diff --git a/analysis/current/analysis-report-hybrid.json b/analysis/current/analysis-report-hybrid.json index ccfbbf88000dcf85c9be65131c7cd5164e597ef2..f6f4746ec4b8f75a8b2f93fc44623b93280b517a 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": "20260702T000029Z", - "generated_at": "2026-07-02T00:13:10Z", + "snapshot_id": "20260702T060019Z", + "generated_at": "2026-07-02T06:11:19Z", "evidence_quality": "full", "llm_enrichment": true, "meta_bugs": [ @@ -1149,14 +1149,228 @@ ] }, { - "cluster_id": "cluster-45672-5", - "summary": "Cluster of 3 issues and 2 PRs centered on issue #46932.", + "cluster_id": "cluster-46894-6", + "summary": "Cluster of 3 issues and 3 PRs centered on issue #46894.", + "status": "open", + "confidence": 0.6, + "canonical_issue_number": 46894, + "canonical_pr_number": 46903, + "issue_numbers": [ + 46894, + 46897, + 46901 + ], + "pr_numbers": [ + 46895, + 46903, + 46958 + ], + "evidence_types": [ + "closing_reference", + "soft_similarity" + ], + "pr_comparisons": [ + { + "left_pr_number": 46895, + "right_pr_number": 46903, + "code_similarity": 0.362, + "size_similarity": 0.443, + "file_overlap": 0.333, + "area_overlap": 0.419, + "patch_similarity": 0.178, + "shared_filenames": [ + "src/transformers/models/cohere_asr/modeling_cohere_asr.py", + "src/transformers/models/cohere_asr/modular_cohere_asr.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/models/cohere_asr/modeling_cohere_asr.py", + "left_ranges": [ + [ + 615, + 627 + ], + [ + 639, + 654 + ] + ], + "right_ranges": [ + [ + 637, + 651 + ] + ] + }, + { + "filename": "src/transformers/models/cohere_asr/modular_cohere_asr.py", + "left_ranges": [ + [ + 478, + 490 + ], + [ + 502, + 517 + ] + ], + "right_ranges": [ + [ + 500, + 514 + ] + ] + } + ] + }, + { + "left_pr_number": 46895, + "right_pr_number": 46958, + "code_similarity": 0.715, + "size_similarity": 0.833, + "file_overlap": 1.0, + "area_overlap": 0.366, + "patch_similarity": 0.802, + "shared_filenames": [ + "src/transformers/models/cohere_asr/modeling_cohere_asr.py", + "src/transformers/models/cohere_asr/modular_cohere_asr.py", + "tests/models/cohere_asr/test_modeling_cohere_asr.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/models/cohere_asr/modeling_cohere_asr.py", + "left_ranges": [ + [ + 615, + 627 + ], + [ + 639, + 654 + ] + ], + "right_ranges": [ + [ + 22, + 28 + ], + [ + 638, + 645 + ] + ] + }, + { + "filename": "src/transformers/models/cohere_asr/modular_cohere_asr.py", + "left_ranges": [ + [ + 478, + 490 + ], + [ + 502, + 517 + ] + ], + "right_ranges": [ + [ + 16, + 22 + ], + [ + 501, + 508 + ] + ] + }, + { + "filename": "tests/models/cohere_asr/test_modeling_cohere_asr.py", + "left_ranges": [ + [ + 35, + 41 + ], + [ + 176, + 214 + ] + ], + "right_ranges": [ + [ + 15, + 21 + ], + [ + 180, + 221 + ] + ] + } + ] + }, + { + "left_pr_number": 46903, + "right_pr_number": 46958, + "code_similarity": 0.325, + "size_similarity": 0.369, + "file_overlap": 0.333, + "area_overlap": 0.364, + "patch_similarity": 0.156, + "shared_filenames": [ + "src/transformers/models/cohere_asr/modeling_cohere_asr.py", + "src/transformers/models/cohere_asr/modular_cohere_asr.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/models/cohere_asr/modeling_cohere_asr.py", + "left_ranges": [ + [ + 637, + 651 + ] + ], + "right_ranges": [ + [ + 22, + 28 + ], + [ + 638, + 645 + ] + ] + }, + { + "filename": "src/transformers/models/cohere_asr/modular_cohere_asr.py", + "left_ranges": [ + [ + 500, + 514 + ] + ], + "right_ranges": [ + [ + 16, + 22 + ], + [ + 501, + 508 + ] + ] + } + ] + } + ] + }, + { + "cluster_id": "cluster-46899-4", + "summary": "Cluster of 2 issues and 2 PRs centered on issue #46932.", "status": "open", "confidence": 0.8, "canonical_issue_number": 46932, "canonical_pr_number": 46933, "issue_numbers": [ - 45672, 46899, 46932 ], @@ -1333,112 +1547,6 @@ } ] }, - { - "cluster_id": "cluster-46894-3", - "summary": "Cluster of 1 issues and 2 PRs centered on issue #46894.", - "status": "open", - "confidence": 0.75, - "canonical_issue_number": 46894, - "canonical_pr_number": 46895, - "issue_numbers": [ - 46894 - ], - "pr_numbers": [ - 46895, - 46958 - ], - "evidence_types": [ - "closing_reference", - "shared_issue_target" - ], - "pr_comparisons": [ - { - "left_pr_number": 46895, - "right_pr_number": 46958, - "code_similarity": 0.715, - "size_similarity": 0.833, - "file_overlap": 1.0, - "area_overlap": 0.366, - "patch_similarity": 0.802, - "shared_filenames": [ - "src/transformers/models/cohere_asr/modeling_cohere_asr.py", - "src/transformers/models/cohere_asr/modular_cohere_asr.py", - "tests/models/cohere_asr/test_modeling_cohere_asr.py" - ], - "shared_file_areas": [ - { - "filename": "src/transformers/models/cohere_asr/modeling_cohere_asr.py", - "left_ranges": [ - [ - 615, - 627 - ], - [ - 639, - 654 - ] - ], - "right_ranges": [ - [ - 22, - 28 - ], - [ - 638, - 645 - ] - ] - }, - { - "filename": "src/transformers/models/cohere_asr/modular_cohere_asr.py", - "left_ranges": [ - [ - 478, - 490 - ], - [ - 502, - 517 - ] - ], - "right_ranges": [ - [ - 16, - 22 - ], - [ - 501, - 508 - ] - ] - }, - { - "filename": "tests/models/cohere_asr/test_modeling_cohere_asr.py", - "left_ranges": [ - [ - 35, - 41 - ], - [ - 176, - 214 - ] - ], - "right_ranges": [ - [ - 15, - 21 - ], - [ - 180, - 221 - ] - ] - } - ] - } - ] - }, { "cluster_id": "cluster-46869-3", "summary": "Cluster of 1 issues and 2 PRs centered on issue #46869.", @@ -2035,7 +2143,7 @@ "cluster_id": "cluster-46305-3", "summary": "Cluster of 1 issues and 2 PRs centered on issue #46305.", "status": "open", - "confidence": 0.55, + "confidence": 0.8, "canonical_issue_number": 46305, "canonical_pr_number": 46310, "issue_numbers": [ @@ -2046,7 +2154,9 @@ 47010 ], "evidence_types": [ - "closing_reference" + "closing_reference", + "shared_issue_target", + "soft_similarity" ], "pr_comparisons": [ { @@ -2174,62 +2284,33 @@ { "left_pr_number": 46970, "right_pr_number": 46986, - "code_similarity": 0.154, - "size_similarity": 0.727, + "code_similarity": 0.01, + "size_similarity": 0.029, "file_overlap": 0.0, "area_overlap": 0.0, - "patch_similarity": 0.057, + "patch_similarity": 0.026, "shared_filenames": [], "shared_file_areas": [] }, { "left_pr_number": 46970, "right_pr_number": 47006, - "code_similarity": 0.105, - "size_similarity": 0.5, + "code_similarity": 0.008, + "size_similarity": 0.02, "file_overlap": 0.0, "area_overlap": 0.0, - "patch_similarity": 0.036, + "patch_similarity": 0.026, "shared_filenames": [], "shared_file_areas": [] }, { "left_pr_number": 46986, "right_pr_number": 47006, - "code_similarity": 0.073, - "size_similarity": 0.364, - "file_overlap": 0.0, - "area_overlap": 0.0, - "patch_similarity": 0.0, - "shared_filenames": [], - "shared_file_areas": [] - } - ] - }, - { - "cluster_id": "cluster-46836-2", - "summary": "Cluster of 2 related pull requests linked by soft_similarity.", - "status": "open", - "confidence": 0.5, - "canonical_issue_number": null, - "canonical_pr_number": 46836, - "issue_numbers": [], - "pr_numbers": [ - 46836, - 46912 - ], - "evidence_types": [ - "soft_similarity" - ], - "pr_comparisons": [ - { - "left_pr_number": 46836, - "right_pr_number": 46912, - "code_similarity": 0.129, - "size_similarity": 0.55, + "code_similarity": 0.144, + "size_similarity": 0.697, "file_overlap": 0.0, "area_overlap": 0.0, - "patch_similarity": 0.129, + "patch_similarity": 0.028, "shared_filenames": [], "shared_file_areas": [] } @@ -2292,292 +2373,25 @@ "shared_file_areas": [] } ] - }, - { - "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": [ { - "cluster_id": "cluster-45672-5", + "cluster_id": "cluster-46894-6", + "canonical_issue_number": 46894, + "duplicate_issue_numbers": [ + 46897, + 46901 + ], + "reason": "Issues in cluster-46894-6 are treated as duplicates because they share closing_reference, soft_similarity evidence." + }, + { + "cluster_id": "cluster-46899-4", "canonical_issue_number": 46932, "duplicate_issue_numbers": [ - 45672, 46899 ], - "reason": "Issues in cluster-45672-5 are treated as duplicates because they share closing_reference, shared_issue_target, soft_similarity evidence." + "reason": "Issues in cluster-46899-4 are treated as duplicates because they share closing_reference, shared_issue_target, soft_similarity evidence." } ], "duplicate_prs": [ @@ -2625,15 +2439,6 @@ "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, @@ -2661,15 +2466,6 @@ "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-45672-5", - "canonical_pr_number": 46933, - "duplicate_pr_numbers": [ - 47008 - ], - "target_issue_number": 46932, - "reason": "PRs in cluster-45672-5 are treated as duplicates because they converge on issue #46932 with closing_reference, shared_issue_target, soft_similarity evidence." - }, { "cluster_id": "cluster-46159-3", "canonical_pr_number": 46174, @@ -2686,16 +2482,7 @@ 47010 ], "target_issue_number": 46305, - "reason": "PRs in cluster-46305-3 are treated as duplicates because they converge on issue #46305 with closing_reference evidence." - }, - { - "cluster_id": "cluster-46836-2", - "canonical_pr_number": 46836, - "duplicate_pr_numbers": [ - 46912 - ], - "target_issue_number": null, - "reason": "PRs in cluster-46836-2 are treated as duplicates because they share soft_similarity evidence." + "reason": "PRs in cluster-46305-3 are treated as duplicates because they converge on issue #46305 with closing_reference, shared_issue_target, soft_similarity evidence." }, { "cluster_id": "cluster-46846-3", @@ -2716,13 +2503,23 @@ "reason": "PRs in cluster-46869-3 are treated as duplicates because they converge on issue #46869 with closing_reference, shared_issue_target evidence." }, { - "cluster_id": "cluster-46894-3", - "canonical_pr_number": 46895, + "cluster_id": "cluster-46894-6", + "canonical_pr_number": 46903, "duplicate_pr_numbers": [ + 46895, 46958 ], "target_issue_number": 46894, - "reason": "PRs in cluster-46894-3 are treated as duplicates because they converge on issue #46894 with closing_reference, shared_issue_target evidence." + "reason": "PRs in cluster-46894-6 are treated as duplicates because they converge on issue #46894 with closing_reference, soft_similarity evidence." + }, + { + "cluster_id": "cluster-46899-4", + "canonical_pr_number": 46933, + "duplicate_pr_numbers": [ + 47008 + ], + "target_issue_number": 46932, + "reason": "PRs in cluster-46899-4 are treated as duplicates because they converge on issue #46932 with closing_reference, shared_issue_target, soft_similarity evidence." }, { "cluster_id": "cluster-46908-2", @@ -2754,10 +2551,10 @@ } ], "best_issue": { - "cluster_id": "cluster-45672-5", - "issue_number": 46932, - "reason": "Issue #46932 is the strongest global issue candidate because it is open, belongs to a cluster with 5 artifacts, and carries 1 discussion comments plus 4 inbound references.", - "score": 156.0 + "cluster_id": "cluster-46894-6", + "issue_number": 46894, + "reason": "Issue #46894 is the strongest global issue candidate because it is open, belongs to a cluster with 6 artifacts, and carries 0 discussion comments plus 6 inbound references.", + "score": 166.0 }, "best_pr": { "cluster_id": "cluster-43979-11", diff --git a/analysis/current/analysis-report-hybrid.llm-reviews.json b/analysis/current/analysis-report-hybrid.llm-reviews.json index 2780e77dd86b2e997c5079b81ebc931f718804fd..326f67316993bb422f2e8d1fbd66a810869e47cd 100644 --- a/analysis/current/analysis-report-hybrid.llm-reviews.json +++ b/analysis/current/analysis-report-hybrid.llm-reviews.json @@ -1,14 +1,14 @@ { "schema_version": "1.0", "repo": "huggingface/transformers", - "snapshot_id": "20260702T000029Z", - "generated_at": "2026-07-02T00:13:10Z", + "snapshot_id": "20260702T060019Z", + "generated_at": "2026-07-02T06:11:19Z", "model": "gpt-5.4-mini?service_tier=flex", "reviews": [ { "label": "issue", "component_index": 1, - "component_count": 25, + "component_count": 26, "review_unit_index": 1, "review_unit_count": 61, "status": "reviewed", @@ -120,8 +120,8 @@ "issue:44292|issue:44918", "issue:45507|issue:45685", "issue:43716|issue:44292", - "issue:44918|issue:45161", "issue:43716|issue:45685", + "issue:44918|issue:45161", "issue:43039|issue:44918", "issue:41762|issue:43638", "issue:43827|issue:44512", @@ -132,14 +132,14 @@ "issue:45237|issue:46015", "issue:43039|issue:44805" ], - "prepared_review_unit_hash": "87a2fa49e705e8f1d149987cb7b82e9c3fd0573dd9e32326ff8031569298a4d3", + "prepared_review_unit_hash": "064d6365c1d30a398675df748be644a53c9e808dd7e34c410a3bc80383480118", "component_budget": { "node_count": 505, "item_count": 505, "soft_pair_count": 1854, - "serialized_chars": 962354, - "estimated_input_tokens": 240589, - "estimated_eval_tokens": 481434 + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 }, "budget": { "node_count": 48, @@ -154,390 +154,390 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly heterogeneous: it mixes unrelated model-specific runtime bugs, docs regressions, and version-compatibility reports. The only clear duplicate pair is the DeepSpeed ZeRO-3/BERT size-0 IndexError reports.", + "summary": "Only one clear duplicate pair appears in this cluster: the DeepSpeed ZeRO-3 BertModel IndexError reports. The rest are mostly loosely related model/loading/device-mismatch/doc issues and should stay separate.", "confidence": 0.84, - "canonical_issue_reason": "issue:43596 is the best representative because it most clearly states the shared ZeRO-3 load failure; issue:43638 is the same error with a narrower BertModel variant.", + "canonical_issue_reason": "issue:43638 is the best canonical issue because it captures the same ZeRO-3 Bert IndexError as issue:43596, but with a more specific repro and more discussion.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43596 is the strongest canonical issue candidate in the cluster: it is concrete, well-scoped, and matches the only truly duplicate issue pair.", + "best_issue_reason": "issue:43638 is the strongest representative of the only real duplicate set in the cluster and is the most specific of the two Bert ZeRO-3 reports.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44322", "right": "issue:45464", "accept": false, - "reason": "Different bugs: Qwen3.5 config attribute error vs chat/completions streaming failure." + "reason": "Different failures: missing Qwen3.5 config fields versus streaming chat/completions API behavior." }, { "left": "issue:45237", "right": "issue:45685", "accept": false, - "reason": "Different model/backend issues: AMD GPU failure vs MPS histogram op not implemented." + "reason": "Both mention runtime/device problems, but one is AMD GPU support for GPT-OSS and the other is an MPS histogram error in MoE code; different bugs." }, { "left": "issue:43716", "right": "issue:45237", "accept": false, - "reason": "Different root causes and model families: dtype mismatch in preprocessing vs GPT-OSS GPU failure." + "reason": "Different code paths: image preprocessor dtype mismatch versus GPT-OSS failing on AMD GPUs." }, { "left": "issue:43828", "right": "issue:45237", "accept": false, - "reason": "Different failures: Phi autocast dtype mismatch vs GPT-OSS on AMD GPUs." + "reason": "Both are dtype-related, but one is an autocast mismatch and the other is an AMD GPU compatibility issue; not the same bug." }, { "left": "issue:45237", "right": "issue:46468", "accept": false, - "reason": "Official Nemotron example broken is a separate issue, not the same runtime bug." + "reason": "Different models and symptoms: GPT-OSS on AMD GPUs versus a broken official Nemotron example." }, { "left": "issue:45237", "right": "issue:45507", "accept": false, - "reason": "Different model/code paths: GPT-OSS AMD GPU failure vs GraniteMoEHybrid invalid method." + "reason": "Different failures in different model families; no shared concrete bug." }, { "left": "issue:30277", "right": "issue:35545", "accept": false, - "reason": "Jamba DeepSpeed collective mismatch is unrelated to ModernBERT ONNX export failure." + "reason": "DeepSpeed ZeRO-3 collective mismatch is unrelated to ModernBERT ONNX export." }, { "left": "issue:35545", "right": "issue:42915", "accept": false, - "reason": "ModernBERT ONNX export and Qwen3Moe FP8 failure are different bugs." + "reason": "ONNX export error and FineGrainedFP8 failure are unrelated model/runtime issues." }, { "left": "issue:35545", "right": "issue:41720", "accept": false, - "reason": "ONNX export failure vs Qwen3 auto device-mapping cuda assert are not the same issue." + "reason": "Different problems: ONNX export versus Qwen3 auto device mapping cudaErrorAssert." }, { "left": "issue:45901", "right": "issue:46015", "accept": false, - "reason": "Both are docs-ish, but one is a table QA crash and the other is outdated RAG examples; not the same bug." + "reason": "A task crash is not the same as outdated RAG examples documentation." }, { "left": "issue:35545", "right": "issue:36010", "accept": false, - "reason": "ModernBERT export error is unrelated to the GenerationMixin import error." + "reason": "Different failures: ONNX export versus GenerationMixin import regression." }, { "left": "issue:44509", "right": "issue:44512", "accept": false, - "reason": "Both are docs cleanup reports, but they target different removed features and are too broad to merge as duplicates." + "reason": "Both are docs cleanups for v5 removals, but they target different deprecated content and are not the same change." }, { "left": "issue:44918", "right": "issue:46165", "accept": false, - "reason": "TRL SFT unpacking failure vs resume-training with trust_remote_code are different training paths and symptoms." + "reason": "TRL SFT embedding unpacking failure is unrelated to resuming training with a locally downloaded remote-code model." }, { "left": "issue:46165", "right": "issue:46869", "accept": false, - "reason": "Resume-training/trust_remote_code problem is unrelated to YOLO multi-GPU device_map_auto failure." + "reason": "Different scenarios and code paths: resume-training/remote-code versus YOLO multi-GPU device-map auto." }, { "left": "issue:44805", "right": "issue:45823", "accept": false, - "reason": "Mask/tensor shape IndexError vs Gemma4 PLE device mismatch are different bugs." + "reason": "Mask-shape IndexError is unrelated to Gemma4 device-map forward mismatch." }, { "left": "issue:43638", "right": "issue:44805", "accept": false, - "reason": "Same broad IndexError family, but BERT ZeRO-3 loading and mask-shape mismatch are different code paths." - }, - { - "left": "issue:43638", - "right": "issue:46165", - "accept": false, - "reason": "BERT ZeRO-3 index error is unrelated to the remote-code resume-training issue." + "reason": "Both are IndexErrors, but one is ZeRO-3 Bert initialization and the other is a mask/tensor shape mismatch; different bugs." }, { "left": "issue:43596", "right": "issue:44805", "accept": false, - "reason": "BERT ZeRO-3 zero-size tensor issue and mask-shape mismatch are not the same bug." + "reason": "Different failure modes and subsystems: ZeRO-3 Bert init versus tensor mask shape mismatch." }, { "left": "issue:44805", "right": "issue:46294", "accept": false, - "reason": "Mask-shape mismatch and Qwen3.5 video token/feature mismatch are different failures." + "reason": "Shape-mismatch IndexError is unrelated to Qwen3.5 video token/template mismatch." }, { "left": "issue:44805", "right": "issue:45161", "accept": false, - "reason": "Tensor mask IndexError vs GPT-OSS MoE TP failure are unrelated." + "reason": "Different models and bugs: mask/tensor shape mismatch versus GPT-OSS MoE tensor-parallel failure." }, { "left": "issue:46165", "right": "issue:46468", "accept": false, - "reason": "Resume-training/trust_remote_code issue is unrelated to the Nemotron example bug." + "reason": "Resume-training remote-code issue is unrelated to the broken Nemotron example." }, { "left": "issue:44560", "right": "issue:46165", "accept": false, - "reason": "Qwen3-vl video StopIteration and resume-training/trust_remote_code are different issues." + "reason": "Video StopIteration in chat template handling is unrelated to resuming training from a local model." }, { "left": "issue:34634", "right": "issue:35545", "accept": false, - "reason": "BarkProcessor voice_preset failure is unrelated to ModernBERT ONNX export." + "reason": "BarkProcessor voice_preset parsing and ModernBERT ONNX export are unrelated." }, { "left": "issue:30277", "right": "issue:43638", "accept": false, - "reason": "Jamba collective mismatch and BERT ZeRO-3 index error are different DeepSpeed failures." + "reason": "DeepSpeed collective mismatch on Jamba-v01 is not the same as the Bert ZeRO-3 IndexError." }, { "left": "issue:43596", "right": "issue:43638", "accept": true, - "reason": "Same ZeRO-3 loading failure: both report index 0 out of bounds on size-0 tensors during BERT model init, differing only by pretrained vs non-pretrained wording." + "reason": "These describe the same ZeRO-3 BertModel IndexError; the non-pretrained Bert repro is just a more specific instance." }, { "left": "issue:43039", "right": "issue:45161", "accept": false, - "reason": "Liger kernel cross_entropy call and GPT-OSS TP failure are unrelated code paths." + "reason": "Liger Kernel cross-entropy routing and GPT-OSS tensor-parallel failure are different issues." }, { "left": "issue:43975", "right": "issue:44457", "accept": false, - "reason": "DeepSeek detokenization bug is unrelated to the LoRA merge/save/reload mismatch report." + "reason": "Incorrect detokenization and LoRA merge/save inconsistency are unrelated bugs." }, { "left": "issue:44292", "right": "issue:44457", "accept": false, - "reason": "Qwen-3-8B-NVFP4 runtime failure is unrelated to LoRA weight merge/save differences." + "reason": "Qwen-3-8B-NVFP4 runtime error is unrelated to LoRA weight merge/reload mismatch." }, { "left": "issue:44841", "right": "issue:46661", "accept": false, - "reason": "Voxtral processor failure and remote-code config_class AttributeError are different problems." + "reason": "Processor failure for Voxtral and remote-code config_class AttributeError are different failures." }, { "left": "issue:43716", "right": "issue:46869", "accept": false, - "reason": "Mistral-3 dtype mismatch is unrelated to the YOLO multi-GPU device-map issue." + "reason": "Mistral dtype mismatch and YOLO multi-GPU device-map failure are unrelated." }, { "left": "issue:24643", "right": "issue:30277", "accept": false, - "reason": "'weight' must be 2-D with DeepSpeed and Jamba collective mismatch are different failures." + "reason": "Both involve DeepSpeed, but the concrete failures are different: 2-D weight error versus collective mismatch." }, { "left": "issue:34928", "right": "issue:44805", "accept": false, - "reason": "FSDP activation-checkpoint recompute mismatch is unrelated to the mask/tensor shape IndexError." + "reason": "Activation checkpointing/FSDP recomputation mismatch is a different bug from the mask-shape IndexError." }, { "left": "issue:45685", "right": "issue:46869", "accept": false, - "reason": "MPS histogram_mps issue in MoE is unrelated to YOLO multi-GPU device-map failure." + "reason": "MPS histogram not implemented for Int is unrelated to YOLO device-map auto failures." }, { "left": "issue:45859", "right": "issue:45902", "accept": false, - "reason": "Same model family, but CPU-offload incompatibility and uninitialized-memory-after-to_empty are not clearly the same bug." + "reason": "Same module family, but one is CPU offload incompatibility and the other is uninitialized memory after materialization; not clearly the same bug." }, { "left": "issue:43638", "right": "issue:45161", "accept": false, - "reason": "BERT ZeRO-3 index error and GPT-OSS TP failure are unrelated." + "reason": "ZeRO-3 Bert initialization and GPT-OSS tensor-parallel issues are unrelated." }, { "left": "issue:34928", "right": "issue:36331", "accept": false, - "reason": "Activation-checkpoint/FSDP recompute mismatch is unrelated to the CustomTrainer compute_loss signature change." + "reason": "FSDP activation-checkpointing tensor recomputation and CustomTrainer compute_loss signature mismatch are unrelated." }, { "left": "issue:43596", "right": "issue:45161", "accept": false, - "reason": "BERT ZeRO-3 index error and GPT-OSS TP failure are different code paths." + "reason": "Different model and stack: Bert ZeRO-3 init versus GPT-OSS TP failure." }, { "left": "issue:43716", "right": "issue:45507", "accept": false, - "reason": "Mistral-3 dtype mismatch and GraniteMoEHybrid invalid method are unrelated." + "reason": "Mistral dtype mismatch and GraniteMoEHybrid invalid method do not share the same underlying bug." }, { "left": "issue:45874", "right": "issue:46294", "accept": false, - "reason": "Gemma4 reverse-embedding lookup is unrelated to Qwen3.5 video feature/token mismatch." + "reason": "Gemma4 reverse-embedding lookup and Qwen3.5 video token mismatch are unrelated." }, { "left": "issue:44292", "right": "issue:45237", "accept": false, - "reason": "Qwen-3-8B-NVFP4 runtime failure is unrelated to GPT-OSS AMD GPU failure." + "reason": "Qwen-3-8B-NVFP4 error and