model families and files; same refactor theme but not the same concrete change." + }, + { + "left": "pull_request:43998", + "right": "pull_request:44000", + "accept": false, + "reason": "Different model families and files; same refactor theme but not the same concrete change." + }, + { + "left": "pull_request:43998", + "right": "pull_request:44001", + "accept": false, + "reason": "Different model families and files; same refactor theme but not the same concrete change." + }, + { + "left": "pull_request:43998", + "right": "pull_request:44003", + "accept": false, + "reason": "Different model families and files; same refactor theme but not the same concrete change." + }, + { + "left": "pull_request:43998", + "right": "pull_request:44004", + "accept": false, + "reason": "Different code paths and files; both mention output tracing, but they are separate model changes." + }, + { + "left": "pull_request:43998", + "right": "pull_request:44029", + "accept": false, + "reason": "Different model families and files; same refactor theme but not the same concrete change." + }, + { + "left": "pull_request:43998", + "right": "pull_request:46106", + "accept": false, + "reason": "Unrelated docs toctree change, not an output-tracing refactor." + }, + { + "left": "pull_request:43999", + "right": "pull_request:44001", + "accept": false, + "reason": "Different model families and files; same refactor theme but not the same concrete change." + }, + { + "left": "pull_request:43999", + "right": "pull_request:44002", + "accept": false, + "reason": "Different model families and files; same refactor theme but not the same concrete change." + }, + { + "left": "pull_request:43999", + "right": "pull_request:44025", + "accept": false, + "reason": "Different model families and files; same refactor theme but not the same concrete change." + }, + { + "left": "pull_request:43999", + "right": "pull_request:44026", + "accept": false, + "reason": "Different model families and files; same refactor theme but not the same concrete change." + }, + { + "left": "pull_request:43999", + "right": "pull_request:44027", + "accept": false, + "reason": "Different model families and files; same refactor theme but not the same concrete change." + }, + { + "left": "pull_request:44000", + "right": "pull_request:44001", + "accept": false, + "reason": "Different model families and files; same refactor theme but not the same concrete change." + }, + { + "left": "pull_request:44000", + "right": "pull_request:44002", + "accept": false, + "reason": "Different model families and files; same refactor theme but not the same concrete change." + }, + { + "left": "pull_request:44000", + "right": "pull_request:44003", + "accept": false, + "reason": "Different model families and files; same refactor theme but not the same concrete change." + }, + { + "left": "pull_request:44000", + "right": "pull_request:44004", + "accept": false, + "reason": "Different code paths and files; both mention output tracing, but they are separate model changes." + }, + { + "left": "pull_request:44000", + "right": "pull_request:44025", + "accept": false, + "reason": "Different model families and files; same refactor theme but not the same concrete change." + }, + { + "left": "pull_request:44001", + "right": "pull_request:44002", + "accept": false, + "reason": "Different model families and files; same refactor theme but not the same concrete change." + }, + { + "left": "pull_request:44001", + "right": "pull_request:44003", + "accept": false, + "reason": "Different model families and files; same refactor theme but not the same concrete change." + }, + { + "left": "pull_request:44001", + "right": "pull_request:44004", + "accept": false, + "reason": "Different code paths and files; both mention output tracing, but they are separate model changes." + }, + { + "left": "pull_request:44001", + "right": "pull_request:44025", + "accept": false, + "reason": "Different model families and files; same refactor theme but not the same concrete change." + }, + { + "left": "pull_request:44001", + "right": "pull_request:46106", + "accept": false, + "reason": "Unrelated docs toctree change, not an output-tracing refactor." + }, + { + "left": "pull_request:44002", + "right": "pull_request:44003", + "accept": false, + "reason": "Different model families and files; same refactor theme but not the same concrete change." + }, + { + "left": "pull_request:44002", + "right": "pull_request:44004", + "accept": false, + "reason": "Different code paths and files; both mention output tracing, but they are separate model