GPT-OSS AMD GPU failure are different model-specific problems." }, { "left": "issue:39290", "right": "issue:45081", "accept": false, - "reason": "Gemma3Config sliding_window_pattern error and Mistral tokenizer regex crash are different bugs." + "reason": "Gemma3TextConfig missing attribute with vLLM is unrelated to the Mistral regex patch crash." }, { "left": "issue:45823", "right": "issue:45874", "accept": false, - "reason": "Both are Gemma4-related, but one is a device mismatch and the other is an expensive reverse-lookup bug." + "reason": "Both are Gemma4 issues, but one is device-mismatch in forward and the other is expensive reverse embedding lookup; different code paths." }, { "left": "issue:45800", "right": "issue:46515", "accept": false, - "reason": "Both mention torch 2.4.1, but one is tied to transformers 5.8.0 and the reports are not the same concrete bug." + "reason": "Both mention torch compatibility, but they are different reported incompatibilities and are not the same bug." }, { "left": "issue:34928", "right": "issue:41720", "accept": false, - "reason": "FSDP activation-checkpoint mismatch and Qwen3 auto device-map cuda assert are unrelated." + "reason": "FSDP checkpointing tensor mismatch and Qwen3 auto-device-map cudaErrorAssert are unrelated." }, { "left": "issue:43425", "right": "issue:44292", "accept": false, - "reason": "Torch 2.10 incompatibility and Qwen-3-8B-NVFP4 runtime failure are different issues." - }, - { - "left": "issue:43828", - "right": "issue:44512", - "accept": false, - "reason": "Phi autocast dtype mismatch and docs about the removed run command are unrelated." + "reason": "General Torch 2.10 incompatibility is not the same as the Qwen-3-8B-NVFP4 runtime error." }, { "left": "issue:43828", "right": "issue:45161", "accept": false, - "reason": "Phi autocast dtype mismatch and GPT-OSS TP failure are different problems." + "reason": "Autocast dtype mismatch and GPT-OSS tensor-parallel failure are distinct issues." }, { "left": "issue:43541", "right": "issue:43716", "accept": false, - "reason": "Mixtral torch-dynamo grouped_mm tracing error is unrelated to the Mistral-3 dtype mismatch." + "reason": "Mixtral torch dynamo grouped_mm tracing error is unrelated to Mistral image-preprocessor dtype mismatch." }, { "left": "issue:43828", "right": "issue:44292", "accept": false, - "reason": "Phi autocast dtype mismatch and Qwen-3-8B-NVFP4 runtime failure are unrelated." + "reason": "Different failures: autocast dtype mismatch versus Qwen-3-8B-NVFP4 runtime error." }, { "left": "issue:43716", "right": "issue:46468", "accept": false, - "reason": "Mistral-3 dtype mismatch is unrelated to the Nemotron example bug." + "reason": "Mistral dtype mismatch and broken Nemotron example are unrelated." }, { "left": "issue:44292", "right": "issue:44918", "accept": false, - "reason": "Qwen-3-8B-NVFP4 runtime failure and Qwen3.5 TRL unpacking failure are different training/runtime issues." + "reason": "Qwen-3-8B-NVFP4 runtime error is unrelated to TRL SFT embedding unpacking." }, { "left": "issue:45507", "right": "issue:45685", "accept": false, - "reason": "GraniteMoEHybrid invalid method and MPS histogram_mps not implemented are different backend/model bugs." + "reason": "Different model/runtime bugs: GraniteMoEHybrid invalid method versus MPS histogram limitation." }, { "left": "issue:43716", "right": "issue:44292", "accept": false, - "reason": "Mistral-3 dtype mismatch and Qwen-3-8B-NVFP4 runtime failure are unrelated." + "reason": "Different model families and failure modes." + }, + { + "left": "issue:43716", + "right": "issue:45685", + "accept": false, + "reason": "Image-preprocessor dtype mismatch is unrelated to an MPS integer histogram error." + }, + { + "left": "issue:44918", + "right": "issue:45161", + "accept": false, + "reason": "TRl embedding unpacking and GPT-OSS tensor-parallel failure are different issues." }, { "left": "issue:43039", "right": "issue:44918", "accept": false, - "reason": "Liger kernel cross_entropy issue is unrelated to Qwen3.5 TRL embedding unpacking." + "reason": "Liger Kernel cross-entropy routing is unrelated to TRL SFT embedding unpacking." }, { "left": "issue:41762", "right": "issue:43638", "accept": false, - "reason": "Same generic zero-sized-tensor IndexError pattern, but Gemma3 conditional generation and BertModel ZeRO-3 loading are too different to treat as one bug." + "reason": "Gemma3 ZeRO-3 loading IndexError is a different bug from the Bert ZeRO-3 IndexError." }, { "left": "issue:43827", "right": "issue:44512", "accept": false, - "reason": "Both are docs regressions, but one is pipeline-task references and the other is the run-command docs; separate cleanup tasks." + "reason": "Both are docs regressions, but they refer to different removed v5 features and should not be merged as the same change." }, { "left": "issue:43828", "right": "issue:45685", "accept": false, - "reason": "Phi autocast dtype mismatch and MPS histogram_mps issue are unrelated." + "reason": "Autocast dtype mismatch and MPS histogram not implemented are unrelated." }, { "left": "issue:43039", "right": "issue:43828", "accept": false, - "reason": "Liger kernel cross_entropy call and Phi autocast dtype mismatch are different issues." + "reason": "Liger Kernel cross-entropy being called and autocast dtype mismatch are different bugs." }, { "left": "issue:30277", "right": "issue:36331", "accept": false, - "reason": "Jamba DeepSpeed collective mismatch is unrelated to the CustomTrainer compute_loss signature change." + "reason": "DeepSpeed collective mismatch is unrelated to the CustomTrainer compute_loss signature change." }, { "left": "issue:34928", "right": "issue:36010", "accept": false, - "reason": "FSDP activation-checkpoint mismatch and GenerationMixin import error are unrelated." + "reason": "Activation-checkpointing/FSDP mismatch and GenerationMixin import failure are unrelated." }, { "left": "issue:45237", "right": "issue:46015", "accept": false, - "reason": "GPT-OSS AMD GPU failure and outdated RAG examples are unrelated." + "reason": "GPU/model runtime failure and outdated RAG docs are not the same underlying issue." }, { "left": "issue:43039", "right": "issue:44805", "accept": false, - "reason": "Liger kernel cross_entropy issue and tensor mask-shape IndexError are unrelated." + "reason": "Liger Kernel cross-entropy routing is unrelated to the tensor mask shape IndexError." } ] }, "evaluator_result": { "accept": true, - "feedback": "The analyst output is grounded in the packet and stays conservative on the soft edges. The canonical issue choice is supported by the clearly duplicated ZeRO-3/BERT pair (issue:43596 and issue:43638), and the other pair verdicts avoid overclaiming similarity. No mergeability concerns apply because there are no PR pairs in the packet." + "feedback": "The analysis is grounded in the packet and stays conservative on the soft edges. The only clearly supported duplicate pair is the ZeRO-3 Bert IndexError reports (43596/43638), and the rest are reasonably kept separate given the titles and available evidence." }, "evaluator_used": true, "retried": false, @@ -548,7 +548,7 @@ { "label": "issue", "component_index": 1, - "component_count": 25, + "component_count": 26, "review_unit_index": 2, "review_unit_count": 61, "status": "reviewed", @@ -609,8 +609,8 @@ ], "soft_pairs": [ "issue:45800|issue:46304", - "issue:43866|issue:46402", "issue:43296|issue:46304", + "issue:43866|issue:46402", "issue:41762|issue:43596", "issue:39290|issue:41720", "issue:43901|issue:44512", @@ -624,8 +624,8 @@ "issue:44863|issue:46015", "issue:45020|issue:46304", "issue:45812|issue:45920", - "issue:43866|issue:44863", "issue:44292|issue:46015", + "issue:43866|issue:44863", "issue:43872|issue:46304", "issue:39290|issue:43531", "issue:41093|issue:43638", @@ -651,292 +651,286 @@ "issue:43572|issue:45070", "issue:45800|issue:45823" ], - "prepared_review_unit_hash": "c9f7fd3b78391e5fef86b5c3af3e6fded97204dbcf1ba3d878a713044590b07d", + "prepared_review_unit_hash": "b5f4c4219911ff1fcade2a7ac1aca13a313ba26f548eaaf30defeb93f85b6547", "component_budget": { "node_count": 505, "item_count": 505, "soft_pair_count": 1854, - "serialized_chars": 962354, - "estimated_input_tokens": 240589, - "estimated_eval_tokens": 481434 + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 42, - "serialized_chars": 40038, - "estimated_input_tokens": 10010, - "estimated_eval_tokens": 20276 + "serialized_chars": 40042, + "estimated_input_tokens": 10011, + "estimated_eval_tokens": 20278 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "Mostly heterogeneous issue set with many false-positive similarity matches. I found two likely duplicate pockets: the v5 AutoTokenizer wrong-token-ID regression and a DeepSpeed ZeRO-3 load-time IndexError; everything else appears to be distinct model- or subsystem-specific failures.", - "confidence": 0.74, - "canonical_issue_reason": "issue:45920 is the strongest representative of the tokenizer regression cluster: it describes the same v5 AutoTokenizer wrong-token-ID bug but across multiple model families, making it broader and more canonical than the narrower Granite-only report.", + "summary": "Most items are unrelated. I only accept two soft duplicates: the ZeRO-3 load-time index-out-of-bounds reports, and the AutoTokenizer wrong-token-IDs regression reports.", + "confidence": 0.9, + "canonical_issue_reason": "issue:45920 is the best canonical issue: it describes the broad v5 AutoTokenizer wrong-token-ID regression across multiple model families, making it the clearest umbrella for the narrower Granite-specific report.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45920 is the best overall issue to anchor this cluster because it is the broadest, most general regression report with the clearest duplicate signal; it also has inbound references and covers the shared root symptom more completely than the narrower variants.", + "best_issue_reason": "issue:45920 is the best overall issue candidate because it is the broadest, most representative regression report in the set and likely subsumes issue:45812.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:45800", "right": "issue:46304", "accept": false, - "reason": "Both are broad compatibility regressions, but they describe different breakages and code paths; not the same underlying bug." + "reason": "Both are torch/version compatibility complaints, but they are about different breakages and code paths; not the same bug." }, { - "left": "issue:43866", - "right": "issue:46402", + "left": "issue:43296", + "right": "issue:46304", "accept": false, - "reason": "One is a corrupted checkpoint report, the other is missing text-encoder weights in a model implementation; different failure modes." + "reason": "Different failures: PaddleOCR-VL/vLLM model loading vs USE_DISTRIBUTED=0 support regression." }, { - "left": "issue:43296", - "right": "issue:46304", + "left": "issue:43866", + "right": "issue:46402", "accept": false, - "reason": "PaddleOCR-VL loading in vLLM and USE_DISTRIBUTED=0/PyTorch support are unrelated problems." + "reason": "Corrupted checkpoint vs weights not loaded; same area, different underlying problem." }, { "left": "issue:41762", "right": "issue:43596", "accept": true, - "reason": "Same DeepSpeed ZeRO-3 load/init IndexError pattern (`index 0 is out of bounds ... size 0`); different model classes, but the failure mode and code path look the same." + "reason": "Both report the same ZeRO-3 load-time index-out-of-bounds failure from an empty/invalid parameter path; model names differ, but the bug looks identical." }, { "left": "issue:39290", "right": "issue:41720", "accept": false, - "reason": "Both mention model/runtime failures, but one is a missing config attribute in vLLM and the other is a CUDA assert with auto device mapping; different bugs." + "reason": "Gemma3 sliding_window_pattern missing in vLLM vs Qwen3 auto device-map CUDA assert are unrelated." }, { "left": "issue:43901", "right": "issue:44512", "accept": false, - "reason": "Both are docs issues, but they cover different documentation mismatches and are not the same change." + "reason": "Both are docs issues, but they cover different documentation problems." }, { "left": "issue:44589", "right": "issue:46304", "accept": false, - "reason": "A float8 storage lookup error and a PyTorch USE_DISTRIBUTED=0 compatibility regression are unrelated." + "reason": "Float8 storage lookup failure is unrelated to the torch distributed compatibility breakage." }, { "left": "issue:45237", "right": "issue:45901", "accept": false, - "reason": "Different domains and symptoms: AMD GPU support vs table-question-answering task crash." + "reason": "Different subsystems and symptoms: AMD GPU support vs table-question-answering crash." }, { "left": "issue:43975", "right": "issue:46015", "accept": false, - "reason": "Incorrect detokenization and outdated RAG examples are separate issues." + "reason": "DeepSeek detokenization bug vs outdated RAG examples; not the same issue." }, { "left": "issue:43039", "right": "issue:44560", "accept": false, - "reason": "Liger kernel cross-entropy routing and Qwen3.5 video StopIteration are unrelated." + "reason": "Liger Kernel cross_entropy fallback vs Qwen3-VL video StopIteration are unrelated." }, { "left": "issue:45823", "right": "issue:46294", "accept": false, - "reason": "Both are model-specific input/device mismatch reports, but they affect different models and different pipelines." + "reason": "Different model-specific bugs: Gemma4 device mismatch vs Qwen3.5 video token/feature mismatch." }, { "left": "issue:44841", "right": "issue:45823", "accept": false, - "reason": "Processor failure for Voxtral and Gemma4 PLE device mismatch are different code paths." + "reason": "Processor failure for Voxtral vs Gemma4 device mismatch; no common code-path bug shown." }, { "left": "issue:43866", "right": "issue:46772", "accept": false, - "reason": "Checkpoint corruption vs NVIDIA ModelOpt compatibility are not the same bug." + "reason": "Checkpoint corruption and NVFP4 compatibility are distinct problems." }, { "left": "issue:44863", "right": "issue:46015", "accept": false, - "reason": "NemotronH checkpoint loading and RAG examples are unrelated." + "reason": "NemotronH loading failure vs RAG docs drift are unrelated." }, { "left": "issue:45020", "right": "issue:46304", "accept": false, - "reason": "Remote-code breakage and USE_DISTRIBUTED=0 support regression are different root causes." + "reason": "Remote-code model breakage and torch distributed support regression are different changes." }, { "left": "issue:45812", "right": "issue:45920", "accept": true, - "reason": "Same v5 AutoTokenizer wrong-token-ID regression; 45920 is the broader report and 45812 is a narrower Granite-specific instance." + "reason": "Same tokenizer regression: wrong token IDs in v5, with 45920 providing the broader multi-model scope and 45812 the Granite-specific instance." }, { - "left": "issue:43866", - "right": "issue:44863", + "left": "issue:44292", + "right": "issue:46015", "accept": false, - "reason": "Corrupted Ovis2 checkpoint and NemotronH checkpoint loading are distinct model-specific failures." + "reason": "Qwen NVFP4 runtime error vs outdated RAG examples; unrelated." }, { - "left": "issue:44292", - "right": "issue:46015", + "left": "issue:43866", + "right": "issue:44863", "accept": false, - "reason": "Qwen-3-8B-NVFP4 runtime failure and RAG docs are unrelated." + "reason": "Checkpoint corruption vs inability to load NemotronH checkpoints are different issues." }, { "left": "issue:43872", "right": "issue:46304", "accept": false, - "reason": "bitsandbytes constructor incompatibility and PyTorch USE_DISTRIBUTED=0 support are different issues." + "reason": "bitsandbytes constructor incompatibility is unrelated to the torch distributed build/runtime regression." }, { "left": "issue:39290", "right": "issue:43531", "accept": false, - "reason": "Both mention sliding-window behavior, but one is a missing attribute in vLLM and the other is a Qwen3-MoE-specific sliding_window problem." + "reason": "Both mention sliding_window, but the affected models and failure modes differ." }, { "left": "issue:41093", "right": "issue:43638", "accept": false, - "reason": "Similar IndexError wording, but one is a mask-shape mismatch and the other is a ZeRO-3 empty-index load failure; not the same bug." + "reason": "Both are index errors, but one is a mask/tensor shape mismatch and the other is a ZeRO-3 empty-tensor load issue." }, { "left": "issue:43792", "right": "issue:45901", "accept": false, - "reason": "Whisper runtime failure and table-question-answering crash are unrelated." + "reason": "Whisper runtime failure vs table QA crash; unrelated." }, { "left": "issue:43854", "right": "issue:43866", "accept": false, - "reason": "GLM-4.7-Flash unit-test load failure and Ovis2 checkpoint corruption are different model issues." + "reason": "Unit-test load failure and corrupted checkpoint are not the same bug." }, { "left": "issue:43425", "right": "issue:45070", "accept": false, - "reason": "Torch 2.10 compatibility and pydantic PretrainedConfig-field regression are separate compatibility problems." + "reason": "Torch 2.10 incompatibility and pydantic PretrainedConfig breakage are different regressions." }, { "left": "issue:43975", "right": "issue:45901", "accept": false, - "reason": "Detokenization regression and table-question-answering crash are unrelated." + "reason": "Detokenization bug vs table QA crash; no shared underlying change." }, { "left": "issue:45362", "right": "issue:45464", "accept": false, - "reason": "Both are Qwen3.5 chat-related, but one is a chat crash and the other is a streaming chat/completions failure; not clearly the same underlying bug." + "reason": "Both involve Qwen3.5, but one is chat crashes and the other is streaming API failure; not clearly the same bug." }, { "left": "issue:43792", "right": "issue:45070", "accept": false, - "reason": "Whisper runtime failure and pydantic model regression are unrelated." + "reason": "Whisper cannot run and a pydantic config regression are unrelated." }, { "left": "issue:43596", "right": "issue:45800", "accept": false, - "reason": "ZeRO-3 load failure and torch/transformers version incompatibility do not share the same concrete bug." + "reason": "ZeRO-3 load index error vs torch 2.4.1 compatibility complaint are different issues." }, { "left": "issue:43541", - "right": "issue:45685", + "right": "issue:46869", "accept": false, - "reason": "Grouped_mm torch-dynamo tracing failure and MPS histogram Int error are different backend-specific issues." + "reason": "Different models and failure modes: torch dynamo/grouped_mm vs YOLO multi-GPU auto device-map." }, { "left": "issue:44292", "right": "issue:45901", "accept": false, - "reason": "Qwen NVFP4 runtime error and table-question-answering crash are unrelated." + "reason": "Qwen NVFP4 runtime failure vs table QA crash are unrelated." }, { "left": "issue:39290", "right": "issue:45823", "accept": false, - "reason": "Missing sliding_window_pattern in vLLM and Gemma4 PLE device mismatch are different bugs." + "reason": "Different models and symptoms; no evidence of one shared bug." }, { "left": "issue:45464", "right": "issue:45685", "accept": false, - "reason": "Qwen3.5 streaming inference failure and MPS Int histogram error are unrelated." + "reason": "Chat/completions streaming failure and MPS histogram int error are unrelated." }, { "left": "issue:36010", "right": "issue:39290", "accept": false, - "reason": "ImportError for GenerationMixin and Gemma3 sliding_window_pattern attribute error are unrelated." + "reason": "ImportError for GenerationMixin and Gemma3 config attribute missing are separate regressions." }, { "left": "issue:43541", "right": "issue:45685", "accept": false, - "reason": "MoE grouped_mm tracing failure and MPS histogram Int failure are different backend/runtime problems." + "reason": "MoE-related titles are not enough; the concrete failures differ." }, { "left": "issue:43541", "right": "issue:43828", "accept": false, - "reason": "Both are MoE/precision-related, but the concrete failures differ: grouped_mm tracing vs autocast dtype mismatch." + "reason": "Grouped_mm tracing error vs autocast dtype mismatch are different concrete bugs." }, { "left": "issue:44841", "right": "issue:45084", "accept": false, - "reason": "Processor failure for Voxtral and a template-node compilation TypeError are unrelated." + "reason": "Processor failure and template-node compile error are unrelated." }, { "left": "issue:43404", "right": "issue:43866", "accept": false, - "reason": "Mistral3 weight-tying bug and Ovis2 checkpoint corruption are different issues." + "reason": "Weight-tying bug in Mistral3 is unrelated to a corrupted checkpoint report." }, { "left": "issue:45507", "right": "issue:46869", "accept": false, - "reason": "GraniteMoEHybrid invalid method and YOLO multi-GPU device-map failure are unrelated." - }, - { - "left": "issue:41093", - "right": "issue:41762", - "accept": false, - "reason": "Different IndexErrors with different symptoms: mask-shape mismatch vs ZeRO-3 empty-index load failure." + "reason": "Invalid method call in GraniteMoEHybrid vs YOLO multi-GPU device-map failure are unrelated." }, { "left": "issue:45399", "right": "issue:46661", "accept": false, - "reason": "Flash-attn fallback gating and remote-code model_class/config_class AttributeError are unrelated." + "reason": "Flash-Attn fallback checks and remote-code config_class errors are different code paths." }, { "left": "issue:43572", "right": "issue:45070", "accept": false, - "reason": "Missing pad_token_idx in StableLmConfig and a PretrainedConfig pydantic regression are different changes." + "reason": "Missing pad_token_idx in StableLmConfig is not the same as the pydantic PretrainedConfig field regression." }, { "left": "issue:45800", "right": "issue:45823", "accept": false, - "reason": "Torch 2.4.1/transformers 5.8.0 incompatibility and Gemma4 device mismatch are unrelated." + "reason": "Torch version compatibility and Gemma4 device mismatch are unrelated." } ] }, "evaluator_result": { "accept": true, - "feedback": "The summary is grounded in the packet and stays conservative on the soft-edge calls. The two accepted duplicate pockets are well supported by the titles/body signals, and the rest of the rejects do not overclaim shared root cause." + "feedback": "The summary is grounded in the packet, and the two accepted soft duplicates are conservative and well-supported by the issue titles: the ZeRO-3 index-out-of-bounds pair and the AutoTokenizer wrong-token-ID pair. The other verdicts stay cautious rather than overclaiming, and the canonical issue choice is reasonable." }, "evaluator_used": true, "retried": false, @@ -947,7 +941,7 @@ { "label": "issue", "component_index": 1, - "component_count": 25, + "component_count": 26, "review_unit_index": 3, "review_unit_count": 61, "status": "reviewed", @@ -1038,237 +1032,234 @@ "issue:45183|issue:46463", "issue:44918|issue:45507" ], - "prepared_review_unit_hash": "1f8d8a6b3faa7805b434732cec57f348b0f5a6e27acebc969e37182a4aaada65", + "prepared_review_unit_hash": "c3209c118ae64a33c897f277cdf4f5e57c6395584fe5f7a9ac3af1a0c402da11", "component_budget": { "node_count": 505, "item_count": 505, "soft_pair_count": 1854, - "serialized_chars": 962354, - "estimated_input_tokens": 240589, - "estimated_eval_tokens": 481434 + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 }, "budget": { "node_count": 47, "item_count": 47, "soft_pair_count": 31, - "serialized_chars": 36244, - "estimated_input_tokens": 9061, - "estimated_eval_tokens": 18378 + "serialized_chars": 36237, + "estimated_input_tokens": 9060, + "estimated_eval_tokens": 18376 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly heterogeneous with many false-positive similarities. Two issue pairs look like genuine duplicates: a remote-code cache/module resolution bug and an input-validation error-message bug. No PRs are present.", - "confidence": 0.61, - "canonical_issue_reason": "No single issue cleanly represents the whole cluster because it mixes unrelated topics. If a representative anchor is needed, issue 45183 is the best fit: it captures the broader underlying input-validation/XOR bug, while 46463 looks like a narrower T5-family instance of the same problem.", - "canonical_pr_reason": null, - "best_issue_reason": "Issue 45183 is the strongest duplicate anchor among the set because it describes the root validation logic problem in the most general way and can plausibly subsume 46463.", - "best_pr_reason": null, + "summary": "This cluster is heterogeneous and does not look like a true duplicate set. The soft pairs mostly share broad symptoms or subsystem labels, but they describe different concrete bugs, models, or code paths, so none should be merged as duplicates.", + "confidence": 0.95, + "canonical_issue_reason": "No single issue is a safe canonical representative: the items span unrelated regressions (FSDP, MoE, video I/O, remote_code, quantization, docs, etc.) and the soft pairs do not match on a concrete underlying bug.", + "canonical_pr_reason": "No pull requests are present in this cluster.", + "best_issue_reason": "No issue is a good global best representative because the cluster is not centered on one bug; choosing any one would mischaracterize the rest of the items.", + "best_pr_reason": "No pull requests are present in this cluster.", "soft_edge_verdicts": [ { "left": "issue:43296", "right": "issue:43366", "accept": false, - "reason": "Different products and failures: PaddleOCR-VL loading in vLLM vs GGUF gpt-oss architecture support." + "reason": "Different problems: PaddleOCR-VL loading failure vs GGUF/gpt-oss architecture support." }, { "left": "issue:44805", "right": "issue:44918", "accept": false, - "reason": "Both mention shape issues, but the code paths differ: generic mask/tensor indexing vs Qwen3.5 embedding unpacking in TRL SFT." + "reason": "Both are shape-related errors, but one is a generic mask/tensor mismatch and the other is Qwen3.5 embedding unpacking with TRL SFT; not the same bug." }, { "left": "issue:43541", "right": "issue:45161", "accept": false, - "reason": "Both involve MoE models, but one is a torch-dynamo grouped_mm tracing crash and the other is tensor-parallel support not working for GPT-OSS." + "reason": "Different failures: torch dynamo grouped_mm tracing on Mixtral vs tensor-parallel-only GPT-OSS MoE behavior." }, { "left": "issue:43828", "right": "issue:45464", "accept": false, - "reason": "Different failure modes: autocast dtype mismatch in Phi-tiny-MoE vs streaming chat/completions on Qwen3.5-0.8B." + "reason": "Different code paths and models: autocast dtype mismatch in Phi-tiny-MoE vs Qwen3.5 streaming chat/completions regression." }, { "left": "issue:44387", "right": "issue:45005", "accept": false, - "reason": "One is int4 quantization memory/OOM regression; the other is tied-weights handling on translation models." + "reason": "OOM under int4 quantization is unrelated to tied-weight issues in translation models." }, { "left": "issue:43366", "right": "issue:45464", "accept": false, - "reason": "GGUF gpt-oss support and Qwen3.5 streaming inference are unrelated." + "reason": "GGUF gpt-oss support is a model-format request, not the Qwen3.5 streaming inference failure." }, { "left": "issue:44291", "right": "issue:45070", "accept": false, - "reason": "Init-empty-weights argument handling and pydantic PretrainedConfig field breakage are separate compatibility bugs." + "reason": "Init-empty-weights/_is_hf_initialized TypeError is unrelated to the PretrainedConfig pydantic regression." }, { "left": "issue:43531", "right": "issue:45070", "accept": false, - "reason": "Qwen3-MoE sliding-window behavior is unrelated to the pydantic model field regression." + "reason": "Qwen3-MoE sliding_window behavior and pydantic PretrainedConfig breakage are different bugs." }, { "left": "issue:41093", "right": "issue:43596", "accept": false, - "reason": "Both are IndexErrors, but one is a mask/tensor size mismatch and the other is a zero3/BertModel empty-index failure." + "reason": "Both raise IndexError, but one is a mask/tensor size mismatch and the other is a DeepSpeed ZeRO-3/BertModel init failure." }, { "left": "issue:43716", "right": "issue:46294", "accept": false, - "reason": "Different multimodal bugs: image preprocessor/model dtype mismatch vs video feature/token count mismatch." + "reason": "Different multimodal mismatches: image-preprocessor dtype issue vs video-token/feature count mismatch in Qwen3.5." }, { "left": "issue:45020", "right": "issue:45800", "accept": false, - "reason": "Remote-code regressions and torch/transformers version incompatibility are different root causes." + "reason": "Remote-code compatibility regression and torch 2.4.1/transformers 5.8.0 incompatibility are distinct issues." }, { "left": "issue:43054", "right": "issue:43994", "accept": false, - "reason": "Same model family, but one reports weak text embeddings while the other reports bad AutoModel/pipeline results; not enough to assert one bug." + "reason": "Both involve SigLIP2, but one is an embedding-quality complaint and the other is a loading/pipeline correctness bug." }, { "left": "issue:44479", "right": "issue:46294", "accept": false, - "reason": "Both are video/multimodal-related, but one is a v5.3.0 regression across several Qwen VL variants and the other is a Qwen3.5-9B chat template mismatch." + "reason": "Broad video-input regression across several models is not the same concrete bug as Qwen3.5 chat_template video feature/token mismatch." }, { "left": "issue:46468", "right": "issue:46869", "accept": false, - "reason": "Official Nemotron example breakage and YOLO multi-GPU device-map failure are unrelated subsystems." + "reason": "Nemotron example breakage and YOLO multi-GPU device-map failure are unrelated." }, { "left": "issue:45759", "right": "issue:45902", "accept": false, - "reason": "Composite-config unwrapping and uninitialized rotary embedding memory are different code paths in Qwen 3.5/3.6 models." + "reason": "Composite Qwen config unwrapping and uninitialized inv_freq after to_empty(cuda) are different Qwen issues." }, { "left": "issue:43596", "right": "issue:46165", "accept": false, - "reason": "Zero3 initialization with BertModel and resume training with trust_remote_code are distinct issues." + "reason": "DeepSpeed zero3 init failure with BertModel is unrelated to resuming training with trust_remote_code." }, { "left": "issue:46640", "right": "issue:46661", "accept": false, - "reason": "Flash Attention 2 fallback problems and remote-code model_class config_class errors are unrelated." + "reason": "Flash Attention 2 issues and remote-code config_class AttributeError are separate code paths." }, { "left": "issue:43716", "right": "issue:44918", "accept": false, - "reason": "One is image preprocessor dtype handling; the other is SFT embedding unpacking for Qwen3.5." + "reason": "Mistral-3 image dtype mismatch and Qwen3.5 embedding unpacking with TRL SFT are unrelated." }, { "left": "issue:39186", "right": "issue:44918", "accept": false, - "reason": "FSDP 2-D weight error and Qwen3.5 embedding unpacking failure do not share the same code path." + "reason": "FSDP weight dimensionality error and Qwen3.5/TRL embedding unpacking are not the same bug." }, { "left": "issue:40990", "right": "issue:44960", "accept": false, - "reason": "High perplexity on gpt-oss-20b and the GLM5 issue are unrelated model-quality reports." + "reason": "High perplexity on gpt-oss-20b and GLM5 are unrelated reports." }, { "left": "issue:43645", "right": "issue:45399", "accept": false, - "reason": "Custom-model notebook initialization breakage and flash-attn2 fallback gating are separate v5 regressions." + "reason": "Custom-model notebook initialization breakage and flash-attn2 fallback gating are different issues." }, { "left": "issue:45632", "right": "issue:45698", - "accept": true, - "reason": "Both point to the same trust_remote_code/custom-module resolution bug: cache/path reuse causes the wrong module to be loaded after save/load." + "accept": false, + "reason": "Both concern remote-code/custom models, but one is cache-path collision and the other is wrong module selection after save/load; not the same concrete defect." }, { "left": "issue:43404", "right": "issue:45005", "accept": false, - "reason": "Tied lm_head weights in Mistral3 and translation-model tied-weights issues are different model families and failure modes." + "reason": "Tied lm_head not tied in Mistral3 is different from tied-weight issues in translation models." }, { "left": "issue:42915", "right": "issue:45399", "accept": false, - "reason": "FineGrainedFP8Config failure on Qwen3Moe and flash-attn2 fallback blocking are unrelated." + "reason": "Qwen3MoE with FineGrainedFP8Config and flash-attn2 fallback gating are unrelated." }, { "left": "issue:43873", "right": "issue:45005", "accept": false, - "reason": "Quantization/offloading behavior and tied-weights handling are different bugs." + "reason": "Quantization/offloading behavior and tied-weight handling are different problems." }, { "left": "issue:44360", "right": "issue:44512", "accept": false, - "reason": "A DSA indexer algorithm note and a docs typo are not the same underlying issue." + "reason": "DSA indexer ReLU discussion and removed run-command documentation are unrelated." }, { "left": "issue:44530", "right": "issue:45005", "accept": false, - "reason": "PagedAttentionCache linear_attention crash and tied-weights issues are unrelated." + "reason": "PagedAttentionCache linear_attention crash on Qwen3.5 is not the same as translation-model tied-weight bugs." }, { "left": "issue:44492", "right": "issue:44512", "accept": false, - "reason": "Cache-strategies typo and removed run-command docs are both documentation-level but not the same bug." + "reason": "A cache-strategy typo and outdated docs about the removed run command are unrelated." }, { "left": "issue:44863", "right": "issue:45005", "accept": false, - "reason": "NemotronH checkpoint loading and translation-model tied weights are unrelated implementation problems." + "reason": "NemotronH checkpoint loading failure is unrelated to tied weights in translation models." }, { "left": "issue:45183", "right": "issue:46463", - "accept": true, - "reason": "These look like the same validation/error-message bug: the broad input_ids/inputs_embeds XOR logic issue, with 46463 as a T5-specific manifestation." + "accept": false, + "reason": "Both are validation/error-message issues, but one is input_ids/inputs_embeds XOR logic and the other is T5 decoder-input validation; not the same bug." }, { "left": "issue:44918", "right": "issue:45507", "accept": false, - "reason": "Qwen3.5 embedding unpacking in TRL and GraniteMoEHybrid invalid method calls are different failures." + "reason": "Qwen3.5 embedding unpacking with TRL SFT and GraniteMoEHybrid invalid-method calls are unrelated." } ] }, - "evaluator_result": { - "accept": true, - "feedback": "The summary is grounded in the packet and stays conservative. It correctly identifies the cluster as mostly heterogeneous, notes the absence of PRs, and only elevates two issue pairs that are plausibly duplicate-style matches based on the titles/body themes. The soft-edge rejections are appropriately cautious rather than overstated." - }, - "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 }, { "label": "issue", "component_index": 1, - "component_count": 25, + "component_count": 26, "review_unit_index": 4, "review_unit_count": 61, "status": "reviewed", @@ -1342,8 +1333,8 @@ "issue:38175|issue:43054", "issue:43950|issue:45902", "issue:44910|issue:45859", - "issue:43054|issue:43493", "issue:41553|issue:43883", + "issue:43054|issue:43493", "issue:42947|issue:44387", "issue:45663|issue:45902", "issue:44403|issue:44451", @@ -1358,223 +1349,223 @@ "issue:43782|issue:44960", "issue:43824|issue:43872" ], - "prepared_review_unit_hash": "3a31d40021edabdf90bbebc04f02a96172b8b4e323a3c3353793749cc41cc458", + "prepared_review_unit_hash": "c81dd76313996f9095ebf799cc756c799d481ee1021ec4d62e05eac36e008ab4", "component_budget": { "node_count": 505, "item_count": 505, "soft_pair_count": 1854, - "serialized_chars": 962354, - "estimated_input_tokens": 240589, - "estimated_eval_tokens": 481434 + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 }, "budget": { "node_count": 47, "item_count": 47, "soft_pair_count": 30, - "serialized_chars": 36285, - "estimated_input_tokens": 9072, - "estimated_eval_tokens": 18400 + "serialized_chars": 36283, + "estimated_input_tokens": 9071, + "estimated_eval_tokens": 18398 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The set is mostly heterogeneous and not a single duplicate cluster. Clear duplicate subgroups are the non-persistent buffer regression issues, the Gemma4 quantized embedder bug, and likely the StaticCache `get_seq_length()` contract regression. Most other pairs are different models, different code paths, or different symptom classes.", - "confidence": 0.88, - "canonical_issue_reason": "Issue #43950 is the clearest representative of the strongest duplicate theme in this set: the transformers 5.x non-persistent-buffer regression. Its title is specific about the failure mode, and it aligns with the related buffer-loading reports in the soft edges.", + "summary": "The cluster has a few solid duplicates around the same concrete regressions, especially the v5 non-persistent-buffer corruption and the StaticCache return-type bug. The SigLIP2 pair is related but not strong enough to merge as the same bug from the available titles alone: one is a broad quality regression and the other is a specific HF-vs-JAX discrepancy. Overall this looks like a mostly conservative cluster with a few clear matches, not a set of indisputable duplicates.", + "confidence": 0.74, + "canonical_issue_reason": "#43950 is the clearest anchor for the v5 non-persistent-buffer corruption regression, and its wording best covers the same issue seen in #43644/#44534.", "canonical_pr_reason": null, - "best_issue_reason": "#43950 is the best anchor issue for the only clearly repeated underlying bug class here; it is broader and more root-cause-oriented than the later symptom reports.", + "best_issue_reason": "#43950 most directly names the concrete bug and regression scope, making it the best duplicate target among the issues here.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43975", "right": "issue:44292", "accept": false, - "reason": "Different failures and different models: DeepSeek tokenizer detokenization vs Qwen-3 NVFP4 loading." + "reason": "Different models and different failure modes; the titles do not indicate one shared bug." }, { "left": "issue:46489", "right": "issue:46491", "accept": false, - "reason": "Both are tokenizer bugs, but they affect different models and different round-trip behaviors; not the same concrete bug." + "reason": "Both are tokenizer regressions, but the affected models and concrete symptoms differ too much to treat as one bug." }, { "left": "issue:43950", "right": "issue:44534", "accept": true, - "reason": "Same underlying regression: transformers v5 non-persistent buffers being corrupted/filled with junk during load." + "reason": "Same v5 regression corrupting non-persistent buffers / filling them with junk." }, { "left": "issue:39692", "right": "issue:43054", "accept": false, - "reason": "Documentation example errors vs model quality/embedding degradation are not the same bug." + "reason": "Doc-example errors vs. a SigLIP2 quality regression; not the same concrete bug." }, { "left": "issue:43644", "right": "issue:43950", - "accept": false, - "reason": "Remote-code config field loss is unrelated to non-persistent buffer corruption." + "accept": true, + "reason": "Both report the same non-persistent-buffer junk/corruption regression in Transformers 5.x." }, { "left": "issue:36331", "right": "issue:41762", "accept": false, - "reason": "Custom-trainer loss-signature error and Gemma3 DeepSpeed load failure are different code paths." + "reason": "Unrelated training API signature break vs. DeepSpeed loading index error." }, { "left": "issue:46899", "right": "issue:46932", "accept": true, - "reason": "These describe the same Gemma4 quantized vision/audio embedder dtype-casting bug; #46932 even references the same issue." + "reason": "Same Gemma4 multimodal 4-bit quantization dtype-casting bug." }, { "left": "issue:43872", "right": "issue:44589", "accept": false, - "reason": "Different quantization/storage incompatibilities with different exception surfaces." + "reason": "Different quantization/storage errors affecting different code paths." }, { "left": "issue:43828", "right": "issue:44928", "accept": false, - "reason": "Autocast dtype mismatch in Phi MoE is not the same as the Qwen3.5 3D position_ids / SDPA fallback issue." + "reason": "Different models and different numerical issues; not the same failure." }, { "left": "issue:45920", "right": "issue:46489", "accept": false, - "reason": "Related tokenizer regression theme, but different affected models and different concrete symptoms; too broad to merge." + "reason": "Tokenizer regressions in different model families, but not one concrete shared bug." }, { "left": "issue:43054", "right": "issue:45701", "accept": false, - "reason": "SigLIP2 embedding quality complaint is unrelated to a general tokenization-change report." + "reason": "SigLIP2 embedding-quality issue vs. a broad tokenization-change report; too different to merge." }, { "left": "issue:45987", "right": "issue:46858", "accept": true, - "reason": "Both point to the same StaticCache `get_seq_length()` contract/regression, just observed in different execution contexts." + "reason": "Same StaticCache get_seq_length return-type/behavior bug; one describes the contract violation, the other the downstream breakage." }, { "left": "issue:38175", "right": "issue:43054", "accept": false, - "reason": "Same model family, but zero probabilities and worse text embeddings are not clearly the same underlying bug." + "reason": "Related SigLIP2 symptom area, but no clear shared underlying bug from the titles alone." }, { "left": "issue:43950", "right": "issue:45902", - "accept": true, - "reason": "Both are the non-persistent-buffer/materialization regression class: corrupted or uninitialized buffers after loading/materialization." + "accept": false, + "reason": "Both mention bad initialization-like behavior, but they are different code paths and different objects." }, { "left": "issue:44910", "right": "issue:45859", "accept": false, - "reason": "Flash-Attention packed-sequence misinterpretation and CPU-offload incompatibility are different problems." + "reason": "Flash-Attention packed-sequence crash vs. CPU-offload incompatibility; unrelated issues." }, { - "left": "issue:43054", - "right": "issue:43493", + "left": "issue:41553", + "right": "issue:43883", "accept": false, - "reason": "Both concern SigLIP2, but one is a quality discrepancy vs the original JAX implementation, not the same bug as worse text embeddings." + "reason": "Bad error message vs. missing attribute; different failures." }, { - "left": "issue:41553", - "right": "issue:43883", + "left": "issue:43054", + "right": "issue:43493", "accept": false, - "reason": "Bad AutoTokenizer error message for Voxtral and Molmo missing `all_tied_weights_keys` are unrelated." + "reason": "Related SigLIP2 area, but one is a broad text-embedding quality regression and the other is a specific HF-vs-JAX discrepancy; titles alone do not prove the same concrete bug." }, { "left": "issue:42947", "right": "issue:44387", "accept": false, - "reason": "Gradient checkpointing behavior and int4 OOM/reserved-memory growth are different issues." + "reason": "Different training/memory regressions with no shared concrete code-path." }, { "left": "issue:45663", "right": "issue:45902", "accept": false, - "reason": "Gemma4 FSDP2 shared-kv-state rebuild bug is unrelated to non-persistent buffer initialization." + "reason": "Different model internals and different memory/state bugs." }, { "left": "issue:44403", "right": "issue:44451", "accept": false, - "reason": "Generic loading noise vs a specific model-loading failure are not duplicates." + "reason": "Generic load-noise complaint vs. a specific model load failure; not the same bug." }, { "left": "issue:43975", "right": "issue:45520", "accept": false, - "reason": "Tokenizer detokenization bug and Python 3.13 import error for flash_attn are unrelated." + "reason": "Tokenizer/detokenization issue vs. import-utils flash_attn KeyError; unrelated." }, { "left": "issue:24643", "right": "issue:28530", "accept": false, - "reason": "DeepSpeed training runtime error and early-stopping metric resolution are different trainer issues." + "reason": "DeepSpeed training runtime error vs. early-stopping metric configuration problem." }, { "left": "issue:42947", "right": "issue:43856", "accept": false, - "reason": "LoRA checkpointing ineffectiveness and Qwen3 MoE memory usage are different training-memory problems." + "reason": "Both are memory-related, but they describe different training settings and different failure modes." }, { "left": "issue:44403", "right": "issue:44488", "accept": false, - "reason": "Different model-loading complaints with no shared code-path evidence." + "reason": "Both mention loading, but they concern different models and no common bug is clear." }, { "left": "issue:43643", "right": "issue:45020", "accept": false, - "reason": "Remote-code config field loss is not the same concrete bug as the broader 'models using remote_code break' report." + "reason": "trust_remote_code missing-fields bug vs. a broad remote_code regression; too general to merge." }, { "left": "issue:41762", "right": "issue:42915", "accept": false, - "reason": "Gemma3 ZeRO-3 load failure and Qwen3 MoE FineGrainedFP8 failure are unrelated." + "reason": "Different models and different failure classes (ZeRO-3 load vs. FineGrainedFP8Config)." }, { "left": "issue:41553", "right": "issue:43881", "accept": false, - "reason": "Voxtral tokenizer error messaging and glm-4v-9b load failure are different artifacts." + "reason": "Different model load failures with no shared concrete cause." }, { "left": "issue:44991", "right": "issue:45399", "accept": false, - "reason": "Tokenizer loading regression and flash-attn2 fallback gating are unrelated." + "reason": "Tokenizer loading regression vs. flash-attn fallback logic; unrelated." }, { "left": "issue:43782", "right": "issue:44960", "accept": false, - "reason": "Qwen3VL weight_only loading failure and GLM5 are different issues." + "reason": "Different model families and different errors; no shared code-path bug." }, { "left": "issue:43824", "right": "issue:43872", "accept": false, - "reason": "Missing model export/import symbol and bitsandbytes parameter incompatibility are unrelated." + "reason": "Import error for a missing class vs. bitsandbytes constructor mismatch; unrelated." } ] }, "evaluator_result": { - "accept": true, - "feedback": "The cluster assessment is grounded in the packet and the soft-edge calls are mostly conservative. The accepted duplicates are well supported by the titles and issue descriptions (#43950/#44534, #46899/#46932, #45987/#46858). The summary also avoids overclaiming a single unified cluster and correctly treats most other pairs as distinct enough to reject." + "accept": false, + "feedback": "Most of the cluster calls are grounded, but the SigLIP2 accept (#43054/#43493) is a bit too strong for the packet: one issue is a broad quality regression and the other is a specific HF-vs-JAX discrepancy, and the titles alone do not prove they are the same bug. The summary\u2019s wording about \u2018genuine duplicate clusters\u2019 is therefore slightly overconfident. The other accepted pairs look conservative enough." }, "evaluator_used": true, - "retried": false, + "retried": true, "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null @@ -1582,7 +1573,7 @@ { "label": "issue", "component_index": 1, - "component_count": 25, + "component_count": 26, "review_unit_index": 5, "review_unit_count": 61, "status": "reviewed", @@ -1651,11 +1642,11 @@ "issue:42491|issue:44368", "issue:45020|issue:45356", "issue:45715|issue:46315", - "issue:43827|issue:44509", - "issue:45488|issue:46489", "issue:43872|issue:46869", + "issue:45488|issue:46489", "issue:45759|issue:46661", "issue:43404|issue:43950", + "issue:43827|issue:44509", "issue:46894|issue:46897", "issue:43872|issue:44291", "issue:45020|issue:46661", @@ -1664,27 +1655,27 @@ "issue:42947|issue:45923", "issue:36331|issue:41093", "issue:41553|issue:44991", - "issue:43854|issue:44863", "issue:43940|issue:45313", + "issue:43854|issue:44863", "issue:44928|issue:45588", - "issue:44479|issue:44560", "issue:41093|issue:44805", + "issue:44479|issue:44560", "issue:24643|issue:30064" ], - "prepared_review_unit_hash": "db1fd57f2b2ddd44ec72c6395b9eb792f7377dd4e30369dfd3fc00bc190f03d6", + "prepared_review_unit_hash": "774bba1b5d32f5ab124400a05017e313cd4ee918d401efdf5f3206bf238b48d6", "component_budget": { "node_count": 505, "item_count": 505, "soft_pair_count": 1854, - "serialized_chars": 962354, - "estimated_input_tokens": 240589, - "estimated_eval_tokens": 481434 + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 28, - "serialized_chars": 36755, + "serialized_chars": 36753, "estimated_input_tokens": 9189, "estimated_eval_tokens": 18634 }, @@ -1693,186 +1684,186 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly a mix of recent transformers regressions across model loading, tokenization, video/audio processing, and docs. Only a few soft pairs look like true duplicates; most are separate per-model failures or adjacent but distinct bugs.", - "confidence": 0.61, - "canonical_issue_reason": "Issue 45020 is the broadest umbrella here: it captures the general \u2018recent transformers versions break models using remote_code\u2019 loading-regression theme that several other reports resemble, and it has enough discussion to serve as a cluster anchor.", + "summary": "Most pairs are only broad thematic matches, not true duplicates. Clear duplicate-looking groups appear for docs mentioning removed pipeline tasks, the `_is_hf_initialized` TypeError, Qwen3-family DeepSpeed ZeRO-3 load failures, the identical mask/tensor shape IndexError, and the training-loss double-shift bug.", + "confidence": 0.73, + "canonical_issue_reason": "Issue 44291 is the broadest/root-cause report for the `_is_hf_initialized` loading regression, making it the best canonical issue among the duplicates in this set.