changes." + }, + { + "left": "pull_request:44002", + "right": "pull_request:44029", + "accept": false, + "reason": "Different model families and files; same refactor theme but not the same concrete change." + }, + { + "left": "pull_request:44002", + "right": "pull_request:46106", + "accept": false, + "reason": "Unrelated docs toctree change, not an output-tracing refactor." + }, + { + "left": "pull_request:44003", + "right": "pull_request:44004", + "accept": false, + "reason": "Different model families and files; same refactor theme but not the same concrete change." + }, + { + "left": "pull_request:44003", + "right": "pull_request:44028", + "accept": false, + "reason": "Different model families and files; same refactor theme but not the same concrete change." + }, + { + "left": "pull_request:44003", + "right": "pull_request:44030", + "accept": false, + "reason": "Different model families and files; same refactor theme but not the same concrete change." + }, + { + "left": "pull_request:43998", + "right": "pull_request:44004", + "accept": false, + "reason": "Different code paths and files; both mention output tracing, but they are separate model changes." + }, + { + "left": "pull_request:44001", + "right": "pull_request:44004", + "accept": false, + "reason": "Different code paths and files; both mention output tracing, but they are separate model changes." + }, + { + "left": "pull_request:44002", + "right": "pull_request:44004", + "accept": false, + "reason": "Different code paths and files; both mention output tracing, but they are separate model changes." + }, + { + "left": "pull_request:44000", + "right": "pull_request:44004", + "accept": false, + "reason": "Different code paths and files; both mention output tracing, but they are separate model changes." + }, + { + "left": "pull_request:44025", + "right": "pull_request:44026", + "accept": false, + "reason": "Different model families and files; same refactor theme but not the same concrete change." + }, + { + "left": "pull_request:44025", + "right": "pull_request:44027", + "accept": false, + "reason": "Different model families and files; same refactor theme but not the same concrete change." + }, + { + "left": "pull_request:44025", + "right": "pull_request:44028", + "accept": false, + "reason": "Different model families and files; same refactor theme but not the same concrete change." + }, + { + "left": "pull_request:44025", + "right": "pull_request:44029", + "accept": false, + "reason": "Different model families and files; same refactor theme but not the same concrete change." + }, + { + "left": "pull_request:44025", + "right": "pull_request:44030", + "accept": false, + "reason": "Different model families and files; same refactor theme but not the same concrete change." + }, + { + "left": "pull_request:44026", + "right": "pull_request:44027", + "accept": false, + "reason": "Different model families and files; same refactor theme but not the same concrete change." + }, + { + "left": "pull_request:44026", + "right": "pull_request:44028", + "accept": false, + "reason": "Different model families and files; same refactor theme but not the same concrete change." + }, + { + "left": "pull_request:44026", + "right": "pull_request:44029", + "accept": false, + "reason": "Different model families and files; same refactor theme but not the same concrete change." + }, + { + "left": "pull_request:44026", + "right": "pull_request:44030", + "accept": false, + "reason": "Different model families and files; same refactor theme but not the same concrete change." + }, + { + "left": "pull_request:44026", + "right": "pull_request:46106", + "accept": false, + "reason": "Unrelated docs toctree change, not an output-tracing refactor." + }, + { + "left": "pull_request:44027", + "right": "pull_request:44028", + "accept": false, + "reason": "Different model families and files; same refactor theme but not the same concrete change." + }, + { + "left": "pull_request:44027", + "right": "pull_request:44029", + "accept": false, + "reason": "Different model families and files; same refactor theme but not the same concrete change." + }, + { + "left": "pull_request:44027", + "right": "pull_request:44030", + "accept": false, + "reason": "Different model families and files; same refactor theme but not the same concrete change." + }, + { + "left": "pull_request:44027", + "right": "pull_request:46106", + "accept": false, + "reason": "Unrelated docs toctree change, not an output-tracing refactor." + }, + { + "left": "pull_request:44028", + "right": "pull_request:44029", + "accept": false, + "reason": "Different model families and files; same refactor theme but not the same concrete change." + }, + { + "left": "pull_request:44028", + "right": "pull_request:44030", + "accept": false, + "reason": "Different model families and files; same refactor theme but not the same concrete change." + }, + { + "left": "pull_request:44029", + "right": "pull_request:44030", + "accept": false, + "reason": "Different model families and files; same refactor theme but not the same concrete change." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 16, + "component_count": 27, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-44662-3", + "nodes": [ + "pull_request:44662", + "pull_request:45075", + "pull_request:45401" + ], + "soft_pairs": [ + "pull_request:44662|pull_request:45401", + "pull_request:44662|pull_request:45075", + "pull_request:45075|pull_request:45401" + ], + "prepared_review_unit_hash": "7625955c16f8b35bb35b75d769b140bf1fb46062d2064b99f7419714bcd66749", + "component_budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 3, + "serialized_chars": 10911, + "estimated_input_tokens": 2728, + "estimated_eval_tokens": 5712 + }, + "budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 3, + "serialized_chars": 10911, + "estimated_input_tokens": 2728, + "estimated_eval_tokens": 5712 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are three separate model-addition PRs for unrelated models; the overlap is mostly boilerplate docs, auto-registry wiring, and tests, not a shared bug or the same change.", + "confidence": 0.98, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": "PR 45075 is the strongest representative if one must be picked: it is a substantial, non-draft model integration with high review activity and broad implementation coverage, but it is still not a duplicate of the others.", + "soft_edge_verdicts": [ + { + "left": "pull_request:44662", + "right": "pull_request:45401", + "accept": false, + "reason": "Different model integrations (PenguinVL vs Voxtral TTS); shared files are only generic auto-registry/docs plumbing, not the same underlying code-path change." + }, + { + "left": "pull_request:44662", + "right": "pull_request:45075", + "accept": false, + "reason": "Both add separate vision-language models, but they target different models and implementations; shared files are generic registration/docs changes, so this is not a duplicate PR." + }, + { + "left": "pull_request:45075", + "right": "pull_request:45401", + "accept": false, + "reason": "DeepSeek-OCR-2 and Voxtral TTS are unrelated model additions with distinct architecture and code paths; overlap is limited to standard Transformers integration scaffolding." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 17, + "component_count": 27, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-44979-2", + "nodes": [ + "pull_request:44979", + "pull_request:45363" + ], + "soft_pairs": [ + "pull_request:44979|pull_request:45363" + ], + "prepared_review_unit_hash": "596f4bbf30794e0eb641bf5f03fcdd76a6804a515b83d58d1708bb3a3ade686a", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4407, + "estimated_input_tokens": 1102, + "estimated_eval_tokens": 2460 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4407, + "estimated_input_tokens": 1102, + "estimated_eval_tokens": 2460 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These PRs are related to kernel/module fusion, but they target different abstractions and code paths: one introduces a generic Module Fusion API, while the other adds n-to-1 fusion through KernelConfig and integration hooks. They are not duplicates.", + "confidence": 0.86, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR #44979 is the cleaner core API proposal: it adds a new `module_fusion.py` module plus tests, centered on the fusion abstraction itself. It is more canonical for the general feature area than the more specialized kernel-config integration PR.", + "best_issue_reason": null, + "best_pr_reason": "PR #45363 is the most complete user-facing implementation in this cluster: it wires fusion into `modeling_utils`, `integrations`, and `KernelConfig`, and includes a runnable example. It is the strongest standalone PR, though it addresses a different mechanism than #44979.", + "soft_edge_verdicts": [ + { + "left": "pull_request:44979", + "right": "pull_request:45363", + "accept": false, + "reason": "They both concern fusion, but #44979 is a generic module fusion API while #45363 implements n-to-1 kernel fusion via `KernelConfig` and integration plumbing. Different concrete mechanisms and merge targets." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 18, + "component_count": 27, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-45086-2", + "nodes": [ + "pull_request:45086", + "pull_request:45317" + ], + "soft_pairs": [ + "pull_request:45086|pull_request:45317" + ], + "prepared_review_unit_hash": "7f07bf70f36bbd57c88a72d1d42bacce5b6a391334cfcd97cf2c3b03a1ee0798", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4414, + "estimated_input_tokens": 1104, + "estimated_eval_tokens": 2464 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4414, + "estimated_input_tokens": 1104, + "estimated_eval_tokens": 2464 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Both PRs fix the same _patch_mistral_regex AttributeError by switching from backend_tokenizer.pre_tokenizer to tokenizer.pre_tokenizer; 45317 is the more complete later patch with tests.", + "confidence": 0.99, + "canonical_issue_reason": null, + "canonical_pr_reason": "pull_request:45317 is the stronger canonical PR: it targets the same bug as 45086, adds tests, and appears to be the fuller implementation of the same fix.", + "best_issue_reason": null, + "best_pr_reason": "pull_request:45317 is the best representative PR because it covers the same concrete code-path bug and includes test coverage, making it the most complete version of the fix.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45086", + "right": "pull_request:45317", + "accept": true, + "reason": "Same underlying AttributeError fix in the same file and code path; the diff is effectively the same change, with 45317 adding tests." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 19, + "component_count": 27, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-45115-2", + "nodes": [ + "pull_request:45115", + "pull_request:45569" + ], + "soft_pairs": [ + "pull_request:45115|pull_request:45569" + ], + "prepared_review_unit_hash": "aedcce9b4cec296758b56f74354b0392bf5007111dadac7f3e7377e232196dc6", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 6066, + "estimated_input_tokens": 1517, + "estimated_eval_tokens": 3290 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 6066, + "estimated_input_tokens": 1517, + "estimated_eval_tokens": 3290 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Two PRs touch Nemotron H files, but they are different changes: one is a base-model refactor/conversion-mapping update, the other is a broader split into NemotronH dense/sparse models with docs and tests. They should not be treated as duplicates.", + "confidence": 0.82, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 45569 is the better canonical PR because it is the broader, more complete implementation of the Nemotron H model split (dense/sparse wrappers, docs, auto mappings, tests), while PR 45115 is a narrower refactor focused on inheritance and conversion mapping.", + "best_issue_reason": null, + "best_pr_reason": "PR 45569 is the strongest representative of the cluster: it contains the main user-facing model restructuring and supporting documentation/tests. PR 45115 is a separate, narrower code-path refactor and not a duplicate.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45115", + "right": "pull_request:45569", + "accept": false, + "reason": "They overlap on Nemotron H base files, but the concrete changes differ: 45115 is a refactor/conversion-mapping and inheritance update, while 45569 adds dense/sparse model variants, docs, auto mappings, and tests. Not the same underlying fix and not a single PR candidate." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 20, + "component_count": 27, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-45332-2", + "nodes": [ + "pull_request:45332", + "pull_request:45333" + ], + "soft_pairs": [ + "pull_request:45332|pull_request:45333" + ], + "prepared_review_unit_hash": "71fc5528e0b8f779659676eed86bec33dfa58af2ac4ef81de8bf0c7448f04f9a", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 5398, + "estimated_input_tokens": 1350, + "estimated_eval_tokens": 2956 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 5398, + "estimated_input_tokens": 1350, + "estimated_eval_tokens": 2956 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Two related but distinct feature PRs: one introduces heterogeneous per-layer config plumbing, the other adds broader heterogeneous model/runtime support on top. They share infrastructure files but are not the same concrete change, so they should not be merged as duplicates.", + "confidence": 0.92, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR #45332 is the broader umbrella change and the better representative of the cluster: it extends the feature into cache, masking, and model code paths, whereas #45333 is limited to per-layer config plumbing.", + "best_issue_reason": null, + "best_pr_reason": "PR #45332 is the strongest single PR to represent this cluster because it covers the larger end-to-end heterogeneous model support effort and contains the most complete implementation surface.