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 45020 is the best single representative because it is the most general and central loading-regression report in this set; the others are narrower manifestations or separate model-specific failures.", + "best_issue_reason": "44291 is the most reusable representative because it describes the underlying TypeError generically rather than a single downstream symptom.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:45685", "right": "issue:45823", "accept": false, - "reason": "Different bugs: MPS histogram/int dtype issue vs Gemma4 PLE device-map forward mismatch." + "reason": "Different failures: MPS histogram/int dtype error vs Gemma4 PLE device placement mismatch." }, { "left": "issue:45081", "right": "issue:45399", "accept": false, - "reason": "Both touch tokenizer/attention internals, but one is a Mistral regex crash and the other is flash-attn fallback gating; not the same code-path bug." + "reason": "Both involve tokenizer/model loading, but one is a Mistral regex crash and the other is flash-attn2 fallback gating; not the same bug." }, { "left": "issue:45020", "right": "issue:45823", "accept": false, - "reason": "Broad remote_code loading regressions vs a specific Gemma4 device-mismatch forward bug." + "reason": "Remote-code model breakage is a different problem from a Gemma4 forward device mismatch." }, { "left": "issue:44661", "right": "issue:45020", "accept": false, - "reason": "Related area, but one is add-new-model-like handling of TOKENIZER_MAPPING_NAMES; the other is a broader remote_code loading regression." + "reason": "Add-new-model-like TOKENIZER_MAPPING_NAMES failure is unrelated to the broader remote-code regression." }, { "left": "issue:43881", "right": "issue:43883", "accept": false, - "reason": "Both are load failures, but they affect different model families and symptoms (GLM-4V vs Molmo tied-weights attribute)." + "reason": "Different model-loading errors for different models; only loosely similar." }, { "left": "issue:43828", "right": "issue:45823", "accept": false, - "reason": "Distinct problems: autocast dtype mismatch in Phi-tiny-MoE vs Gemma4 device placement during forward." + "reason": "Autocast dtype mismatch on Phi MoE is not the same as a device-map forward mismatch." }, { "left": "issue:42491", "right": "issue:44368", "accept": false, - "reason": "Different LoRA/tie-word-embeddings warnings vs a cross-version Qwen3 MOE compatibility issue; not the same concrete bug." + "reason": "Both mention LoRA/Qwen embedding tying, but they describe different symptoms and code paths." }, { "left": "issue:45020", "right": "issue:45356", "accept": false, - "reason": "Both involve tokenizer/model regressions, but Kimi-K2.5 codec handling is a separate tokenizer-specific issue." + "reason": "Broad remote-code breakage vs a specific Kimi-K2.5 tokenizer regression; not the same underlying bug." }, { "left": "issue:45715", "right": "issue:46315", "accept": false, - "reason": "Same tokenizer API family, but opposite directions and different performance hot paths; not a single bug." + "reason": "These are inverse tokenizer-performance regressions; similar pattern, but not the same defect." }, { - "left": "issue:43827", - "right": "issue:44509", - "accept": true, - "reason": "Both are the same docs cleanup: references to removed v5 pipeline tasks (summarization/translation/text2text-generation) still remain." + "left": "issue:43872", + "right": "issue:46869", + "accept": false, + "reason": "Bitsandbytes/`_is_hf_initialized` incompatibility is unrelated to YOLO multi-GPU device-map failure." }, { "left": "issue:45488", "right": "issue:46489", "accept": false, - "reason": "Both are tokenizer regressions, but they hit different model families and different tokenizer mechanics." - }, - { - "left": "issue:43872", - "right": "issue:46869", - "accept": false, - "reason": "Bitsandbytes `_is_hf_initialized` constructor issue vs a YOLO multi-GPU device-map problem; unrelated failures." + "reason": "Both concern tokenizer regressions in DeepSeek-family models, but the described bugs and code paths differ." }, { "left": "issue:45759", "right": "issue:46661", "accept": false, - "reason": "Both concern model construction, but one is composite Qwen text_config unwrapping and the other is remote-code config_class handling." + "reason": "Both involve model/config loading APIs, but one is `text_config` unwrapping and the other is missing `config_class` on remote-code models." }, { "left": "issue:43404", "right": "issue:43950", "accept": false, - "reason": "Different regressions: tied lm_head weights in Mistral3 vs silent corruption of non-persistent buffers on from_pretrained." + "reason": "Untied lm_head weights and silent non-persistent buffer corruption are different regressions." + }, + { + "left": "issue:43827", + "right": "issue:44509", + "accept": true, + "reason": "Same docs regression: both say removed pipeline tasks are still referenced after v5 removals." }, { "left": "issue:46894", "right": "issue:46897", - "accept": false, - "reason": "Same training-loss pattern, but these are different model families; not enough to treat as one bug." + "accept": true, + "reason": "Same underlying training-loss double-shift bug, reported for two model families." }, { "left": "issue:43872", "right": "issue:44291", "accept": true, - "reason": "Same underlying `_is_hf_initialized` keyword regression during model initialization/loading; 43872 is the bitsandbytes manifestation, 44291 the generic init_empty_weights case." + "reason": "Same `_is_hf_initialized` TypeError during model loading; 44291 gives the generic root-cause framing." }, { "left": "issue:45020", "right": "issue:46661", "accept": false, - "reason": "Both mention remote-code models, but the failure modes differ: general loading breakage vs missing config_class on AutoModel.from_pretrained." + "reason": "Remote-code models failing generally is broader than the specific missing `config_class` AttributeError." }, { "left": "issue:43792", "right": "issue:44220", "accept": false, - "reason": "The titles suggest a possible relation, but one is a generic Whisper failure and the other is a specific fbank helper issue; not enough evidence for duplicate status." + "reason": "Whisper model run failure and `_torch_extract_fbank_features()` are not the same issue." }, { "left": "issue:43257", "right": "issue:45305", "accept": false, - "reason": "Different DeepSpeed bugs: weight conversion/loading under accelerate+deepspeed vs gradient averaging under GAS." + "reason": "Both involve DeepSpeed/Qwen, but one is weight conversion and the other is gradient averaging; different bugs." }, { "left": "issue:42947", "right": "issue:45923", "accept": false, - "reason": "One is LoRA gradient checkpointing ineffective; the other is a missing supports_gradient_checkpointing flag on a trust_remote_code variant." + "reason": "Gradient checkpointing with PEFT LoRA is unrelated to a missing supports_gradient_checkpointing flag on a trust_remote_code variant." }, { "left": "issue:36331", "right": "issue:41093", "accept": false, - "reason": "Different trainer/runtime errors: unexpected compute_loss kwarg vs mask/index shape mismatch." + "reason": "Unexpected `num_items_in_batch` kwarg is unrelated to the mask/index shape mismatch." }, { "left": "issue:41553", "right": "issue:44991", "accept": false, - "reason": "Both are tokenizer-loading issues, but the model families and failure causes differ." - }, - { - "left": "issue:43854", - "right": "issue:44863", - "accept": false, - "reason": "Different checkpoints and model classes; both fail to load, but not the same underlying bug." + "reason": "Both are tokenizer-related, but the Voxtral error-message issue and est-roberta tokenizer loading failure are different bugs." }, { "left": "issue:43940", "right": "issue:45313", + "accept": true, + "reason": "Same Qwen3-family DeepSpeed ZeRO-3 weight-loading failure; different variants of the same missing-weights path." + }, + { + "left": "issue:43854", + "right": "issue:44863", "accept": false, - "reason": "Both are ZeRO-3 weight-loading failures in Qwen variants, but the affected model paths and missing-weight symptoms are distinct enough to keep separate." + "reason": "GLM-4.7-Flash test loading failure and NemotronH checkpoint loading failure are unrelated." }, { "left": "issue:44928", "right": "issue:45588", "accept": false, - "reason": "Different failures: RLHF NaNs from 3D position_ids/SDPA fallback vs flash-attention sink-less-model crash on s_aux=None." + "reason": "RLHF gradient explosion from 3D position_ids is unrelated to flash_attention s_aux handling." }, { - "left": "issue:44479", - "right": "issue:44560", + "left": "issue:41093", + "right": "issue:44805", "accept": true, - "reason": "44560 looks like a specific manifestation of the broader v5.3.0 Qwen video-input regression described in 44479." + "reason": "Same mask-vs-indexed-tensor shape mismatch error; 44805 is the generic follow-up report." }, { - "left": "issue:41093", - "right": "issue:44805", + "left": "issue:44479", + "right": "issue:44560", "accept": false, - "reason": "Both are mask shape errors, but the actual failing models/code paths are not shown to be the same." + "reason": "Both are video-input regressions, but they affect different model families and failure modes." }, { "left": "issue:24643", "right": "issue:30064", "accept": false, - "reason": "Completely different subsystems: DeepSpeed training weight-dimension error vs image processor void segmentation map handling." + "reason": "DeepSpeed training weight-shape runtime error is unrelated to void segmentation map processing." } ] }, "evaluator_result": { "accept": true, - "feedback": "The summary is grounded in the packet: issue 45020 is a reasonable broad anchor for the remote-code loading-regression subset, and the narrative correctly characterizes the rest as a mix of distinct model-specific or subsystem-specific bugs. The three accepted soft pairs are conservative and supported by the titles/body themes, while the rejected pairs avoid overclaiming duplicate status." + "feedback": "The grouping is grounded in the packet and the soft-edge verdicts are suitably conservative. The accepted pairs are defensible as same-bug or same-regression matches, and the rejected pairs avoid overstating similarity." }, "evaluator_used": true, "retried": false, @@ -1883,7 +1874,7 @@ { "label": "issue", "component_index": 1, - "component_count": 25, + "component_count": 26, "review_unit_index": 6, "review_unit_count": 61, "status": "reviewed", @@ -1944,8 +1935,8 @@ ], "soft_pairs": [ "issue:44530|issue:44863", - "issue:44863|issue:46402", "issue:41553|issue:44488", + "issue:44863|issue:46402", "issue:43366|issue:43828", "issue:45325|issue:45902", "issue:43366|issue:43531", @@ -1978,14 +1969,14 @@ "issue:42491|issue:43720", "issue:43596|issue:44661" ], - "prepared_review_unit_hash": "0fea99f1ddb23d48ed92cd5416a16218005ad7fb37b91f2fcd6080137c09c055", + "prepared_review_unit_hash": "e74c5f95845bfd863fc633d7c50b7e5f56721e755ef12dfb8d6e5280f7681009", "component_budget": { "node_count": 505, "item_count": 505, "soft_pair_count": 1854, - "serialized_chars": 962354, - "estimated_input_tokens": 240589, - "estimated_eval_tokens": 481434 + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 }, "budget": { "node_count": 48, @@ -2000,225 +1991,225 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly a collection of unrelated Transformers bug reports, but there are four clear duplicate pairs: the custom-model initialization regression, the ZeRO-3/BertModel IndexError, the SigLIP2 bad-output reports, and the Gemma4 4-bit dtype-cast reports.", - "confidence": 0.86, - "canonical_issue_reason": "Issue 43646 is the best canonical issue: it is the broadest, least user-specific wording of the custom model initialization regression and matches the clearer duplicate pair with 43645.", + "summary": "The cluster is mostly heterogeneous. Only two soft-edge pairs are accepted as true duplicates-like matches: SigLIP2 inference regression (38175/43994) and the Transformers 5 custom-model initialization regression (43645/43646). The Bert/DeepSpeed zero3 issues are not supported as a duplicate pair.", + "confidence": 0.91, + "canonical_issue_reason": "issue:43646 is the best canonical issue because it states the broad Transformers 5 custom-model initialization regression in a generic, reusable way; issue:43645 is the narrower notebook-flavored manifestation.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 43646 is the strongest representative issue in the set because it captures a general regression cleanly, whereas most other items are narrower or unrelated symptoms.", + "best_issue_reason": "issue:43646 is the strongest cluster representative: concise, generic, and centered on the underlying regression rather than one environment-specific symptom.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44530", "right": "issue:44863", "accept": false, - "reason": "Different failures on different model families (Qwen3.5 cache handling vs NemotronH checkpoint loading)." + "reason": "Different bugs: Qwen3.5 PagedAttentionCache linear_attention crash vs NemotronH checkpoint loading failure." }, { - "left": "issue:44863", - "right": "issue:46402", + "left": "issue:41553", + "right": "issue:44488", "accept": false, - "reason": "Both are loading problems, but they affect unrelated models and different code paths." + "reason": "Both are loading problems, but they concern different models and different failure modes." }, { - "left": "issue:41553", - "right": "issue:44488", + "left": "issue:44863", + "right": "issue:46402", "accept": false, - "reason": "Both mention loading errors, but the models and failure modes differ; not the same bug." + "reason": "Different model-loading regressions: NemotronH checkpoint compatibility vs Sam3Model text encoder weights not loading." }, { "left": "issue:43366", "right": "issue:43828", "accept": false, - "reason": "Unrelated issues: GGUF gpt-oss support vs Phi MoE autocast dtype mismatch." + "reason": "Unrelated issues: gpt-oss GGUF support request vs Phi-tiny-MoE autocast dtype mismatch." }, { "left": "issue:45325", "right": "issue:45902", "accept": false, - "reason": "Both involve rotary/position logic, but they describe different regressions and different models." + "reason": "Different Qwen internals and different defects: rope index scaling vs inv_freq materialization after meta->to_empty(cuda)." }, { "left": "issue:43366", "right": "issue:43531", "accept": false, - "reason": "Different Qwen/GGUF concerns; no shared concrete bug." + "reason": "Different problems: gpt-oss support request vs sliding_window bug in Qwen3-MoE." }, { "left": "issue:43596", "right": "issue:46515", "accept": false, - "reason": "ZeRO-3/Bert index error and torch 2.4.1 incompatibility are different problems." + "reason": "Bert/DeepSpeed zero3 IndexError is a different bug from a generic torch 2.4.1 incompatibility report." }, { "left": "issue:44155", "right": "issue:45584", "accept": false, - "reason": "Different audio generation bugs with different triggers and symptoms." + "reason": "Audio-related but distinct code paths: track-to-track token leak vs empty-transcription handling after align_special_tokens." }, { "left": "issue:45663", "right": "issue:45823", "accept": false, - "reason": "Both are Gemma4, but one is an FSDP2 shared_kv_states bug and the other is a device-map mismatch." + "reason": "Same model family, but different bugs: per-layer shared_kv_states rebuild under FSDP2 vs device mismatch with device_map=\"auto\"." }, { "left": "issue:43638", "right": "issue:46515", "accept": false, - "reason": "A Bert/ZeRO-3 IndexError is not the same as a generic torch version incompatibility." + "reason": "The Bert zero3 training IndexError is not the same as a generic torch version incompatibility issue." }, { "left": "issue:43366", "right": "issue:44589", "accept": false, - "reason": "Different issues: model support vs missing Float8 storage handling." + "reason": "Unrelated: gpt-oss GGUF support vs Float8_e4m3fnStorage lookup failure." }, { "left": "issue:43950", "right": "issue:45834", "accept": false, - "reason": "Non-persistent buffer loading corruption is unrelated to Kosmos2.5 long-input indexing." + "reason": "Different regressions: non-persistent buffer corruption in from_pretrained() vs Kosmos2.5 long-ocr index error." }, { "left": "issue:43824", "right": "issue:43957", "accept": false, - "reason": "ImportError for a specific class is different from meta-device loading failures." + "reason": "Different loading failures: missing Qwen2_5_VL export vs meta-device loading breakage on some models." }, { "left": "issue:42915", "right": "issue:43957", "accept": false, - "reason": "Qwen3Moe FineGrainedFP8Config failure and meta-device loading issues are different regressions." + "reason": "Different regressions: Qwen3Moe FineGrainedFP8Config failure vs meta-device model loading breakage." }, { "left": "issue:43716", "right": "issue:44560", "accept": false, - "reason": "Different multimodal model bugs with different error modes." + "reason": "Different models and failure modes: Mistral-3 image-preprocessor dtype mismatch vs Qwen3-vl-embedding video StopIteration." }, { "left": "issue:44910", "right": "issue:46858", "accept": false, - "reason": "Both are attention/cache related, but the root causes and failing paths are different." + "reason": "Both involve cache/positions, but these are distinct bugs with different triggers and backends." }, { "left": "issue:43645", "right": "issue:43646", "accept": true, - "reason": "Same Transformers 5.0.0 custom-model initialization regression; 43645 is just the Jupyter-specific manifestation." + "reason": "Same Transformers 5 custom-model initialization regression; 43645 is the notebook-specific symptom of the broader bug in 43646." }, { "left": "issue:43550", "right": "issue:44910", "accept": false, - "reason": "torch.compile+SDPA on Bamba is unrelated to Qwen3.5 flash-attention packed-sequence misinterpretation." + "reason": "Different Qwen/Bamba failures: torch.compile with SDPA on Bamba-9B-v2 vs flash-attention crash from Qwen3.5 3D position_ids handling." }, { "left": "issue:44560", "right": "issue:45685", "accept": false, - "reason": "Video StopIteration on Qwen3-vl-embedding is unrelated to the MPS histogram error in MoE." + "reason": "Different platform/model bugs: Qwen3-vl video StopIteration vs MPS histogram not implemented for Int." }, { "left": "issue:42617", "right": "issue:43366", "accept": false, - "reason": "A script launch failure is unrelated to GGUF gpt-oss support." + "reason": "3d_parallel.py execution failure is unrelated to gpt-oss GGUF architecture support." }, { "left": "issue:36010", "right": "issue:42915", "accept": false, - "reason": "GenerationMixin import issues and Qwen3Moe FP8 issues are unrelated." + "reason": "ImportError for GenerationMixin and Qwen3Moe FineGrainedFP8Config failure are unrelated bugs." }, { "left": "issue:44560", "right": "issue:44918", "accept": false, - "reason": "Different Qwen video/training failures with no shared code-path bug." + "reason": "Different Qwen3-related failures: video StopIteration vs TRL SFT trainer embedding unpacking." }, { "left": "issue:44683", "right": "issue:46515", "accept": false, - "reason": "Compiled flex_attention on torch>=2.9 is not the same as a torch 2.4.1 incompatibility." + "reason": "Both are torch-version-related, but the concrete failures are different." }, { "left": "issue:45812", "right": "issue:46489", "accept": false, - "reason": "Both are tokenizer regressions, but they affect different model families and likely different fixes." + "reason": "Both are tokenizer regressions, but they affect different model families and appear to have different root causes." }, { "left": "issue:45823", "right": "issue:46515", "accept": false, - "reason": "Gemma4 device-map mismatch is unrelated to a generic torch version incompatibility." + "reason": "Gemma4 device mismatch with device_map=\"auto\" is a different bug from a generic torch 2.4.1 incompatibility." }, { "left": "issue:44661", "right": "issue:45823", "accept": false, - "reason": "TOKENIZER_MAPPING_NAMES add-new-model-like failure is unrelated to Gemma4 PLE device placement." + "reason": "Tokenizer-mapping/add-new-model-like failure is unrelated to Gemma4 PLE device mismatch." }, { "left": "issue:38175", "right": "issue:43994", "accept": true, - "reason": "Same SigLIP2 model regression: one report sees zero probabilities, the other nonsensical outputs through AutoModel/pipeline." + "reason": "Same SigLIP2 inference regression on google/siglip2-base-patch16-224; \"zero probabilities\" and \"nonsensical results\" point to the same underlying bug." }, { "left": "issue:44560", "right": "issue:44805", "accept": false, - "reason": "Different shape/IndexError bugs in unrelated execution paths." + "reason": "Different failures: Qwen3-vl embedding video StopIteration vs a generic mask-shape IndexError." }, { "left": "issue:45672", "right": "issue:46899", - "accept": true, - "reason": "Same Gemma4 4-bit quantization dtype-cast problem in multimodal/audio embedders." + "accept": false, + "reason": "Both are Gemma4 quantization-related, but they are different concrete bugs in different modules." }, { "left": "issue:43632", "right": "issue:45834", "accept": false, - "reason": "_is_hf_initialized flag breakage is unrelated to Kosmos2.5 OCR indexing." + "reason": "_is_hf_initialized flag regression and Kosmos2.5 long OCR index error are different issues." }, { "left": "issue:44877", "right": "issue:45030", "accept": false, - "reason": "Both involve config validation, but they are different model/config regressions." + "reason": "Both are config-validation problems, but they target different models and different validation changes." }, { "left": "issue:43638", "right": "issue:44661", "accept": false, - "reason": "ZeRO-3 Bert initialization and add-new-model-like token mapping failures are unrelated." + "reason": "Bert zero3 IndexError and tokenizer-mapping add-new-model-like failure are unrelated defects." }, { "left": "issue:42491", "right": "issue:43720", "accept": false, - "reason": "Different bugs: Qwen3 MoE LoRA compatibility vs BitNet packed-weight loading." + "reason": "Different subsystems and failure modes: Qwen3_MOE LoRA incompatibility across hf4.x/hf5.x vs BitNet packed-weight loading in accelerate." }, { "left": "issue:43596", - "right": "issue:43638", - "accept": true, - "reason": "Same deepspeed ZeRO-3/BertModel IndexError; the pretrained vs non-pretrained distinction does not change the underlying bug." + "right": "issue:44661", + "accept": false, + "reason": "Different bugs: Bert zero3 IndexError vs tokenizer-mapping/add-new-model-like failure." } ] }, "evaluator_result": { - "accept": true, - "feedback": "The summary is grounded in the packet, and the soft-edge verdicts stay conservative overall. The four accepted duplicate pairs are supported by closely aligned titles and the pair scores, and the rejected pairs generally avoid overclaiming by distinguishing different models or code paths." + "accept": false, + "feedback": "The analysis is mostly careful on the soft-edge verdicts, but the summary overstates the evidence by claiming \u201cthree true duplicate-like issue pairs.\u201d In the packet, only two pairs are actually accepted as duplicates (38175/43994 and 43645/43646); the Bert/DeepSpeed zero3 issues are not supported as a true duplicate pair. Please keep the cluster summary tighter and grounded in the explicit soft-edge outcomes." }, "evaluator_used": true, - "retried": false, + "retried": true, "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null @@ -2226,7 +2217,7 @@ { "label": "issue", "component_index": 1, - "component_count": 25, + "component_count": 26, "review_unit_index": 7, "review_unit_count": 61, "status": "reviewed", @@ -2312,25 +2303,25 @@ "issue:44960|issue:46050", "issue:44164|issue:44291", "issue:46897|issue:46901", - "issue:45488|issue:45812", "issue:45362|issue:45406", + "issue:45488|issue:45812", "issue:45276|issue:45335", "issue:44910|issue:45588" ], - "prepared_review_unit_hash": "d73ad4761ffb442856624da3995bbd4b478166c8034b2d3141aa6c5a0e07547b", + "prepared_review_unit_hash": "a6d6ec9bbb8a7d90ec4c081687246d952bb5285ff923b1be74ac67d4e68f1ddc", "component_budget": { "node_count": 505, "item_count": 505, "soft_pair_count": 1854, - "serialized_chars": 962354, - "estimated_input_tokens": 240589, - "estimated_eval_tokens": 481434 + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 30, - "serialized_chars": 36656, + "serialized_chars": 36654, "estimated_input_tokens": 9164, "estimated_eval_tokens": 18584 }, @@ -2339,201 +2330,201 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly heterogeneous. The only accepted soft duplicate pairs are the Transformers v5 non-persistent-buffer corruption reports and the resize_token_embeddings propagation bug across model-specific embedding/output layers. The rest share only broad model or subsystem themes, not the same concrete defect.", - "confidence": 0.82, - "canonical_issue_reason": "Issue 43644 is the clearest representative of the v5 non-persistent-buffer corruption reports and is very close in wording and symptom to 44534.", + "summary": "Most items are distinct model- or subsystem-specific regressions. I only accept a few soft duplicates where the titles point to the same concrete bug: non-persistent buffers being junk-filled in v5, resize_token_embeddings not propagating to output embeddings in related model variants, and the training-loss double-shift bug.", + "confidence": 0.9, + "canonical_issue_reason": "issue:43644 is the clearest, most concrete regression report in the set and has the exact duplicate-style paraphrase in issue:44534.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 43644 is the best overall issue representative here because it names the v5 buffer corruption symptom directly and has the strongest linkage to a matching follow-up report.", + "best_issue_reason": "issue:43644 is the best representative issue here because it states the bug precisely and matches the closest duplicate candidate without ambiguity.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43317", "right": "issue:44756", "accept": false, - "reason": "Both involve loading/offload behavior, but one is a device_map offload failure and the other is a Strix Halo mmap/OOM issue." + "reason": "Different failures: device_map/offload loading error vs mmap/OOM mitigation on Strix Halo." }, { "left": "issue:44661", "right": "issue:44805", "accept": false, - "reason": "Different bugs: tokenizer-mapping setup failure versus a mask/tensor shape mismatch." + "reason": "Unrelated bugs: tokenizer mapping/model addition logic vs mask shape mismatch at runtime." }, { "left": "issue:38175", "right": "issue:39692", "accept": false, - "reason": "Same model family, but the symptoms differ: zero probabilities versus docs/example model-processor mismatch and quantization failure." + "reason": "Same model family, but one is zero probabilities and the other is doc-example/model-processor/quantization errors; not the same concrete bug." }, { "left": "issue:43644", "right": "issue:44534", "accept": true, - "reason": "Same Transformers v5 non-persistent-buffer corruption bug; the titles describe the same symptom with slightly different wording." + "reason": "Same regression: both report Transformers v5 filling non-persistent buffers with junk." }, { "left": "issue:41720", "right": "issue:42915", "accept": false, - "reason": "Different Qwen3 failures: auto device-mapping CUDA assert versus FineGrainedFP8Config load failure." + "reason": "Different Qwen3 issues with different failure modes and code paths." }, { "left": "issue:42915", "right": "issue:44756", "accept": false, - "reason": "Unrelated issues: Qwen3 MoE FP8 config handling versus mmap/OOM on Strix Halo." + "reason": "Unrelated: Qwen3 MoE FP8 failure vs mmap/OOM on Strix Halo." }, { "left": "issue:42491", "right": "issue:45276", "accept": false, - "reason": "Both are Qwen3-related, but one is a LoRA compatibility problem and the other is a resize_token_embeddings regression." + "reason": "Different problems: LoRA compatibility across versions vs resize_token_embeddings propagation." }, { "left": "issue:44960", "right": "issue:44977", "accept": false, - "reason": "Different model and failure mode; no clear shared concrete bug." + "reason": "Both are GLM/Qwen-related, but the reported failures are different model/runtime issues." }, { "left": "issue:44451", "right": "issue:44488", "accept": false, - "reason": "Same general loading theme, but they affect different models and do not show one shared defect." + "reason": "Both are model-loading complaints, but they concern different checkpoints and no shared concrete root cause is evident." }, { "left": "issue:44863", "right": "issue:46846", "accept": false, - "reason": "NemotronH checkpoint loading versus missing linear_attn entries in tensor-parallel planning are different problems." + "reason": "Different models and different failures: NemotronH checkpoint loading vs missing linear_attn TP plan entries." }, { "left": "issue:46459", "right": "issue:46512", "accept": false, - "reason": "Both involve dtype, but one is composite-checkpoint loading and the other is from_config leaving config.dtype stale." + "reason": "Related around dtype handling, but one is loading composite checkpoints and the other is from_config/config-state mismatch." }, { "left": "issue:45313", "right": "issue:45357", "accept": false, - "reason": "Different code paths: ZeRO-3 weight loading versus save_pretrained writing incorrect visual encoder keys." + "reason": "Different bugs: DeepSpeed ZeRO-3 loading vs incorrect visual encoder keys on save_pretrained." }, { "left": "issue:45507", "right": "issue:46468", "accept": false, - "reason": "Different model/example breakages with no evidence of the same method-call bug." + "reason": "Different model implementations and different invalid-method/example failures." }, { "left": "issue:44488", "right": "issue:44991", "accept": false, - "reason": "Both are tokenizer/model-loading regressions, but they appear to be different tokenizer issues for different models." + "reason": "Both are tokenizer/model loading issues, but they affect different repos and no same underlying bug is shown." }, { "left": "issue:43644", "right": "issue:43645", "accept": false, - "reason": "Non-persistent-buffer corruption is unrelated to the Jupyter/custom-model initialization regression." + "reason": "One is the non-persistent-buffer junk regression; the other is Jupyter custom-model initialization, a different issue." }, { "left": "issue:43257", "right": "issue:43854", "accept": false, - "reason": "Both are loading issues, but one is accelerate+deepspeed conversion and the other is a unit-test model-loading failure." + "reason": "Different loading paths and failure symptoms: accelerate+deepspeed MoE weight conversion vs unit-test model loading." }, { "left": "issue:39692", "right": "issue:43994", "accept": false, - "reason": "Related SigLIP2 area, but the reported failures differ and do not clearly point to the same bug." + "reason": "Same SigLIP2 family, but doc-example mistakes are not the same as the nonsensical-output regression." }, { "left": "issue:43856", "right": "issue:43873", "accept": false, - "reason": "Training memory inefficiency and quantization/offload behavior are not the same defect." + "reason": "Both mention memory/offloading around Qwen3 MoE, but they are different problems with different concrete causes." }, { "left": "issue:43854", "right": "issue:43931", "accept": false, - "reason": "Different load failures: GLM-4.7-Flash unit-test loading versus Qwen3-VL weight-shape mismatch." + "reason": "Different models and errors: GLM-4.7 test loading vs Qwen3-VL shape mismatch." }, { "left": "issue:42491", "right": "issue:45292", "accept": false, - "reason": "Shared resize wording is not enough; these appear to be different bugs." + "reason": "Different underlying bugs; only loosely related by model state/embedding behavior." }, { "left": "issue:43856", "right": "issue:44387", "accept": false, - "reason": "Both mention memory, but one is Qwen3 MoE training efficiency and the other is increased reserved CUDA memory under int4 quantization." + "reason": "Different memory issues: training efficiency vs int4 reserved-memory/OOM regression." }, { "left": "issue:43708", "right": "issue:44075", "accept": false, - "reason": "Trainer checkpoint-resume step counting and SGD argument handling are unrelated." + "reason": "Trainer checkpoint-resume math bug is unrelated to SGD optimizer args not being used." }, { "left": "issue:42491", "right": "issue:45335", "accept": false, - "reason": "Different model-specific regressions; not enough evidence that they are the same issue." + "reason": "LoRA/version compatibility issue is unrelated to resize_token_embeddings behavior." }, { "left": "issue:44960", "right": "issue:46050", "accept": false, - "reason": "Different models and different save/load behavior; no shared concrete bug." + "reason": "Different bugs: GLM5 model issue vs quantized model save/load corruption." }, { "left": "issue:44164", "right": "issue:44291", "accept": false, - "reason": "Different serialization/load paths: extra_state handling versus unexpected _is_hf_initialized argument." + "reason": "Different save/load failures: extra_state handling vs unexpected _is_hf_initialized argument under init_empty_weights." }, { "left": "issue:46897", "right": "issue:46901", - "accept": false, - "reason": "They share a loss-shift pattern, but the model-specific implementations are different and not clearly one shared bug." + "accept": true, + "reason": "Same bug pattern: training loss is being double-shifted instead of using the labels as-is." }, { - "left": "issue:45488", - "right": "issue:45812", + "left": "issue:45362", + "right": "issue:45406", "accept": false, - "reason": "Both are tokenizer regressions, but they affect different tokenizer classes and model families." + "reason": "Different components and failures: chat crash for Qwen3.5-35B vs Gemma4Processor missing _tokenizer in serve." }, { - "left": "issue:45362", - "right": "issue:45406", + "left": "issue:45488", + "right": "issue:45812", "accept": false, - "reason": "Qwen3.5 chat crash and Gemma4 processor crash are unrelated failures." + "reason": "Both are tokenizer regressions, but the affected families and failure mechanisms differ." }, { "left": "issue:45276", "right": "issue:45335", "accept": true, - "reason": "Likely the same resize_token_embeddings propagation bug: model-specific embedding/output layers are not updated consistently." + "reason": "Same concrete bug: resize_token_embeddings does not propagate updated embeddings to output/decoder embeddings across related model variants." }, { "left": "issue:44910", "right": "issue:45588", "accept": false, - "reason": "Both involve flash attention, but the concrete crash causes are different." + "reason": "Both involve flash attention, but the crashes come from different triggers and code paths." } ] }, "evaluator_result": { - "accept": false, - "feedback": "The cluster summary is broadly aligned with the packet, and the two accepted soft pairs are plausible. However, the writeup overstates evidence in a few places: \"precise, reproducible\" is not supported by the packet, \"exact duplicate counterpart\" is stronger than the available title-level evidence, and \"same underlying API regression\" for the resize_token_embeddings pair should be phrased more cautiously as a likely/shared bug. Because precision is the goal, this should be tightened before accepting." + "accept": true, + "feedback": "The summary is grounded in the packet and the conservative rejections are mostly well supported. The accepted soft duplicates also look plausibly mergeable/same-bug from the titles: non-persistent buffer junk in v5, resize_token_embeddings propagation, and the training-loss double-shift pattern. No clear overstatement stands out." }, "evaluator_used": true, - "retried": true, + "retried": false, "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null @@ -2541,7 +2532,7 @@ { "label": "issue", "component_index": 1, - "component_count": 25, + "component_count": 26, "review_unit_index": 8, "review_unit_count": 61, "status": "reviewed", @@ -2565,6 +2556,7 @@ "issue:43645", "issue:43646", "issue:43708", + "issue:43828", "issue:43854", "issue:43873", "issue:43957", @@ -2575,7 +2567,6 @@ "issue:44292", "issue:44387", "issue:44402", - "issue:44451", "issue:44479", "issue:44488", "issue:44756", @@ -2631,162 +2622,162 @@ "issue:45292|issue:45335", "issue:43645|issue:44164", "issue:43646|issue:43708", - "issue:44402|issue:44451" + "issue:43828|issue:43957" ], - "prepared_review_unit_hash": "18e6c018e55714104174f9ae6b7fa90a3aa00a4adbacd45310ab1a4eeb9cad3b", + "prepared_review_unit_hash": "fa648452b1ce020897d1f1356ae4d7ae281ace42b515367be977d381644b4292", "component_budget": { "node_count": 505, "item_count": 505, "soft_pair_count": 1854, - "serialized_chars": 962354, - "estimated_input_tokens": 240589, - "estimated_eval_tokens": 481434 + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 31, - "serialized_chars": 37061, - "estimated_input_tokens": 9266, - "estimated_eval_tokens": 18788 + "serialized_chars": 37088, + "estimated_input_tokens": 9272, + "estimated_eval_tokens": 18800 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly a grab-bag of unrelated Transformers bug reports. The only clear duplicate subgroup is the `resize_token_embeddings` regression around output embeddings, where a broad generic report and two model-specific variants overlap.", - "confidence": 0.84, - "canonical_issue_reason": "issue:45292 is the broadest representative of the only real duplicate subgroup here: `resize_token_embeddings` not updating output embeddings / related embedding modules.", + "summary": "This cluster is mostly a grab-bag of unrelated Transformers bugs across different models and subsystems. The only clearly duplicate-like family is the resize_token_embeddings/output_embeddings reports, with the generic issue serving as the best representative. No PRs are present.", + "confidence": 0.31, + "canonical_issue_reason": "Issue 45292 is the most generic and reusable description of the resize_token_embeddings bug, and it best represents the close duplicate-like Gemma4/output_embeddings reports.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45292 is the best issue to anchor the embedding-resize subgroup because it is the most generic phrasing and can absorb the Gemma4/T5Gemma variants.", + "best_issue_reason": "Issue 45292 has the broadest framing of the underlying symptom and would be the best standalone tracker for the resize_token_embeddings family.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44402", "right": "issue:44488", "accept": false, - "reason": "Both are tokenizer/model loading problems, but the specific failure modes and affected models differ; not the same underlying bug." + "reason": "Different problems: token vocab mismatch vs model loading failure for a specific checkpoint." }, { "left": "issue:39692", "right": "issue:44960", "accept": false, - "reason": "Unrelated topics: doc/example errors versus a GLM5 model issue." + "reason": "Unrelated model/docs issue versus a GLM5 loading/title-only issue." }, { "left": "issue:41720", "right": "issue:44155", "accept": false, - "reason": "Different subsystems and symptoms: Qwen device-mapping crash vs AudioFlamingo batched inference leakage." + "reason": "Different code paths and symptoms: Qwen3 device mapping CUDA assert vs AudioFlamingo batched inference leak." }, { "left": "issue:43425", "right": "issue:43957", "accept": false, - "reason": "Both involve model loading/runtime compatibility, but one is a Torch version incompatibility and the other is a `meta` loading regression." + "reason": "Torch version compatibility and meta-device loading failures are separate issues." }, { "left": "issue:45663", "right": "issue:45759", "accept": false, - "reason": "Different code paths: Gemma4 FSDP attention state handling vs multimodal Qwen config unwrapping." + "reason": "Different model families and bugs: Gemma4 FSDP2 attention state vs Qwen multimodal config unwrapping." }, { "left": "issue:42947", "right": "issue:45704", "accept": false, - "reason": "Gradient checkpointing/LoRA behavior is unrelated to the T5 apex RMSNorm memory leak." + "reason": "Gradient checkpointing/LoRA behavior is unrelated to T5 apex RMSNorm memory leak." }, { "left": "issue:43404", "right": "issue:43454", "accept": false, - "reason": "Similar theme of weight tying, but these are different model classes and likely separate implementation bugs." + "reason": "Both mention lm_head tying, but they are different model implementations and likely separate fixes." }, { "left": "issue:44291", "right": "issue:44756", "accept": false, - "reason": "`init_empty_weights` argument handling is unrelated to the Strix Halo mmap OOM workaround." + "reason": "init_empty_weights argument handling and Strix Halo mmap/OOM are unrelated." }, { "left": "issue:43873", "right": "issue:44387", "accept": false, - "reason": "Both touch quantization memory, but one is offloading behavior and the other is increased reserved memory under int4; not the same concrete bug." + "reason": "Both involve quantization/memory, but the concrete failure modes differ too much to treat as one bug." }, { "left": "issue:44079", "right": "issue:46829", "accept": false, - "reason": "Both are output-structure issues, but they affect different output classes and symptom paths." + "reason": "ModelOutput key assignment and VideoPrism output fields are different issues." }, { "left": "issue:45663", "right": "issue:45859", "accept": false, - "reason": "Gemma4 FSDP attention state handling is unrelated to Qwen3_5Moe rotary embedding CPU offload." + "reason": "Different bugs in different models: Gemma4 FSDP2 vs Qwen3_5MoE CPU offload incompatibility." }, { "left": "issue:44292", "right": "issue:44912", "accept": false, - "reason": "Separate quantization failures on different models; not enough evidence of one shared bug." + "reason": "Different quantization formats and failure modes: NVFP4 load error vs MXFP4 fallback to bf16." }, { "left": "issue:36010", "right": "issue:39404", "accept": false, - "reason": "Import path breakage and Whisper pipeline behavior are unrelated." + "reason": "ImportError for GenerationMixin is unrelated to Whisper return_language pipeline regression." }, { "left": "issue:44155", "right": "issue:45823", "accept": false, - "reason": "Different model families and failure modes: token leakage in batched inference vs device mismatch in forward." + "reason": "AudioFlamingo batched inference leak and Gemma4 device_map mismatch are separate problems." }, { "left": "issue:44387", "right": "issue:44756", "accept": false, - "reason": "Both mention OOM, but one is int4 quantization memory growth and the other is mmap on Strix Halo; distinct causes." + "reason": "Both can end in OOM, but one is int4 quantization memory growth and the other is Strix Halo mmap behavior." }, { "left": "issue:42947", "right": "issue:43381", "accept": false, - "reason": "Both mention gradient checkpointing, but one is ineffective under PEFT LoRA while the other is eval-mode incompatibility." + "reason": "Both involve gradient checkpointing, but one is ineffective under PEFT LoRA and the other is invalid in eval mode." }, { "left": "issue:43957", "right": "issue:44756", "accept": false, - "reason": "`torch.device('meta')` loading regressions and Strix Halo mmap OOM are different issues." + "reason": "Meta-device loading and mmap-related OOM on Strix Halo are unrelated." }, { "left": "issue:43335", "right": "issue:45663", "accept": false, - "reason": "Single-layer sparse config bug is unrelated to Gemma4 FSDP attention state rebuilding." + "reason": "SwitchTransformers sparse layer construction and Gemma4 FSDP2 attention state are unrelated." }, { "left": "issue:44863", "right": "issue:46468", "accept": false, - "reason": "Nemotron checkpoint loading and the official Nemotron example being broken are related in theme but not clearly the same bug." + "reason": "NemotronH checkpoint loading and a broken official Nemotron example are not the same concrete bug." }, { "left": "issue:45276", "right": "issue:45292", "accept": true, - "reason": "Both report `resize_token_embeddings` failing to update output embeddings; 45276 is a Gemma4-specific manifestation of the broader bug in 45292." + "reason": "These are effectively the same resize_token_embeddings/output_embeddings bug, with 45276 being a model-specific variant of the generic issue." }, { "left": "issue:44479", "right": "issue:44877", "accept": false, - "reason": "Different model/config regressions: video input handling vs strict config loading for granite_speech." + "reason": "Different model/regression contexts: video input regression vs strict config blocking granite_speech loading." }, { "left": "issue:43381", @@ -2798,61 +2789,61 @@ "left": "issue:44164", "right": "issue:44756", "accept": false, - "reason": "Save/load `extra_state` handling is unrelated to disabling mmap for OOM." + "reason": "save/from_pretrained extra_state handling and Strix Halo mmap/OOM are unrelated." }, { "left": "issue:42796", "right": "issue:46326", "accept": false, - "reason": "Tokenizer BPE marker handling and `continue_final_message` are different tokenizer bugs." + "reason": "Tokenizer BPE-marker handling and continue_final_message behavior are different tokenizer bugs." }, { "left": "issue:45588", "right": "issue:45663", "accept": false, - "reason": "Flash-attention sink-less crash and Gemma4 FSDP attention state rebuilding are separate code paths." + "reason": "flash_attention sink-less crash and Gemma4 FSDP2 shared_kv state are unrelated." }, { "left": "issue:45684", "right": "issue:45698", "accept": false, - "reason": "Custom-model save permissions and wrong-module loading after save are both save/load related but distinct failures." + "reason": "Permissions propagation during save_pretrained and loading the wrong custom module are separate save/load bugs." }, { "left": "issue:43854", "right": "issue:45216", "accept": false, - "reason": "Different regressions: unit-test loading of GLM-4.7-Flash vs Qwen3.5 checkpoint saving." + "reason": "Model load failure in tests and Qwen3.5 save_pretrained regression do not match." }, { "left": "issue:45292", "right": "issue:45335", - "accept": true, - "reason": "Both are the same `resize_token_embeddings` family of bugs: 45292 covers output embeddings, and 45335 shows the decoder embedding not being updated." + "accept": false, + "reason": "They are related in theme, but 45335 is a T5Gemma-specific decoder.embed_tokens symptom and not clearly the same concrete bug as the generic output_embeddings report." }, { "left": "issue:43645", "right": "issue:44164", "accept": false, - "reason": "Custom-model Jupyter init breakage is unrelated to `extra_state` save/load handling." + "reason": "Custom model initialization in Jupyter and extra_state save/load handling are separate problems." }, { "left": "issue:43646", "right": "issue:43708", "accept": false, - "reason": "Custom model initialization under Transformers 5 and trainer resume/max_steps calculation are unrelated." + "reason": "Transformers 5.0 custom model init breakage and Trainer max_steps recalculation are unrelated." }, { - "left": "issue:44402", - "right": "issue:44451", + "left": "issue:43828", + "right": "issue:43957", "accept": false, - "reason": "Different model-loading/tokenizer problems; same broad area but not the same concrete bug." + "reason": "Autocast dtype mismatch in Phi-tiny-MoE and meta-device loading failures are different issues." } ] }, "evaluator_result": { "accept": true, - "feedback": "The analysis is grounded in the packet and stays conservative on the soft edges. The resize_token_embeddings subgroup is the only clearly supported duplicate cluster, and the accepted links (45276\u219445292, 45292\u219445335) are well justified by the titles and high similarity. The rest of the pair rejections do not overstate the evidence." + "feedback": "The summary is grounded in the packet and stays conservative. The only clearly related family identified is the resize_token_embeddings/output_embeddings cluster, and the accepted 45276\u201345292 soft-edge verdict is supported by the titles and high similarity. No PR-pair mergeability issues arise because there are no PRs in the packet." }, "evaluator_used": true, "retried": false, @@ -2863,7 +2854,7 @@ { "label": "issue", "component_index": 1, - "component_count": 25, + "component_count": 26, "review_unit_index": 9, "review_unit_count": 61, "status": "reviewed", @@ -2878,7 +2869,6 @@ "issue:43299", "issue:43425", "issue:43452", - "issue:43596", "issue:43611", "issue:43645", "issue:43716", @@ -2887,11 +2877,12 @@ "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", @@ -2923,11 +2914,11 @@ "issue:46901" ], "soft_pairs": [ - "issue:43828|issue:43957", + "issue:44402|issue:44451", "issue:46894|issue:46901", "issue:44589|issue:44756", - "issue:46315|issue:46491", "issue:43425|issue:43828", + "issue:46315|issue:46491", "issue:45812|issue:46315", "issue:39404|issue:45701", "issue:44530|issue:45084", @@ -2955,241 +2946,234 @@ "issue:45479|issue:46894", "issue:43856|issue:44756", "issue:44164|issue:46821", - "issue:45479|issue:46897", - "issue:43596|issue:44683" + "issue:45479|issue:46897" ], - "prepared_review_unit_hash": "af35413b0b53d90f4c02e7d1d4989b03b9cc35bed98074983582bb267780aacd", + "prepared_review_unit_hash": "a5cb1f0edeeab89741f5362db5ab851149cc2a6f802923f38e3530ead78d7117", "component_budget": { "node_count": 505, "item_count": 505, "soft_pair_count": 1854, - "serialized_chars": 962354, - "estimated_input_tokens": 240589, - "estimated_eval_tokens": 481434 + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 34, - "serialized_chars": 38131, - "estimated_input_tokens": 9533, - "estimated_eval_tokens": 19322 + "soft_pair_count": 33, + "serialized_chars": 37758, + "estimated_input_tokens": 9440, + "estimated_eval_tokens": 19136 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is heterogeneous: several items are broad v5 loading/tokenizer regressions, but most pairs describe different code paths or model families. I did not find any soft-edge pair that is safe to merge as a duplicate.", - "confidence": 0.69, - "canonical_issue_reason": "issue:43611 is the broadest, most general v5 loading-regression report in the set, so it is the least-specific representative of the cluster.", + "summary": "The candidates are mostly broad, thematically similar regressions, but they do not look like the same concrete bug. I would keep them separate; the only close pair is the two training-loss double-shift reports, but they still appear model-specific rather than one shared code-path fix.", + "confidence": 0.82, + "canonical_issue_reason": "issue:46901 is the most concrete, technically actionable issue in the set: it states the exact label-shift mistake (`labels[..., 1:]` vs `labels`).", "canonical_pr_reason": null, - "best_issue_reason": "issue:43611 is the best single representative of the overall loading-regression theme, though the cluster is still too mixed to treat as one duplicate bug.", + "best_issue_reason": "issue:46901 is the best representative issue because it is precise, reproducible, and directly points to the faulty loss computation; the other nearby reports are broader or more model-specific.