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45332", + "right": "pull_request:45333", + "accept": false, + "reason": "Related feature work, but not the same underlying change. #45333 adds per-layer config support; #45332 adds model/runtime support that builds on heterogeneous config. They could land in sequence, not as one duplicate PR." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 21, + "component_count": 27, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-45415-2", + "nodes": [ + "pull_request:45415", + "pull_request:45425" + ], + "soft_pairs": [ + "pull_request:45415|pull_request:45425" + ], + "prepared_review_unit_hash": "b22e43170fd6bbee7deecd104c7ab5a6d10424e633b1442eb78c052cb1776955", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 5513, + "estimated_input_tokens": 1379, + "estimated_eval_tokens": 3014 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 5513, + "estimated_input_tokens": 1379, + "estimated_eval_tokens": 3014 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Both PRs are typing-focused, but they are different changes: #45415 is a broad type-checking sweep across many modules plus pyproject config, while #45425 adds specific typing helpers/protocols for modeling_utils/PEFT-related code. The overlap is incidental, not the same underlying fix.", + "confidence": 0.91, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR #45425 is the more self-contained, focused typing change; PR #45415 is a much broader repo-wide type-checking sweep.", + "best_issue_reason": null, + "best_pr_reason": "PR #45425 is the best representative of the cluster because it makes a concrete typing-infrastructure change in a narrower code path, whereas #45415 is a large mixed typing cleanup.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45415", + "right": "pull_request:45425", + "accept": false, + "reason": "Shared files are incidental; #45415 is a broad type-checking sweep, while #45425 adds specific typing protocols/helpers for different code paths. They are not the same concrete change and would not plausibly merge into one PR." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 22, + "component_count": 27, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-45487-4", + "nodes": [ + "pull_request:45487", + "pull_request:46109", + "pull_request:46135", + "pull_request:46151" + ], + "soft_pairs": [ + "pull_request:45487|pull_request:46151", + "pull_request:46109|pull_request:46151", + "pull_request:46135|pull_request:46151" + ], + "prepared_review_unit_hash": "00963de295bad7b69ebeffc1381f45996e149fc4a321a2c49074ac12e1f8d83d", + "component_budget": { + "node_count": 4, + "item_count": 4, + "soft_pair_count": 3, + "serialized_chars": 9913, + "estimated_input_tokens": 2479, + "estimated_eval_tokens": 5214 + }, + "budget": { + "node_count": 4, + "item_count": 4, + "soft_pair_count": 3, + "serialized_chars": 9913, + "estimated_input_tokens": 2479, + "estimated_eval_tokens": 5214 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These PRs are not duplicates of one another: they cover different code paths and different kinds of changes (AltCLIP/ChineseCLIP model-parallel bugfixes, distributed trainer test parity tweaks, Kosmos2.5 integration-test expectations, and pipeline token-classification expectations).", + "confidence": 0.96, + "canonical_issue_reason": null, + "canonical_pr_reason": "No single canonical PR: the cluster mixes unrelated fixes, so there is no one artifact that subsumes the others.", + "best_issue_reason": null, + "best_pr_reason": "PR 45487 is the strongest standalone change because it fixes a concrete runtime bug in model code; the others are narrower test-expected-output adjustments in unrelated areas.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45487", + "right": "pull_request:46151", + "accept": false, + "reason": "Different features and code paths: AltCLIP/ChineseCLIP model-parallel device handling vs token-classification pipeline test expectations." + }, + { + "left": "pull_request:46109", + "right": "pull_request:46151", + "accept": false, + "reason": "Both are test-only, but they target different subsystems and issues: distributed trainer device-string parity vs pipeline token-classification expectations." + }, + { + "left": "pull_request:46135", + "right": "pull_request:46151", + "accept": false, + "reason": "Kosmos2.5 integration-test expectations are unrelated to token-classification pipeline expectations; not the same underlying change." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 23, + "component_count": 27, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-45513-3", + "nodes": [ + "pull_request:45513", + "pull_request:46084", + "pull_request:46098" + ], + "soft_pairs": [ + "pull_request:45513|pull_request:46084", + "pull_request:46084|pull_request:46098" + ], + "prepared_review_unit_hash": "5a4046297a4240a7284b1773aaf360039955d995c9829af3ec0b8cc726524609", + "component_budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 2, + "serialized_chars": 7426, + "estimated_input_tokens": 1857, + "estimated_eval_tokens": 3970 + }, + "budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 2, + "serialized_chars": 7426, + "estimated_input_tokens": 1857, + "estimated_eval_tokens": 3970 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Two PRs are about the same Mamba2 `use_cache`/`seq_len > 1` bug; the Qwen3.5 PR is a different model and code path. PR #46098 is the best representative because it is the most complete Mamba2 fix and explicitly targets the tracked issue.", + "confidence": 0.92, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR #46098 is the cleanest representative of the duplicated Mamba2 bug fix: it explicitly targets issue #46032, covers the seq_len>1 cached path comprehensively, and includes tests/docs. PR #46084 is a narrower variant of the same fix, while #45513 is unrelated.", + "best_issue_reason": null, + "best_pr_reason": "PR #46098 is the best global PR candidate because it directly addresses the tracked Mamba2 cache bug with the broadest fix surface and validation, making it the most complete canonical change among the set.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45513", + "right": "pull_request:46084", + "accept": false, + "reason": "Different models and bugs: Qwen3.5 linear-attention cached forward vs. Mamba2 `use_cache` handling. Similar symptom, but not the same concrete code-path change and not mergeable into one PR." + }, + { + "left": "pull_request:46084", + "right": "pull_request:46098", + "accept": true, + "reason": "Same underlying Mamba2 `use_cache` with `seq_len > 1` bug on the same code path, with #46098 essentially extending/subsuming the narrower fix in #46084." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet: 45513 is clearly a different Qwen3.5-related PR, while 46084 and 46098 share the same Mamba2 issue target and files. The soft-edge verdicts are conservative, and treating 46084 and 46098 as mergeable for the same bugfix is supported by the evidence." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 24, + "component_count": 27, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-45548-2", + "nodes": [ + "pull_request:45548", + "pull_request:45662" + ], + "soft_pairs": [ + "pull_request:45548|pull_request:45662" + ], + "prepared_review_unit_hash": "9555e949d36abd8cbf394499ae8b8d9b20a049de2a1d36b6be7fe651e85be000", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4508, + "estimated_input_tokens": 1127, + "estimated_eval_tokens": 2510 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4508, + "estimated_input_tokens": 1127, + "estimated_eval_tokens": 2510 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Two open PRs touch expert-parallel loading, but they fix different bugs: one is about preserving distributed_config / EP state when loading via accelerate+ZeRO-3, the other is about preventing expert weights from being overwritten by rank-0 broadcast under FSDP2.", + "confidence": 0.91, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 45662 is the better cluster representative because it addresses a concrete EP/FSDP2 overwrite path with broader code impact; PR 45548 is a separate, narrower accelerate/ZeRO-3 configuration-loading fix.", + "best_issue_reason": null, + "best_pr_reason": "PR 45662 is the strongest standalone PR in this set: it targets the actual expert-parameter overwrite failure mode across multiple loading modules, while 45548 is a smaller serialization/loading adjustment.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45548", + "right": "pull_request:45662", + "accept": false, + "reason": "Both are about expert parallelism, but the underlying bugs differ: 45548 fixes EP state/config handling during accelerate+ZeRO-3 loading, while 45662 fixes experts being overwritten by rank-0 broadcast in FSDP2. They would not be merged into one PR." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 25, + "component_count": 27, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-45639-2", + "nodes": [ + "pull_request:45639", + "pull_request:45645" + ], + "soft_pairs": [ + "pull_request:45639|pull_request:45645" + ], + "prepared_review_unit_hash": "cc6329d876355e0560bfba5d39a6ed1d4b8816e5eaac13c5a613de81142a9761", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4589, + "estimated_input_tokens": 1148, + "estimated_eval_tokens": 2552 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4589, + "estimated_input_tokens": 1148, + "estimated_eval_tokens": 2552 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Both pull requests address the same xdist-related captured_info artifact collision problem; PR 45645 is the broader, more complete version that also updates CI log collection. PR 45639 looks like a narrower subset of the same fix.", + "confidence": 0.88, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 45645 is the better canonical candidate because it includes the core xdist-safe output-file change plus the CI workflow adjustment to collect multiple captured_info files, making it the more complete fix.", + "best_issue_reason": null, + "best_pr_reason": "PR 45645 is the best overall PR because it addresses the underlying collision bug and also preserves debug-log visibility in CI, whereas PR 45639 only covers the output-file side of the change.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45639", + "right": "pull_request:45645", + "accept": true, + "reason": "Both PRs target the same concrete bug: xdist collisions for captured_info artifacts in testing_utils. PR 45645 appears to supersede the narrower implementation in 45639 with the same core fix plus CI handling." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 26, + "component_count": 27, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "cache", + "cache_hit": true, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-45737-2", + "nodes": [ + "pull_request:45737", + "pull_request:45738" + ], + "soft_pairs": [ + "pull_request:45737|pull_request:45738" + ], + "prepared_review_unit_hash": "cc2ade73a49ec6a8883a7a24c433988645588bcd049e8aad2bce56b1b24daa1c", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4131, + "estimated_input_tokens": 1033, + "estimated_eval_tokens": 2322 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4131, + "estimated_input_tokens": 1033, + "estimated_eval_tokens": 2322 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Both PRs address the same MusicgenMelody cache initialization bug in `forward()`: replacing `EncoderDecoderCache(DynamicCache, DynamicCache)` with a single `DynamicCache`. PR #45737 is the more complete duplicate because it bundles the fix with a regression test, while #45738 is the fix-only variant.", + "confidence": 0.97, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR #45737 is the best canonical PR because it contains the same code fix as #45738 plus a regression test, making it the most complete representation of the bug fix.", + "best_issue_reason": null, + "best_pr_reason": "PR #45737 is the strongest overall candidate: it fixes the concrete cache bug and adds a targeted regression test, whereas #45738 only changes the implementation.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45737", + "right": "pull_request:45738", + "accept": true, + "reason": "Same underlying code-path bug in `musicgen_melody` cache setup; the diffs match on the core fix and the test-only addition in #45737 could plausibly live in the same PR as #45738." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The assessment is grounded in the packet: both PRs make the same `musicgen_melody` cache initialization change, and #45737 only adds a regression test on top of that fix. The soft-edge accept is conservative enough because the diffs are mergeable into one PR for the same bugfix." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 27, + "component_count": 27, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-46069-4", + "nodes": [ + "pull_request:46069", + "pull_request:46073", + "pull_request:46154", + "pull_request:46196" + ], + "soft_pairs": [ + "pull_request:46069|pull_request:46073", + "pull_request:46069|pull_request:46154", + "pull_request:46069|pull_request:46196", + "pull_request:46073|pull_request:46154", + "pull_request:46073|pull_request:46196", + "pull_request:46154|pull_request:46196" + ], + "prepared_review_unit_hash": "c749d9042e9101fc8ed99af5adbaebb8183df06ea9e3584be5091188b08b45b5", + "component_budget": { + "node_count": 4, + "item_count": 4, + "soft_pair_count": 6, + "serialized_chars": 10494, + "estimated_input_tokens": 2624, + "estimated_eval_tokens": 5504 + }, + "budget": { + "node_count": 4, + "item_count": 4, + "soft_pair_count": 6, + "serialized_chars": 10494, + "estimated_input_tokens": 2624, + "estimated_eval_tokens": 5504 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These PRs all add SDPA/attention-dispatch support, but each targets a different model family and code path, so they are related feature ports rather than duplicates.", + "confidence": 0.98, + "canonical_issue_reason": null, + "canonical_pr_reason": "No single canonical PR: each change is model-specific (ShieldGemma2, CTRL, T5, SeamlessM4T/v2) and would not be merged into one concrete fix.", + "best_issue_reason": null, + "best_pr_reason": "PR 46196 is the broadest representative because it covers both SeamlessM4T and SeamlessM4T v2, but it is still not a duplicate of the other model-specific dispatch PRs.", + "soft_edge_verdicts": [ + { + "left": "pull_request:46069", + "right": "pull_request:46073", + "accept": false, + "reason": "Same general SDPA theme, but ShieldGemma2 and CTRL are different model implementations with separate dispatch paths." + }, + { + "left": "pull_request:46069", + "right": "pull_request:46154", + "accept": false, + "reason": "Both add attention-dispatch support, but ShieldGemma2 and T5 are unrelated code paths and not one concrete bug fix." + }, + { + "left": "pull_request:46069", + "right": "pull_request:46196", + "accept": false, + "reason": "ShieldGemma2 image-classification attention dispatch is distinct from SeamlessM4T/v2 seq2seq attention dispatch." + }, + { + "left": "pull_request:46073", + "right": "pull_request:46154", + "accept": false, + "reason": "CTRL and T5 are separate model families; this is the same feature class, not the same underlying PR change." + }, + { + "left": "pull_request:46073", + "right": "pull_request:46196", + "accept": false, + "reason": "CTRL SDPA dispatch and SeamlessM4T/v2 SDPA dispatch touch different architectures and cannot plausibly be one merged fix." + }, + { + "left": "pull_request:46154", + "right": "pull_request:46196", + "accept": false, + "reason": "T5 and SeamlessM4T/v2 both add SDPA support, but they are independent model-specific implementations, not duplicates." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + } + ] +} diff --git a/snapshots/20260525T120039Z/analysis-runs/hybrid-model-20260525t120039z/manifest.json b/snapshots/20260525T120039Z/analysis-runs/hybrid-model-20260525t120039z/manifest.json new file mode 100644 index 0000000000000000000000000000000000000000..a58372a18cc78307f9bd590349495c3706242461 --- /dev/null +++ b/snapshots/20260525T120039Z/analysis-runs/hybrid-model-20260525t120039z/manifest.json @@ -0,0 +1,14 @@ +{ + "analysis_id": "hybrid-model-20260525t120039z", + "artifacts": { + "hybrid": "snapshots/20260525T120039Z/analysis-runs/hybrid-model-20260525t120039z/analysis-report-hybrid.json", + "hybrid_reviews": "snapshots/20260525T120039Z/analysis-runs/hybrid-model-20260525t120039z/analysis-report-hybrid.llm-reviews.json" + }, + "channel": "canonical", + "model": null, + "published_at": "2026-05-25T12:15:45Z", + "repo": "huggingface/transformers", + "schema_version": 1, + "snapshot_id": "20260525T120039Z", + "variant": "hybrid" +} diff --git a/snapshots/20260525T120039Z/manifest.json b/snapshots/20260525T120039Z/manifest.json index 6cc985d6c44e621934938d98140a38c6969546cb..3023bd3b7163350ac5809767b34935ad859f71e1 100644 --- a/snapshots/20260525T120039Z/manifest.json +++ b/snapshots/20260525T120039Z/manifest.json @@ -32,6 +32,24 @@ "timeline_events": 643 }, "extracted_at": "2026-05-25T12:00:39Z", + "published_analysis": { + "canonical_analysis_id": "hybrid-model-20260525t120039z", + "runs": { + "hybrid-model-20260525t120039z": { + "analysis_id": "hybrid-model-20260525t120039z", + "artifacts": { + "hybrid": "snapshots/20260525T120039Z/analysis-runs/hybrid-model-20260525t120039z/analysis-report-hybrid.json", + "hybrid_reviews": "snapshots/20260525T120039Z/analysis-runs/hybrid-model-20260525t120039z/analysis-report-hybrid.llm-reviews.json" + }, + "channel": "canonical", + "manifest_path": "snapshots/20260525T120039Z/analysis-runs/hybrid-model-20260525t120039z/manifest.json", + "model": null, + "published_at": "2026-05-25T12:15:45Z", + "variant": "hybrid" + } + }, + "schema_version": 1 + }, "repo": "huggingface/transformers", "snapshot_id": "20260525T120039Z", "watermark": {