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:43828", - "right": "issue:43957", + "left": "issue:44402", + "right": "issue:44451", "accept": false, - "reason": "Different bugs: autocast dtype mismatch at runtime vs meta-device model loading failure." + "reason": "Both are tokenizer-loading problems, but one is a vocab-size metadata bug and the other is a specific model load failure; no evidence they share the same root cause." }, { "left": "issue:46894", "right": "issue:46901", "accept": false, - "reason": "Same symptom pattern, but different model-specific training-loss implementations; not the same code path." + "reason": "They describe the same symptom family, but on different model families with no proof of a shared code path; treat as related reports, not one duplicate bug." }, { "left": "issue:44589", "right": "issue:44756", "accept": false, - "reason": "Unrelated failures: missing Float8 storage vs mmap/OOM behavior on Strix Halo." + "reason": "Completely different failures: missing Float8 storage vs mmap/OOM behavior on Strix Halo." }, { - "left": "issue:46315", - "right": "issue:46491", + "left": "issue:43425", + "right": "issue:43828", "accept": false, - "reason": "Both tokenizer-related, but one is a slowdown in convert_tokens_to_ids and the other is a round-trip correctness regression." + "reason": "One is a broad Torch 2.10 compatibility complaint, the other is a specific autocast dtype mismatch in Phi-tiny-MoE." }, { - "left": "issue:43425", - "right": "issue:43828", + "left": "issue:46315", + "right": "issue:46491", "accept": false, - "reason": "Torch version compatibility vs a multimodal autocast dtype mismatch are different problems." + "reason": "Both touch tokenization, but one is a slowdown regression and the other is a round-trip correctness bug; different underlying issues." }, { "left": "issue:45812", "right": "issue:46315", "accept": false, - "reason": "Wrong token IDs for Granite models is a correctness bug, not the same as a slow conversion path." + "reason": "Granite wrong token IDs and slow `convert_tokens_to_ids` are distinct tokenization regressions." }, { "left": "issue:39404", "right": "issue:45701", "accept": false, - "reason": "Whisper pipeline return_language regression is unrelated to a generic tokenization-change report." + "reason": "Whisper pipeline `return_language` and a generic tokenization change are unrelated behaviors." }, { "left": "issue:44530", "right": "issue:45084", "accept": false, - "reason": "Different failures in different subsystems: PagedAttentionCache crash vs template-node compilation error." + "reason": "Different problems: Qwen3.5 cache group-type crash vs a template compilation TypeError." }, { "left": "issue:43611", "right": "issue:43645", "accept": false, - "reason": "Both concern v5 model loading, but one is base_model_prefix handling and the other is notebook/custom-model initialization." + "reason": "Both are v5 load/init regressions, but one is about `base_model_prefix` loading and the other about custom models in Jupyter; not the same bug." }, { "left": "issue:43425", "right": "issue:46869", "accept": false, - "reason": "Torch incompatibility and YOLO multi-GPU device_map behavior are not the same underlying bug." + "reason": "Torch version incompatibility and multi-GPU `device_map=\"auto\"` failure are unrelated." }, { "left": "issue:39404", "right": "issue:46665", "accept": false, - "reason": "Whisper pipeline behavior and from_pretrained inside Python classes are distinct issues." + "reason": "Whisper pipeline regression is unrelated to inconsistent `from_pretrained` behavior inside Python classes." }, { "left": "issue:44683", "right": "issue:45823", "accept": false, - "reason": "Compiled flex_attention on newer torch and Gemma4 device mismatch are unrelated code paths." + "reason": "Compiled flex_attention on torch >= 2.9 and Gemma4 device-mapping mismatch are different code-path failures." }, { "left": "issue:43425", "right": "issue:43716", "accept": false, - "reason": "General Torch compatibility and Mistral-3 image-preprocessor dtype mismatch are different bugs." + "reason": "A Torch compatibility complaint is not the same as an image preprocessor/model dtype mismatch." }, { "left": "issue:43425", "right": "issue:45823", "accept": false, - "reason": "Torch compatibility issue vs a forward-pass device mismatch in Gemma4 PLE." + "reason": "These are different regressions in different subsystems; shared dependency version alone is not enough." }, { "left": "issue:44479", "right": "issue:44805", "accept": false, - "reason": "Different regressions: video-input handling for Qwen VL models vs a generic mask-shape IndexError." + "reason": "Both are runtime errors, but one is a Qwen video-input regression and the other is a mask shape mismatch; no shared root cause is evident." }, { "left": "issue:44589", "right": "issue:45084", "accept": false, - "reason": "Different failure modes in unrelated parts of the stack." + "reason": "Unrelated failures: missing Float8 storage vs Mistral tokenizer regex patch compilation." }, { "left": "issue:43906", "right": "issue:44530", "accept": false, - "reason": "An isolated reproduction of one issue is not the same bug as the PagedAttentionCache crash." + "reason": "A reproduction of an earlier issue and a Qwen3.5 cache crash are not the same bug." }, { "left": "issue:43901", "right": "issue:44509", "accept": false, - "reason": "Both are docs issues, but they refer to different removed/deprecated APIs and different documentation sections." + "reason": "Both concern docs, but they document different removed/changed pipeline behaviors; not a duplicate." }, { "left": "issue:43994", "right": "issue:44079", "accept": false, - "reason": "Wrong/nonsensical model outputs are not the same as ModelOutput key-assignment behavior." + "reason": "SigLIP2 nonsensical outputs and `ModelOutput` key assignment are unrelated." }, { "left": "issue:36032", "right": "issue:43452", "accept": false, - "reason": "Tokenizer add_special_tokens conflict and gguf_file loading failures are separate loading bugs." + "reason": "Tokenizer method-name conflict and gguf loading breakage are different failures." }, { "left": "issue:44265", "right": "issue:44898", "accept": false, - "reason": "torch.export failure due to torch_compilable_check is unrelated to Perceiver interpolation handling." + "reason": "`torch.export.export` failing under `torch_compilable_check` is not the same as Perceiver non-default-resolution failure." }, { "left": "issue:39404", "right": "issue:44912", "accept": false, - "reason": "Whisper pipeline regression and MXFP4 quantization loading are different model-specific problems." + "reason": "Whisper pipeline behavior and MXFP4 quantization fallback are unrelated." }, { "left": "issue:46050", "right": "issue:46326", "accept": false, - "reason": "Quantized save/load breakage and mistral_common continue_final_message behavior are unrelated." + "reason": "Quantized-model save/load corruption and a tokenizer `continue_final_message` bug do not share a concrete code path." }, { "left": "issue:44164", "right": "issue:45759", "accept": false, - "reason": "extra_state save/load handling and text_config unwrapping are different model-loading mechanisms." + "reason": "Extra state handling in save/from_pretrained is unrelated to `text_config` unwrapping for multimodal configs." }, { "left": "issue:43906", "right": "issue:46326", "accept": false, - "reason": "A reproduction issue and a tokenizer message-continuation bug are not the same underlying defect." + "reason": "The earlier issue reproduction and the mistral_common tokenizer bug are distinct." }, { "left": "issue:44671", "right": "issue:44991", "accept": false, - "reason": "CamemBERT masked-LM predictions and EMBEDDIA tokenizer loading failures are different tokenizer/model bugs." + "reason": "CamemBERT masked-LM prediction drift and an Estonian RoBERTa tokenizer load failure are different issues." }, { "left": "issue:44991", "right": "issue:45081", "accept": false, - "reason": "Tokenizer loading failure and a Mistral regex patch crash are distinct issues." + "reason": "Both involve tokenizer loading, but one is a model-specific tokenizer failure and the other is a Mistral regex-patching crash." }, { "left": "issue:45823", "right": "issue:46899", "accept": false, - "reason": "Device-map forward mismatch vs unsafe dtype cast in multimodal embedders are different bugs." + "reason": "Same model family, but one is a device-map mismatch and the other is a 4-bit quantization dtype-cast bug." }, { "left": "issue:43299", "right": "issue:43931", "accept": false, - "reason": "Both involve Qwen3-VL family loading, but one is a broad v5 regression and the other is a specific weight-shape mismatch." + "reason": "Both involve Qwen3-VL loading, but the reported failure modes differ: dev version breakage vs weight-shape mismatch." }, { "left": "issue:45479", "right": "issue:46894", "accept": false, - "reason": "Different loss bugs in different model families; not a shared code path." + "reason": "A degenerate single-label classification loss bug is unrelated to the ASR training-loss double-shift report." }, { "left": "issue:43856", "right": "issue:44756", "accept": false, - "reason": "Training memory usage and Strix Halo mmap/OOM are unrelated." + "reason": "Qwen3 MoE training memory usage and Strix Halo mmap OOM are different problems." }, { "left": "issue:44164", "right": "issue:46821", "accept": false, - "reason": "extra_state handling in save/from_pretrained is unrelated to TimesFM window_size attribute errors." + "reason": "Saving/loading extra_state and a TimesFM `window_size` AttributeError are unrelated." }, { "left": "issue:45479", "right": "issue:46897", "accept": false, - "reason": "Different model families and different loss implementations, despite both being double-shift reports." - }, - { - "left": "issue:43596", - "right": "issue:44683", - "accept": false, - "reason": "Deepspeed zero3/BertModel indexing failure is unrelated to compiled flex_attention on newer torch." + "reason": "Sequence-classification zero-loss behavior and Florence2 training-loss double-shift are different bugs, even if both concern loss computation." } ] }, @@ -3203,7 +3187,7 @@ { "label": "issue", "component_index": 1, - "component_count": 25, + "component_count": 26, "review_unit_index": 10, "review_unit_count": 61, "status": "reviewed", @@ -3228,6 +3212,7 @@ "issue:43482", "issue:43541", "issue:43582", + "issue:43596", "issue:43611", "issue:43632", "issue:43643", @@ -3262,6 +3247,7 @@ "issue:46489" ], "soft_pairs": [ + "issue:43596|issue:44683", "issue:44488|issue:44987", "issue:44488|issue:45081", "issue:43611|issue:43644", @@ -3274,8 +3260,8 @@ "issue:44737|issue:45859", "issue:43643|issue:44530", "issue:44661|issue:44683", - "issue:40990|issue:43421", "issue:42915|issue:44530", + "issue:40990|issue:43421", "issue:44451|issue:44987", "issue:43541|issue:45084", "issue:34567|issue:40444", @@ -3293,228 +3279,237 @@ "issue:43883|issue:44488", "issue:43746|issue:46326" ], - "prepared_review_unit_hash": "172267466cf93b6772be53c62897409b7e57e84fdb8d4a08e0c768ea2e2aa5c9", + "prepared_review_unit_hash": "bcf95be2e34c28ba27587ebc250e9ff99173dfc7dd6c599c70b46b16100cc974", "component_budget": { "node_count": 505, "item_count": 505, "soft_pair_count": 1854, - "serialized_chars": 962354, - "estimated_input_tokens": 240589, - "estimated_eval_tokens": 481434 + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 }, "budget": { - "node_count": 47, - "item_count": 47, - "soft_pair_count": 30, - "serialized_chars": 36111, - "estimated_input_tokens": 9028, - "estimated_eval_tokens": 18312 + "node_count": 48, + "item_count": 48, + "soft_pair_count": 31, + "serialized_chars": 37015, + "estimated_input_tokens": 9254, + "estimated_eval_tokens": 18764 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is thematically loose: many items are tokenizer/model-loading regressions, but the soft pairs usually point to different concrete code paths, models, or failure modes. I would not merge any of the soft candidates as duplicates. If one issue must represent the cluster, 44462 is the closest broad tokenizer-loading regression, but it is still only a loose umbrella fit.", - "confidence": 0.9, - "canonical_issue_reason": "No strong canonical duplicate exists; if forced to pick a representative, issue 44462 is the broadest tokenizer-loading regression and has inbound references, so it is the closest cluster anchor.", + "summary": "These are mostly unrelated closed issues spanning tokenizer loading, model init, multimodal, and backend-specific regressions. Only one soft pair looks plausibly duplicate; the rest are too semantically different to merge safely.", + "confidence": 0.58, + "canonical_issue_reason": "Issue 44462 is the most representative and concrete candidate: it describes a broad tokenizer-loading regression, is fairly specific, and has follow-up references pointing to the same area.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 44462 is the best representative of the cluster\u2019s loose common theme: a user-facing tokenizer/loading regression affecting multiple downstream reports.", + "best_issue_reason": "Issue 44462 is the strongest standalone representative of the cluster because it is actionable, generalized, and directly about a core loading path rather than a model-specific symptom.", "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:43596", + "right": "issue:44683", + "accept": false, + "reason": "Different failures in different code paths: DeepSpeed ZeRO init vs flex_attention compilation on newer torch." + }, { "left": "issue:44488", "right": "issue:44987", "accept": false, - "reason": "Both are load failures, but one is model-specific tokenization loading for cjvt/sleng-bert and the other is a transformers>=5.1 regression on physical-intelligence/fast; different bugs." + "reason": "Both concern model loading, but they point to different repositories/models and likely different root causes." }, { "left": "issue:44488", "right": "issue:45081", "accept": false, - "reason": "cjvt/sleng-bert loading vs Mistral regex patch/backend_tokenizer crash are different tokenizer paths and failure causes." + "reason": "Repo-specific tokenizer loading failure vs a Mistral regex patch crash; not the same bug." }, { "left": "issue:43611", "right": "issue:43644", "accept": false, - "reason": "base_model_prefix loading failure and junk-filled non-persistent buffers are both v5 regressions, but they are distinct initialization/loading problems." + "reason": "Both are v5 regressions, but one is base_model_prefix loading and the other is non-persistent buffer initialization; too different." }, { "left": "issue:42796", "right": "issue:43404", "accept": false, - "reason": "Tokenizer BPE marker handling and Mistral3 lm_head weight tying are separate bugs in different parts of the stack." + "reason": "Tokenizer BPE-marker handling and tied lm_head weights are unrelated issues in Mistral3." }, { "left": "issue:44462", "right": "issue:46489", "accept": false, - "reason": "Ignoring tokenizer.json and DeepSeek-Coder v1 incorrect output are related only broadly to tokenizers; they are not the same concrete defect." + "reason": "Both are tokenizer-related, but one is about AutoTokenizer ignoring tokenizer.json and the other is a DeepSeek-Coder output regression." }, { "left": "issue:36032", "right": "issue:42915", "accept": false, - "reason": "T5Tokenizer add_special_tokens conflict and Qwen3Moe FineGrainedFP8Config failure are unrelated." + "reason": "T5 tokenizer method conflict vs Qwen3Moe FP8 config failure; unrelated subsystems." }, { "left": "issue:46285", "right": "issue:46402", "accept": false, - "reason": "CLIP structure inconsistency and Sam3Model text encoder weights not loading are model-specific and have different root causes." + "reason": "CLIP model structure change vs Sam3 text-encoder weights not loading; different models and symptoms." }, { "left": "issue:43716", "right": "issue:45084", "accept": false, - "reason": "Image preprocessor dtype mismatch and a compilation/template-node TypeError are unrelated failures." + "reason": "Image preprocessor dtype mismatch and template compilation failure are unrelated." }, { "left": "issue:39404", "right": "issue:44671", "accept": false, - "reason": "Whisper return_language pipeline breakage and CamemBERT masked-LM prediction errors are different model/code-path regressions." + "reason": "Whisper pipeline language return regression vs CamemBERT masked-LM prediction bug; no shared bug." }, { "left": "issue:44737", "right": "issue:45859", "accept": false, - "reason": "Both involve device placement, but XLNet CPU arange and Qwen3_5Moe CPU-offload rotary embedding are separate bugs." + "reason": "XLNet CPU placement issue and Qwen3_5Moe CPU-offload incompatibility are different implementations." }, { "left": "issue:43643", "right": "issue:44530", "accept": false, - "reason": "trust_remote_code missing fields in AutoConfig and PagedAttentionCache linear_attention crash are unrelated." + "reason": "AutoConfig trust_remote_code field loss vs PagedAttentionCache linear_attention crash; unrelated." }, { "left": "issue:44661", "right": "issue:44683", "accept": false, - "reason": "Model registration in TOKENIZER_MAPPING_NAMES and compiled flex_attention on torch>=2.9 do not share a concrete bug." + "reason": "Model registration logic and flex_attention compilation failure do not share a concrete bug." }, { - "left": "issue:40990", - "right": "issue:43421", + "left": "issue:42915", + "right": "issue:44530", "accept": false, - "reason": "High perplexity on gpt-oss-20b and runtime special-token post-processor updates are different symptoms and code paths." + "reason": "Different Qwen-family failures with different error modes and likely different fixes." }, { - "left": "issue:42915", - "right": "issue:44530", + "left": "issue:40990", + "right": "issue:43421", "accept": false, - "reason": "Qwen3Moe FP8 config failure and Qwen3.5 PagedAttentionCache linear_attention crash are not the same issue." + "reason": "Perplexity regression on a model/dataset benchmark vs runtime special-token post-processor sync bug." }, { "left": "issue:44451", "right": "issue:44987", "accept": false, - "reason": "Loading vesteinn/ScandiBERT and loading physical-intelligence/fast on newer transformers are different model-specific load failures." + "reason": "Both are loading failures, but the affected models and likely root causes differ." }, { "left": "issue:43541", "right": "issue:45084", "accept": false, - "reason": "Mixtral grouped_mm tracing error and non-template-node compile error are distinct compiler/runtime failures." + "reason": "Torch dynamo grouped_mm tracing failure vs template-node compilation failure; distinct compiler issues." }, { "left": "issue:34567", "right": "issue:40444", "accept": false, - "reason": "TrainerState token counter regression and Qwen2.5-VL iterable-dataset multi-image failure are unrelated." + "reason": "Trainer token-count tracking and IterableDataset multimodal finetuning failure are not the same bug." }, { "left": "issue:43644", "right": "issue:43646", - "accept": false, - "reason": "Non-persistent buffer junk and custom model initialization failures are both v5 regressions, but not the same bug." + "accept": true, + "reason": "Both describe the same v5 model-initialization regression: custom model initialization is broken, and non-persistent buffers being filled with junk is a specific manifestation of it." }, { "left": "issue:43582", "right": "issue:43632", "accept": false, - "reason": "AppleSilicon caching_allocator_warmup TypeError and _is_hf_initialized flag breakage are unrelated." + "reason": "Apple Silicon allocator warmup TypeError and _is_hf_initialized flag breakage are unrelated." }, { "left": "issue:40444", "right": "issue:46326", "accept": false, - "reason": "Qwen2.5-VL multi-image finetuning failure and mistral_common continue_final_message behavior are different model/tokenizer issues." + "reason": "IterableDataset multimodal finetuning failure vs mistral_common continue_final_message bug; different code paths." }, { "left": "issue:42175", "right": "issue:46178", "accept": false, - "reason": "Missing tensorflow in the torch extra is a packaging/dependency issue; tensorflow_text not in BACKENDS_MAPPING is a backend registration bug." + "reason": "Backend packaging/install issue and missing BACKENDS_MAPPING entry are different problems." }, { "left": "issue:44206", "right": "issue:44220", "accept": false, - "reason": "Both are audio-feature issues, but one is an unsupported center arg in LasrFeatureExtractor and the other is a lower-level _torch_extract_fbank_features problem; not clearly the same defect." + "reason": "Both audio-related, but one is an unsupported center argument and the other is a fbank feature extraction issue." }, { "left": "issue:36032", "right": "issue:43645", "accept": false, - "reason": "T5Tokenizer add_special_tokens conflict and Jupyter custom model initialization breakage are unrelated." + "reason": "Tokenizer method-name conflict and Jupyter custom model initialization breakage are not the same issue." }, { "left": "issue:44811", "right": "issue:45584", "accept": false, - "reason": "Whisper batch_decode ignoring skip_special_tokens and empty transcription failure after align_special_tokens are related only by subsystem, not a single concrete bug." + "reason": "Whisper decoding skip_special_tokens bug vs empty transcription after align_special_tokens; related area, but not the same failure." }, { "left": "issue:43452", "right": "issue:43482", "accept": false, - "reason": "gguf_file breaking from_pretrained APIs and Qwen2.5-GGUF loading failure are both GGUF-related but not the same specific regression." + "reason": "gguf_file loading breakage and Qwen2.5-GGUF v5 loading failure are different model-loading problems." }, { "left": "issue:44987", "right": "issue:44991", "accept": false, - "reason": "Both are tokenizer/model loading failures on v5, but they involve different models and likely different root causes." + "reason": "Both are tokenizer/model loading regressions, but they affect different models and likely different fixes." }, { "left": "issue:45084", "right": "issue:45507", "accept": false, - "reason": "Compilation of non-template nodes and GraniteMoEHybrid calling an invalid method are unrelated." + "reason": "Template compilation error vs GraniteMoEHybrid invalid method call; unrelated." }, { "left": "issue:43317", "right": "issue:43482", "accept": false, - "reason": "device_map=auto offload/dequantization failure and Qwen2.5-GGUF loading failure are different loading paths." + "reason": "Dequantized model offload failure vs Qwen2.5-GGUF loading regression; different bugs." }, { "left": "issue:43883", "right": "issue:44488", "accept": false, - "reason": "Missing all_tied_weights_keys in MolmoForCausalLM and loading cjvt/sleng-bert are unrelated model-specific issues." + "reason": "Missing tied-weights attribute on a model class vs inability to load a specific tokenizer/model; unrelated." }, { "left": "issue:43746", "right": "issue:46326", "accept": false, - "reason": "GraniteSpeech PEFT adapter local-checkpoint loading and mistral_common continue_final_message are different failures in different model families." + "reason": "PEFT local-checkpoint loading failure vs mistral_common message-format bug; different surfaces and fixes." } ] }, - "evaluator_result": null, - "evaluator_used": false, + "evaluator_result": { + "accept": true, + "feedback": "The analysis is grounded in the packet and the soft-edge judgments are conservative. The one accepted pair, issue:43644 and issue:43646, looks plausibly mergeable as the same v5 custom model initialization regression with one being a specific symptom. The rest of the pairs are rejected with cautious, bug-specific reasons. The representative issue choice is also reasonable and not overstated." + }, + "evaluator_used": true, "retried": false, - "accepted_nontrivial_soft_edge": false, + "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null }, { "label": "issue", "component_index": 1, - "component_count": 25, + "component_count": 26, "review_unit_index": 11, "review_unit_count": 61, "status": "reviewed", @@ -3581,8 +3576,8 @@ "issue:44403|issue:45081", "issue:43792|issue:46869", "issue:42915|issue:45084", - "issue:44849|issue:45325", "issue:43377|issue:43819", + "issue:44849|issue:45325", "issue:43404|issue:44466", "issue:44464|issue:45987", "issue:44987|issue:45081", @@ -3605,20 +3600,20 @@ "issue:45216|issue:45357", "issue:44991|issue:46315" ], - "prepared_review_unit_hash": "90bb67bd0fe2238ddb939ed7947abd8263ff3d6af2ee4992e4601fbb61e6d466", + "prepared_review_unit_hash": "303e391b421c229000056ee4bdd5050b8e258923ca0bd3fde8a16cf971d263bf", "component_budget": { "node_count": 505, "item_count": 505, "soft_pair_count": 1854, - "serialized_chars": 962354, - "estimated_input_tokens": 240589, - "estimated_eval_tokens": 481434 + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 30, - "serialized_chars": 36925, + "serialized_chars": 36927, "estimated_input_tokens": 9232, "estimated_eval_tokens": 18720 }, @@ -3627,198 +3622,198 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "Most candidate pairs are topical false positives rather than true duplicates. I only accept two soft edges: the Qwen video-input regression pair (44479/45325) and the Qwen3.5 save_pretrained regression pair (45216/45357).", - "confidence": 0.78, - "canonical_issue_reason": "issue:44479 is the best umbrella issue in this set because it describes the broad Qwen multimodal video regression, and issue:45325 reads like a specific manifestation of that same bug. The rest of the items are largely unrelated, so no other issue is a better global anchor.", + "summary": "The cluster is mostly heterogeneous, with one clear duplicate pair around Qwen3.5 save_pretrained checkpoint corruption. The rest are distinct bugs across different models, subsystems, and failure modes, so they should stay separate.", + "confidence": 0.71, + "canonical_issue_reason": "issue:45216 is the broader, earlier report of the Qwen3.5 save_pretrained regression; issue:45357 looks like a later, narrower manifestation focused on incorrect visual encoder keys.", "canonical_pr_reason": null, - "best_issue_reason": "issue:44479 is the strongest representative because it captures the broader regression affecting multiple Qwen multimodal variants, whereas 45325 is narrower and model-specific.", + "best_issue_reason": "issue:45216 is the best representative of the only true duplicate pair because it captures the broader checkpoint-writing bug, while issue:45357 is a more specific follow-up.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43262", "right": "issue:43377", "accept": false, - "reason": "Both involve audio/MIMI-related behavior, but one is a sampling-rate default bug and the other is a batching/padding-mask mismatch; different code paths." + "reason": "Both are audio-related, but one is a chat-template sampling-rate default issue and the other is missing padding-mask support causing batched/single mismatches." }, { "left": "issue:43208", "right": "issue:47013", "accept": false, - "reason": "Same model family, but 43208 is a broad multi-bug training report while 47013 is a specific unpacking crash in train mode; not the same bug." + "reason": "Both mention xLSTM, but one is a set of training blockers for small models and the other is a specific return_last_states unpack crash." }, { "left": "issue:43646", "right": "issue:44534", "accept": false, - "reason": "Custom model initialization in v5 is unrelated to non-persistent buffer corruption." + "reason": "Both are v5/custom-model regressions, but one is notebook/custom init breakage and the other is non-persistent buffer junk serialization." }, { "left": "issue:43645", "right": "issue:44534", "accept": false, - "reason": "Jupyter/custom model init regression is not the same underlying issue as junk-filled non-persistent buffers." + "reason": "Same broad v5 area, but different concrete failures: Jupyter custom model init versus buffer state corruption." }, { "left": "issue:44403", "right": "issue:45081", "accept": false, - "reason": "Loading-noise issue and Mistral regex patch crash are different behaviors with different failure modes." + "reason": "Unnecessary loading noise is unrelated to a tokenizer regex patch crash in Mistral." }, { "left": "issue:43792", "right": "issue:46869", "accept": false, - "reason": "Whisper loading failure and YOLO multi-GPU device-map failure are unrelated." + "reason": "Different models and different code paths: Whisper load failure versus YOLO multi-GPU device-map failure." }, { "left": "issue:42915", "right": "issue:45084", "accept": false, - "reason": "FineGrainedFP8Config failure and template-node compilation error are distinct bugs." + "reason": "One is a FineGrainedFP8Config failure and the other is a template-compilation TypeError; they are not the same bug." }, { - "left": "issue:44849", - "right": "issue:45325", + "left": "issue:43377", + "right": "issue:43819", "accept": false, - "reason": "Different Qwen model families and different symptoms; one is hidden-state output handling, the other is video/rope indexing." + "reason": "Both are audio model issues, but one is batching/padding-mask behavior and the other is DAC latent reconstruction mismatch." }, { - "left": "issue:43377", - "right": "issue:43819", + "left": "issue:44849", + "right": "issue:45325", "accept": false, - "reason": "Both are audio-model discrepancies, but the causes differ: missing padding-mask support versus missing STE in DAC.from_latents." + "reason": "Both are Qwen-family regressions, but one affects output_hidden_states and the other affects VL rope index position scaling." }, { "left": "issue:43404", "right": "issue:44466", "accept": false, - "reason": "Both touch tied lm_head weights, but one is a model-specific untied-weight bug and the other is save/serialization inconsistency; not the same concrete fix." + "reason": "Both touch tied weights, but one is a model-class weight-tying bug and the other is device-dependent save_pretrained serialization." }, { "left": "issue:44464", "right": "issue:45987", "accept": false, - "reason": "Chunked generation inconsistency under compiled forward is unrelated to StaticCache sequence-length type contract." + "reason": "Different generation issues: compiled-forward chunked generation inconsistency versus StaticCache returning the wrong type." }, { "left": "issue:44987", "right": "issue:45081", "accept": false, - "reason": "Tokenizer loading failure for one model and Mistral regex-patching crash are different code paths." + "reason": "Tokenizer loading regression for one model is unrelated to the Mistral backend_tokenizer crash." }, { "left": "issue:28530", "right": "issue:34567", "accept": false, - "reason": "Early stopping metric selection and num_input_tokens_seen tracking are separate Trainer issues." + "reason": "Early stopping metric lookup and num_input_tokens_seen updates are unrelated training-state problems." }, { "left": "issue:43994", "right": "issue:44610", "accept": false, - "reason": "Different multimodal preprocessing problems for different models; not the same bug." + "reason": "Different multimodal preprocessing bugs: wrong AutoModel/pipeline outputs versus wrong processor image size." }, { "left": "issue:42915", "right": "issue:44291", "accept": false, - "reason": "Both involve model loading/configuration in v5, but the specific failures and root causes differ." + "reason": "Both involve loading/init, but one is FP8 configuration failure and the other is an unexpected _is_hf_initialized argument in init_empty_weights." }, { "left": "issue:44008", "right": "issue:46899", "accept": false, - "reason": "Gemma 3n variable-name collision and Gemma4 4-bit dtype-cast regression are distinct bugs." + "reason": "Different Gemma multimodal failures: variable-name collision versus unsafe dtype casting under 4-bit quantization." }, { "left": "issue:44451", "right": "issue:45081", "accept": false, - "reason": "ScandiBERT loading failure and Mistral tokenizer regex crash are unrelated." + "reason": "A generic model load failure is not the same as a Mistral tokenizer patching crash." }, { "left": "issue:43883", "right": "issue:44991", "accept": false, - "reason": "Missing all_tied_weights_keys and tokenizer loading failure for est-roberta are different issues." + "reason": "Molmo tied-weights attribute error and EMBEDDIA tokenizer loading failure are unrelated." }, { "left": "issue:36032", "right": "issue:43742", "accept": false, - "reason": "Tokenizer special-token conflict and MobileLLM key error are not the same failure." + "reason": "Tokenizer special-token construction and MobileLLM key loading are different failure modes." }, { "left": "issue:44062", "right": "issue:44987", "accept": false, - "reason": "AddedToken keyword collision and physical-intelligence/fast tokenizer load failure are unrelated." + "reason": "AddedToken kwarg duplication is unrelated to the physical-intelligence/fast loading regression." }, { "left": "issue:44265", "right": "issue:44610", "accept": false, - "reason": "torch.export + torch_compilable_check is unrelated to OmDet-Turbo image-size preprocessing." + "reason": "torch.export export failure and OmDet-Turbo processor size mismatch do not share the same underlying bug." }, { "left": "issue:43632", "right": "issue:46821", "accept": false, - "reason": "_is_hf_initialized breakage and TimesFM window_size AttributeError are different regression classes." + "reason": "_is_hf_initialized/custom-model breakage is unrelated to the TimesFM window_size attribute error." }, { "left": "issue:43089", "right": "issue:45987", "accept": false, - "reason": "Generation performance overhead and StaticCache return-type bug are not the same underlying problem." + "reason": "Generation overhead from syncs/dispatch is a performance issue, not the same as a cache-length type contract bug." }, { "left": "issue:43322", "right": "issue:43540", "accept": false, - "reason": "Both are multimodal load/process failures, but they affect different model stacks and different failure points." + "reason": "Different multimodal bugs: Llava Next loading segfault versus Qwen3OmniMoe video-input processing ValueError." }, { "left": "issue:42915", "right": "issue:44589", "accept": false, - "reason": "FineGrainedFP8Config failure is unrelated to Float8 storage deserialization." + "reason": "FineGrainedFP8Config failure and missing Float8 storage lookup are both fp8-adjacent but different concrete errors." }, { "left": "issue:44479", "right": "issue:45325", - "accept": true, - "reason": "45325 is a specific Qwen2.5-VL manifestation of the broader v5.3 video-input regression described in 44479; same code-path class." + "accept": false, + "reason": "One is a video-input regression across several Qwen models; the other is a still-image position-id scaling bug." }, { "left": "issue:36032", "right": "issue:41720", "accept": false, - "reason": "Tokenizer constructor conflict and Qwen3 auto device-map CUDA assert are unrelated." + "reason": "Tokenizer add_special_tokens conflict is unrelated to Qwen3 auto device-mapping cudaErrorAssert." }, { "left": "issue:43531", "right": "issue:44877", "accept": false, - "reason": "Sliding-window behavior in Qwen3-MoE and strict config loading for granite_speech are separate issues." + "reason": "Qwen3-MoE sliding_window behavior and granite_speech strict-config loading are unrelated." }, { "left": "issue:45216", "right": "issue:45357", "accept": true, - "reason": "Both report the same Qwen3.5 save_pretrained regression around incorrect checkpoint serialization/visual encoder keys; 45357 looks like a follow-up reproduction." + "reason": "Same underlying Qwen3.5 save_pretrained checkpoint corruption; 45357 is a narrower report about incorrect visual encoder keys." }, { "left": "issue:44991", "right": "issue:46315", "accept": false, - "reason": "Tokenizer loading breakage and slower convert_tokens_to_ids are different tokenizer issues with different symptoms and fixes." + "reason": "Tokenizer loading failure and slow convert_tokens_to_ids in slow tokenizers are different bugs." } ] }, "evaluator_result": { "accept": true, - "feedback": "The summary is grounded in the packet and stays conservative. The two accepted soft edges are plausibly the same bug class from the titles/body context: the Qwen video/rope regression pair and the Qwen3.5 save_pretrained regression pair. The rest are reasonably rejected as different failure modes or model-specific regressions. No overclaiming stands out." + "feedback": "The summary is grounded in the packet: the titles support a mostly heterogeneous cluster with one plausible duplicate pair, issue:45216 and issue:45357, both about Qwen3.5 save_pretrained checkpoint/key corruption. The rejected soft pairs are phrased conservatively and stay close to the specific failure modes in the titles, without overstating the evidence." }, "evaluator_used": true, "retried": false, @@ -3829,7 +3824,7 @@ { "label": "issue", "component_index": 1, - "component_count": 25, + "component_count": 26, "review_unit_index": 12, "review_unit_count": 61, "status": "reviewed", @@ -3896,14 +3891,14 @@ "issue:44220|issue:44683", "issue:45812|issue:46491", "issue:42796|issue:43746", - "issue:43381|issue:45923", "issue:44403|issue:44991", + "issue:43381|issue:45923", "issue:43632|issue:44756", "issue:33666|issue:40444", "issue:40990|issue:44811", "issue:43792|issue:44589", - "issue:42796|issue:46710", "issue:44291|issue:44589", + "issue:42796|issue:46710", "issue:44079|issue:45663", "issue:35707|issue:40444", "issue:43482|issue:43525", @@ -3920,220 +3915,220 @@ "issue:43611|issue:43646", "issue:43377|issue:45491" ], - "prepared_review_unit_hash": "b1d43cf63d6caf26d33ca1e71bd066b5a70fe4f3151be00a37a52aca87404737", + "prepared_review_unit_hash": "b1cf07fe662836f58dfb6255d49094051ea743edffe0e31061f4b124ef94cdac", "component_budget": { "node_count": 505, "item_count": 505, "soft_pair_count": 1854, - "serialized_chars": 962354, - "estimated_input_tokens": 240589, - "estimated_eval_tokens": 481434 + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 }, "budget": { "node_count": 47, "item_count": 47, "soft_pair_count": 31, - "serialized_chars": 36384, - "estimated_input_tokens": 9096, - "estimated_eval_tokens": 18448 + "serialized_chars": 36386, + "estimated_input_tokens": 9097, + "estimated_eval_tokens": 18450 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is heterogeneous: it mixes unrelated tokenizer regressions, multimodal training/runtime bugs, packaging issues, and several distinct Transformers v5 compatibility problems. I don\u2019t see a true duplicate pair, so no soft edges should be merged.", - "confidence": 0.91, - "canonical_issue_reason": "No single canonical issue cleanly represents the cluster; the closest thing is 43611 because it is the broadest v5 loading regression, but the rest of the items are largely unrelated or only loosely v5-adjacent.", + "summary": "The soft pairs are mostly false positives: they share broad themes like tokenizers, v5 regressions, or multimodal training, but each looks like a different underlying bug. No pair is clearly the same concrete fix target, so I would not merge any of them as duplicates.", + "confidence": 0.93, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "43611 is the best representative only in the sense that it is a broad, generic loading regression; however, the overall set is too diverse for a strong canonical issue.", + "best_issue_reason": null, "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43425", "right": "issue:44403", "accept": false, - "reason": "Different problems: Torch 2.10 incompatibility vs noisy transformer loading logs." + "reason": "One is a Torch 2.10 compatibility problem; the other is noisy model-loading behavior. Same broad area, different bug." }, { "left": "issue:43716", "right": "issue:46899", "accept": false, - "reason": "Both are multimodal dtype issues, but they hit different models and different code paths." + "reason": "Both mention dtype issues in multimodal paths, but they involve different models and different failure points/casts." }, { "left": "issue:40444", "right": "issue:45325", "accept": false, - "reason": "Qwen2.5-VL training with multi-image batches is not the same bug as still-image rope position scaling." + "reason": "Different Qwen2.5-VL bugs: iterable dataset multi-image finetuning vs still-image rope/position-id scaling." }, { "left": "issue:43611", "right": "issue:44534", "accept": false, - "reason": "Both are Transformers v5 regressions, but one is base_model_prefix loading and the other is non-persistent buffer initialization." + "reason": "Both are Transformers v5 regressions, but one is model loading/base_model_prefix and the other is buffer initialization garbage." }, { "left": "issue:43296", "right": "issue:43531", "accept": false, - "reason": "Different failures: PaddleOCR-VL load in vLLM vs Qwen3-MoE sliding_window behavior." + "reason": "Different subsystems: vLLM loading failure for PaddleOCR-VL vs Qwen3-MoE sliding_window behavior." }, { "left": "issue:44220", "right": "issue:44683", "accept": false, - "reason": "Audio feature extraction bug vs torch flex_attention compile failure." + "reason": "Audio feature extraction bug vs compiled flex_attention failure; unrelated code paths." }, { "left": "issue:45812", "right": "issue:46491", "accept": false, - "reason": "Both are tokenizer regressions, but they affect different models and different behaviors." + "reason": "Both are tokenizer regressions, but one is AutoTokenizer model-ID selection and the other is a leading-space round-trip issue in CodeLlama." }, { "left": "issue:42796", "right": "issue:43746", "accept": false, - "reason": "Tokenizer BPE-marker handling is unrelated to GraniteSpeech PEFT checkpoint loading." + "reason": "Tokenizer BPE marker handling vs GraniteSpeech PEFT checkpoint loading; unrelated." }, { - "left": "issue:43381", - "right": "issue:45923", + "left": "issue:44403", + "right": "issue:44991", "accept": false, - "reason": "Eval-mode gradient checkpointing use is different from a missing supports_gradient_checkpointing flag." + "reason": "Generic loading noise is not the same as a specific tokenizer-loading failure for EMBEDDIA/est-roberta." }, { - "left": "issue:44403", - "right": "issue:44991", + "left": "issue:43381", + "right": "issue:45923", "accept": false, - "reason": "Generic loading noise is not the same as tokenizer loading failure for a specific model." + "reason": "Eval-mode gradient checkpointing error vs missing supports_gradient_checkpointing flag on a trust_remote_code model." }, { "left": "issue:43632", "right": "issue:44756", "accept": false, - "reason": "Both are v5-related, but one is _is_hf_initialized and the other is mmap/OOM behavior." + "reason": "Different failures: _is_hf_initialized regression vs mmap OOM workaround." }, { "left": "issue:33666", "right": "issue:40444", "accept": false, - "reason": "Multi-GPU Qwen2-VL training and multi-image iterable dataset failures are different bugs." + "reason": "Different Qwen2-VL vs Qwen2.5-VL training issues with no concrete shared bug." }, { "left": "issue:40990", "right": "issue:44811", "accept": false, - "reason": "Perplexity regression on GPT-OSS is unrelated to Whisper batch_decode skip_special_tokens handling." + "reason": "Perplexity anomaly on gpt-oss-20b is unrelated to Whisper batch_decode skip_special_tokens handling." }, { "left": "issue:43792", "right": "issue:44589", "accept": false, - "reason": "Whisper runtime failure is not the same as the Float8 storage-object TypeError." + "reason": "Whisper runtime failure vs missing Float8 storage type; different causes." }, { - "left": "issue:42796", - "right": "issue:46710", + "left": "issue:44291", + "right": "issue:44589", "accept": false, - "reason": "Tokenizer BPE-marker bug vs model accuracy regression after upgrade." + "reason": "init_empty_weights/_is_hf_initialized regression is distinct from Float8 storage lookup failure." }, { - "left": "issue:44291", - "right": "issue:44589", + "left": "issue:42796", + "right": "issue:46710", "accept": false, - "reason": "init_empty_weights/_is_hf_initialized argument error is separate from Float8 storage lookup failure." + "reason": "Tokenizer BPE marker bug vs downstream accuracy regression after upgrade; not the same concrete defect." }, { "left": "issue:44079", "right": "issue:45663", "accept": false, - "reason": "ModelOutput key assignment bug is unrelated to Gemma4 shared_kv_states rebuilding." + "reason": "ModelOutput key assignment bug is unrelated to Gemma4 FSDP2 shared_kv_states KeyError." }, { "left": "issue:35707", "right": "issue:40444", "accept": false, - "reason": "Progressive generation with inputs_embeds/past_key_values is unrelated to Qwen2.5-VL multi-image finetuning." + "reason": "Progressive generation with inputs_embeds/past_key_values is unrelated to multimodal iterable-dataset finetuning." }, { "left": "issue:43482", "right": "issue:43525", "accept": false, - "reason": "GGUF loading under v5 and Llama4Config missing pad_token_id are different model-load failures." + "reason": "GGUF loading failure in Transformers v5 and missing pad_token_id on Llama4Config are different bugs." }, { "left": "issue:43632", "right": "issue:44534", "accept": false, - "reason": "Both mention v5, but they affect different internal mechanisms and symptoms." + "reason": "Both are v5 regressions, but one concerns initialization flags and the other non-persistent buffer values." }, { "left": "issue:43531", "right": "issue:44291", "accept": false, - "reason": "Qwen3-MoE sliding_window and init_empty_weights signature errors are unrelated." + "reason": "Qwen3-MoE sliding_window behavior is unrelated to init_empty_weights argument handling." }, { "left": "issue:43421", "right": "issue:44568", "accept": false, - "reason": "Runtime post-processor updates and add_special_tokens BOS/EOS regression are distinct tokenizer issues." + "reason": "Runtime post-processor sync for special tokens is different from add_special_tokens not inserting BOS/EOS." }, { "left": "issue:43482", "right": "issue:45538", "accept": false, - "reason": "GGUF loading failure is unrelated to CLIPTokenizer max length behavior." + "reason": "GGUF loading failure and CLIPTokenizer max_length default are unrelated." }, { "left": "issue:42175", "right": "issue:45542", "accept": false, - "reason": "Packaging/extra dependency omission is different from a tensorboard-without-TensorFlow backend error." + "reason": "Both involve TensorFlow optional deps, but one is missing backend install metadata and the other is a runtime undefined backend error." }, { "left": "issue:41720", "right": "issue:45081", "accept": false, - "reason": "Qwen3 auto device mapping CUDA assert is unrelated to Mistral regex patching." + "reason": "Qwen3 auto device mapping cudaErrorAssert is unrelated to Mistral tokenizer regex patch crashing." }, { "left": "issue:44079", "right": "issue:44534", "accept": false, - "reason": "ModelOutput key assignment and non-persistent buffer junk are different regressions." + "reason": "ModelOutput key handling and non-persistent buffer junk are distinct Transformer v5 regressions." }, { "left": "issue:43317", "right": "issue:43856", "accept": false, - "reason": "Device-map dequantized load failure is unrelated to Qwen3 MoE memory inefficiency." + "reason": "Dequantized model loading on gpu+cpu offload is unrelated to Qwen3 MoE training memory use." }, { "left": "issue:44462", "right": "issue:45812", "accept": false, - "reason": "Ignoring tokenizer.json and wrong Granite token IDs are not the same underlying bug." + "reason": "Both are AutoTokenizer regressions, but one is ignoring tokenizer.json while the other is wrong IDs on Granite models; not clearly the same bug." }, { "left": "issue:43296", "right": "issue:46821", "accept": false, - "reason": "PaddleOCR-VL load failure and TimesFM window_size AttributeError are unrelated." + "reason": "PaddleOCR-VL load failure is unrelated to TimesFM window_size AttributeError." }, { "left": "issue:43611", "right": "issue:43646", "accept": false, - "reason": "Both are v5 loading/init regressions, but the concrete failures differ enough that they should not be merged." + "reason": "Both mention Transformers 5.0.0, but model loading with base_model_prefix and custom model initialization are not the same defect." }, { "left": "issue:43377", "right": "issue:45491", "accept": false, - "reason": "Both are batching-related, but one is missing padding-mask support and the other is NaNs from sliding-window attention." + "reason": "Missing padding-mask support in MIMI encoder is a different bug from Gemma3 NaN embeddings with mixed-length sliding-window batches." } ] }, @@ -4147,7 +4142,7 @@ { "label": "issue", "component_index": 1, - "component_count": 25, + "component_count": 26, "review_unit_index": 13, "review_unit_count": 61, "status": "reviewed", @@ -4228,8 +4223,8 @@ "issue:43606|issue:44265", "issue:43262|issue:43746", "issue:44479|issue:45823", - "issue:35707|issue:41720", "issue:43828|issue:46821", + "issue:35707|issue:41720", "issue:34634|issue:35707", "issue:39692|issue:40444", "issue:44186|issue:44265", @@ -4237,231 +4232,228 @@ "issue:41720|issue:45823", "issue:43606|issue:43819" ], - "prepared_review_unit_hash": "f2879daca8a8c1a5f9e5b23160d06d009e80102ac35022037f97811394d7a621", + "prepared_review_unit_hash": "0918d6736931199622805e0ad6f82733baad936f3ce8c6a7ef03552dd00f8112", "component_budget": { "node_count": 505, "item_count": 505, "soft_pair_count": 1854, - "serialized_chars": 962354, - "estimated_input_tokens": 240589, - "estimated_eval_tokens": 481434 + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 }, "budget": { "node_count": 47, "item_count": 47, "soft_pair_count": 30, - "serialized_chars": 36273, - "estimated_input_tokens": 9069, - "estimated_eval_tokens": 18394 + "serialized_chars": 36270, + "estimated_input_tokens": 9068, + "estimated_eval_tokens": 18392 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "These items are mostly unrelated closed issues. The only soft edge that looks like the same underlying bug is the SigLIP2 pair; the rest differ by model, code path, or failure mode too much for duplicate triage.", - "confidence": 0.87, - "canonical_issue_reason": "issue:43493 is the clearest root-cause report for the SigLIP2 HF-vs-JAX discrepancy, while issue:43994 reads like a downstream symptom of that same problem.", + "summary": "I did not find any soft pair that looks like the same concrete bug/change. Most pairs share only a broad subsystem or a model family name, but the failure modes are distinct enough to keep separate.", + "confidence": 0.9, + "canonical_issue_reason": "issue:43873 is the broadest still-open report and has the most follow-up activity/inbound references, so it is the best umbrella issue among this set.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43493 is the best representative issue because it describes the underlying implementation mismatch directly and is broader than the user-facing symptom report.", + "best_issue_reason": "issue:43873 is a reasonable global best issue because it is active, open, and broad enough to collect related offloading/quantization reports.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44509", "right": "issue:46909", "accept": false, - "reason": "Both are documentation bugs, but they concern different docs and different content errors." + "reason": "Both are documentation problems, but one is about removed pipeline tasks while the other is a broken DIA example; different docs bugs." }, { "left": "issue:41553", "right": "issue:43582", "accept": false, - "reason": "An AutoTokenizer/Voxtral error-message issue is unrelated to the Apple Silicon allocator TypeError." + "reason": "Tokenizer loading error for Voxtral and an Apple Silicon allocator TypeError are unrelated failures." }, { "left": "issue:42175", "right": "issue:43296", "accept": false, - "reason": "Package backend dependency coverage and PaddleOCR-VL loading are different problems." + "reason": "Package extra/backend dependency issue and a PaddleOCR-VL loading failure are different code paths." }, { "left": "issue:43644", "right": "issue:44079", "accept": false, - "reason": "Both involve v5 regressions, but one is buffer initialization and the other is ModelOutput key assignment." + "reason": "Non-persistent buffer serialization and ModelOutput key assignment are separate implementation bugs." }, { "left": "issue:44829", "right": "issue:45479", "accept": false, - "reason": "Both mention degenerate training/loss, but they affect different model families and appear to have different causes." + "reason": "Both mention bad training behavior, but one is specific to flash_attention_3 and the other to num_labels=1 classification loss handling." }, { "left": "issue:43818", "right": "issue:46402", "accept": false, - "reason": "Video-LLaVA architecture issues are unrelated to Sam3Model text-encoder weight loading." + "reason": "Different models and different bugs: Video-LLaVA architecture/weights vs Sam3 text encoder loading." }, { "left": "issue:43818", "right": "issue:45357", "accept": false, - "reason": "Video-LLaVA tower/attention problems are not the same as Qwen3.5 save_pretrained key serialization bugs." + "reason": "Video-LLaVA temporal-attention/weights issue is unrelated to Qwen3.5 visual encoder key serialization." }, { "left": "issue:40444", "right": "issue:45381", "accept": false, - "reason": "Both are multimodal Qwen2.5-VL issues, but one is iterable-dataset training with multiple images and the other is video position-id computation." + "reason": "Multiple-images iterable dataset failure and video vision_position_ids regression are both Qwen2.5-VL related but not the same concrete bug." }, { "left": "issue:44265", "right": "issue:45072", "accept": false, - "reason": "torch.export with torch_compilable_check is unrelated to bfloat16 dtype mismatches in specific models." + "reason": "torch.export with torch_compilable_check and bfloat16 dtype mismatches are different failures." }, { "left": "issue:43493", "right": "issue:43994", - "accept": true, - "reason": "Both describe the same SigLIP2 implementation discrepancy leading to incorrect outputs across HF usage paths." + "accept": false, + "reason": "Both are about SigLIP2, but one reports implementation discrepancy and the other nonsensical inference results; not clearly the same bug." }, { "left": "issue:45325", "right": "issue:45834", "accept": false, - "reason": "Qwen2.5-VL rope/position-id scaling is unrelated to Kosmos2.5 OCR index errors." + "reason": "Wrong rope/position scaling in Qwen2.5-VL is unrelated to Kosmos2.5 OCR index errors." }, { "left": "issue:45325", "right": "issue:45381", "accept": false, - "reason": "These are both Qwen2.5-VL video/vision bugs, but they concern different indexing logic and failure modes." + "reason": "These both involve Qwen2.5-VL vision positions, but one is still-image rope scaling and the other is video vision_position_ids; too different to merge." }, { "left": "issue:43454", "right": "issue:44466", "accept": false, - "reason": "One is missing lm_head tying in AyaVision; the other is inconsistent serialization behavior in v5." + "reason": "Both mention tied lm_head weights, but one is failure to tie weights and the other is device-dependent serialization inconsistency." }, { "left": "issue:42915", "right": "issue:43482", "accept": false, - "reason": "Qwen3Moe FP8 configuration failure is unrelated to Qwen2.5-GGUF loading in v5." + "reason": "Qwen3Moe FP8 config failure and Qwen2.5-GGUF loading on v5 are unrelated model-loading issues." }, { "left": "issue:44291", "right": "issue:44877", "accept": false, - "reason": "Unexpected _is_hf_initialized argument during init_empty_weights is a different issue from strict config rejection of granite_speech." + "reason": "Unexpected init_empty_weights argument and strict config rejection are different load-time problems." }, { "left": "issue:43873", "right": "issue:45701", "accept": false, - "reason": "Quantization/offloading behavior and tokenization changes are unrelated." + "reason": "Offloading/quantization behavior and tokenization changes are unrelated." }, { "left": "issue:43701", "right": "issue:45325", "accept": false, - "reason": "Checkpoint key mismatch has no clear overlap with Qwen2.5-VL rope indexing." + "reason": "Checkpoint key mismatch has nothing to do with Qwen2.5-VL rope indexing." }, { "left": "issue:42796", "right": "issue:45381", "accept": false, - "reason": "Tokenizer BPE marker handling is unrelated to Qwen2.5-VL video position-id logic." + "reason": "Tokenizer BPE marker handling and video vision position bugs are unrelated." }, { "left": "issue:40990", "right": "issue:45584", "accept": false, - "reason": "High perplexity on GPT-OSS and Whisper empty-transcription failure are different model/task bugs." + "reason": "High perplexity on GPT-OSS and Whisper empty-transcription generation failure are different issues." }, { "left": "issue:43606", "right": "issue:44265", "accept": false, - "reason": "CPU offload device mismatch for Bark is unrelated to torch.export failures on torch_compilable_check." + "reason": "CPU offload device mismatch for Bark and export failure for torch_compilable_check are unrelated." }, { "left": "issue:43262", "right": "issue:43746", "accept": false, - "reason": "Audio processor sampling-rate defaults and GraniteSpeech PEFT checkpoint loading are different issues." + "reason": "Audio chat-template sampling-rate default and GraniteSpeech PEFT checkpoint loading are separate bugs." }, { "left": "issue:44479", "right": "issue:45823", "accept": false, - "reason": "Qwen video-input regression and Gemma4 PLE device mismatch are different model-specific failures." + "reason": "Video-input regression across Qwen models and Gemma4 PLE device mismatch are different code paths." }, { - "left": "issue:35707", - "right": "issue:41720", + "left": "issue:43828", + "right": "issue:46821", "accept": false, - "reason": "inputs_embeds/past_key_values generation is unrelated to Qwen3 auto device-mapping CUDA asserts." + "reason": "Autocast dtype mismatch in Phi-tiny-MoE and TimesFM window_size AttributeError are unrelated." }, { - "left": "issue:43828", - "right": "issue:46821", + "left": "issue:35707", + "right": "issue:41720", "accept": false, - "reason": "Autocast dtype mismatch in Phi-tiny-MoE is unrelated to TimesFM window_size AttributeError." + "reason": "inputs_embeds/past_key_values generation and Qwen3 auto device mapping cudaErrorAssert are different failures." }, { "left": "issue:34634", "right": "issue:35707", "accept": false, - "reason": "BarkProcessor voice_preset and progressive generation with past_key_values are unrelated." + "reason": "Bark voice_preset and progressive generation with past_key_values do not describe the same bug." }, { "left": "issue:39692", "right": "issue:40444", "accept": false, - "reason": "SigLIP2 docs-example issues and Qwen2.5-VL multi-image finetuning failures do not describe the same bug." + "reason": "SigLIP2 doc example errors and Qwen2.5-VL multi-image finetuning failure are unrelated." }, { "left": "issue:44186", "right": "issue:44265", "accept": false, - "reason": "LayoutLMv2 tokenizer crashes are unrelated to torch.export with torch_compilable_check." + "reason": "LayoutLMv2 tokenization crashes and torch.export export failures are different issues." }, { "left": "issue:43454", "right": "issue:43550", "accept": false, - "reason": "AyaVision lm_head tying and Bamba torch.compile SDPA failures are different code paths." + "reason": "AyaVision lm_head tying/serialization and Bamba torch.compile SDPA failure are unrelated." }, { "left": "issue:41720", "right": "issue:45823", "accept": false, - "reason": "Qwen3 auto device mapping CUDA assert and Gemma4 PLE forward device mismatch are unrelated." + "reason": "Qwen3 auto device mapping crash and Gemma4 PLE device mismatch are separate model/runtime bugs." }, { "left": "issue:43606", "right": "issue:43819", "accept": false, - "reason": "Bark CPU-offload device mismatch is unrelated to DAC.from_latents vs forward mismatch with missing STE." + "reason": "Bark CPU offload device mismatch and DAC.from_latents/STE mismatch are unrelated." } ] }, - "evaluator_result": { - "accept": true, - "feedback": "The assessment is grounded in the packet and the soft-edge verdicts are conservative. The SigLIP2 pair (43493/43994) is the only clearly related duplicate-like match by title/body summary, while the other rejected pairs are plausibly different bugs, model families, or failure modes. No overstatement stands out." - }, - "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 }, { "label": "issue", "component_index": 1, - "component_count": 25, + "component_count": 26, "review_unit_index": 14, "review_unit_count": 61, "status": "reviewed", @@ -4524,8 +4516,8 @@ "issue:43257|issue:45923", "issue:44568|issue:45920", "issue:42796|issue:46209", - "issue:46305|issue:46315", "issue:43454|issue:45072", + "issue:46305|issue:46315", "issue:45823|issue:46821", "issue:43881|issue:44488", "issue:40990|issue:42947", @@ -4541,8 +4533,8 @@ "issue:44387|issue:45704", "issue:41720|issue:44977", "issue:36331|issue:42915", - "issue:44466|issue:45357", "issue:40990|issue:45701", + "issue:44466|issue:45357", "issue:43450|issue:45701", "issue:44877|issue:44991", "issue:43065|issue:46402", @@ -4551,20 +4543,20 @@ "issue:44186|issue:45072", "issue:43873|issue:44756" ], - "prepared_review_unit_hash": "df401687d91f02d4d7db23d548dc290a5fc7cb5c34c76e6729c3efa1a751fe96", + "prepared_review_unit_hash": "58ace76568ecf16b07eace4af5990b1adc580b6784f0e310803cc15290e9c0fd", "component_budget": { "node_count": 505, "item_count": 505, "soft_pair_count": 1854, - "serialized_chars": 962354, - "estimated_input_tokens": 240589, - "estimated_eval_tokens": 481434 + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 }, "budget": { "node_count": 47, "item_count": 47, "soft_pair_count": 30, - "serialized_chars": 36064, + "serialized_chars": 36063, "estimated_input_tokens": 9016, "estimated_eval_tokens": 18288 }, @@ -4573,192 +4565,192 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "These items are mostly unrelated single-bug reports. The only loose theme is quantization/offloading, but the titles still point to different concrete failure modes, so I would not merge any of the suggested soft pairs.", - "confidence": 0.49, - "canonical_issue_reason": "Issue 43873 is the broadest and most centrally referenced report in the only near-related theme here (offloading + quantization), so it is the best umbrella representative if a canonical issue must be picked.", + "summary": "All proposed soft edges look like false positives: the issues share broad subsystem wording, but each points to a different concrete bug, model, or code path.", + "confidence": 0.95, + "canonical_issue_reason": "No exact duplicate exists in this set; if one issue must serve as the cluster anchor, issue 43873 is the closest because it is active, discussion-heavy, and broadly representative of the offloading/quantization theme seen in the nearest neighbors.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 43873 has the widest scope, the most discussion/inbound references among the closest matches, and reads like the umbrella problem rather than a narrow symptom.", + "best_issue_reason": "Issue 43873 is the strongest representative candidate by activity and breadth, but it should not be treated as a true duplicate of the others.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43377", "right": "issue:44610", "accept": false, - "reason": "Both involve batched inputs, but one is a MIMI encoder padding-mask bug and the other is a video processor output-shape issue; different components and failure modes." + "reason": "Different model/components: MIMI encoder padding-mask batching bug vs OmDet-Turbo processor/model input-size mismatch." }, { "left": "issue:43257", "right": "issue:45923", "accept": false, - "reason": "Different models and different bugs: MoE weight conversion/loading under accelerate+deepspeed vs a missing gradient-checkpointing capability flag." + "reason": "Both are Qwen-related, but one is MoE weight conversion/loading and the other is a missing gradient-checkpointing flag." }, { "left": "issue:44568", "right": "issue:45920", "accept": false, - "reason": "Both are tokenizer regressions, but one is a model-specific add_special_tokens/BOS-EOS bug while the other is a broader wrong-token-id regression across multiple tokenizers." + "reason": "Both are tokenizer regressions, but they affect different tokenizers and failure modes (special tokens vs wrong IDs)." }, { "left": "issue:42796", "right": "issue:46209", "accept": false, - "reason": "Tokenizer BPE-marker handling and FP8 inference on B200 are unrelated issues." + "reason": "Tokenizer BPE-marker handling is unrelated to FP8 inference failure on B200." }, { - "left": "issue:46305", - "right": "issue:46315", + "left": "issue:43454", + "right": "issue:45072", "accept": false, - "reason": "One is an off-by-one in sliding-window masking; the other is a tokenizer performance regression in convert_tokens_to_ids." + "reason": "One is a tied-weights serialization/generation bug; the other is a bfloat16 dtype mismatch in inference." }, { - "left": "issue:43454", - "right": "issue:45072", + "left": "issue:46305", + "right": "issue:46315", "accept": false, - "reason": "lm_head weight tying and bfloat16 dtype mismatches are distinct code paths and symptoms." + "reason": "Off-by-one sliding-window masking is a different bug from convert_tokens_to_ids performance regression." }, { "left": "issue:45823", "right": "issue:46821", "accept": false, - "reason": "Device mismatch in Gemma4 PLE forward pass is not the same bug as TimesFM's window_size AttributeError." + "reason": "Device-mapping forward mismatch and a window_size AttributeError are different failure paths." }, { "left": "issue:43881", "right": "issue:44488", "accept": false, - "reason": "Both are loading failures, but they concern different models and likely different loader problems." + "reason": "Separate model-loading failures with different models and likely different causes." }, { "left": "issue:40990", "right": "issue:42947", "accept": false, - "reason": "High perplexity on a specific model is unrelated to gradient checkpointing being ineffective with PEFT LoRA." + "reason": "High perplexity on one model/dataset is not the same as gradient-checkpointing ineffectiveness with LoRA." }, { "left": "issue:42617", "right": "issue:42915", "accept": false, - "reason": "A failure to run a script and a FineGrainedFP8Config model failure are different problems." + "reason": "3d_parallel execution failure and Qwen3 MoE FP8 config failure are unrelated." }, { "left": "issue:43377", "right": "issue:44898", "accept": false, - "reason": "Both mention batched/shape-related behavior, but one is about encoder padding masks and the other about Perceiver image resolution handling." + "reason": "Both involve batched-vs-single behavior, but one is missing padding-mask support while the other is an image-size/interpolation failure." }, { "left": "issue:43454", "right": "issue:43720", "accept": false, - "reason": "Tied-weight serialization and packed-weight unpacking are different bugs in different loading/saving paths." + "reason": "Different bugs: lm_head weight tying/serialization vs packed-weight unpacking during accelerate loading." }, { "left": "issue:43749", "right": "issue:44534", "accept": false, - "reason": "FSDP CPU RAM-efficient loading and non-persistent buffer junk are separate regression reports, not the same concrete bug." + "reason": "FSDP CPU-efficient loading breakage and non-persistent buffer junk are separate regressions." }, { "left": "issue:43317", "right": "issue:43873", "accept": false, - "reason": "These are the closest pair, but one is a specific device_map/offload load failure and the other is a broader offloading-with-quantization complaint; not enough evidence they are the same concrete bug." + "reason": "Both mention offload/quantization, but one is dequantized model loading with device_map=auto and the other is a broader quantization offloading problem." }, { "left": "issue:43572", "right": "issue:45759", "accept": false, - "reason": "Config missing pad_token_idx and unwrapping text_config for composite multimodal configs are different configuration issues." + "reason": "Missing pad_token_idx in a config is unrelated to AutoModelForCausalLM not unwrapping text_config for composite multimodal configs." }, { "left": "issue:43322", "right": "issue:46050", "accept": false, - "reason": "A segmentation fault while loading Llava Next is not the same as sinq quantization being broken across save/load." + "reason": "Segmentation fault while loading Llava Next is not the same as sinq quantized model save/load corruption." }, { "left": "issue:41720", "right": "issue:43322", "accept": false, - "reason": "Both are model-loading failures, but one is an auto-device-map cudaErrorAssert and the other is a segmentation fault in Llava Next loading." + "reason": "Qwen3 auto device-mapping cuda assert and Llava Next loading segfault are different model/load paths." }, { "left": "issue:36331", "right": "issue:42617", "accept": false, - "reason": "Unexpected compute_loss kwargs and a parallel script execution failure are unrelated." + "reason": "Trainer compute_loss signature change is unrelated to the 3d_parallel runtime failure." }, { "left": "issue:44387", "right": "issue:45704", "accept": false, - "reason": "Int4 quantization OOM and an apex RMSNorm memory leak are different performance/memory issues with different root causes." + "reason": "Both are memory-related, but int4 CUDA reserved-memory growth is not the same as T5 using apex.FusedRMSNorm." }, { "left": "issue:41720", "right": "issue:44977", "accept": false, - "reason": "cudaErrorAssert under auto device mapping and flash-attention generation issues are different failure modes." + "reason": "Both are Qwen-family reports, but the failures are different: device mapping cuda assert vs flash-attention generation issues." }, { "left": "issue:36331", "right": "issue:42915", "accept": false, - "reason": "The compute_loss signature change and FineGrainedFP8Config failure do not describe the same bug." + "reason": "Unexpected compute_loss argument and Qwen3 MoE FP8 conversion failure are unrelated bugs." }, { - "left": "issue:44466", - "right": "issue:45357", + "left": "issue:40990", + "right": "issue:45701", "accept": false, - "reason": "Both concern save_pretrained serialization, but one is tied-weight lm_head serialization and the other is incorrect visual-encoder key saving for Qwen3.5." + "reason": "Extremely high perplexity and tokenization-version-dependent tokenization changes are different symptoms and code paths." }, { - "left": "issue:40990", - "right": "issue:45701", + "left": "issue:44466", + "right": "issue:45357", "accept": false, - "reason": "Different scope and symptoms: a perplexity regression on one model versus broad tokenization differences across versions." + "reason": "Both touch serialization, but one concerns tied lm_head weights while the other concerns incorrect visual encoder keys." }, { "left": "issue:43450", "right": "issue:45701", "accept": false, - "reason": "Video processor batch shape bugs are unrelated to tokenizer behavior changes." + "reason": "Video processor batch shape bug is unrelated to a tokenizer version changing tokenization." }, { "left": "issue:44877", "right": "issue:44991", "accept": false, - "reason": "These are both loading regressions, but one is a config strictness problem and the other is tokenizer loading for a specific model family." + "reason": "Strict config rejection for granite_speech and tokenizer loading failure for est-roberta are different loaders and causes." }, { "left": "issue:43065", "right": "issue:46402", "accept": false, - "reason": "Both mention Sam3, but dummy Conv2d in the pixel decoder and missing text encoder weights are distinct problems in different parts of the model." + "reason": "Sam3PixelDecoder dummy Conv2d and Sam3Model text encoder weights not loading are different parts of the model." }, { "left": "issue:41720", "right": "issue:42617", "accept": false, - "reason": "Auto device mapping cuda asserts and a 3d_parallel script failure are not the same underlying issue." + "reason": "Different failures: Qwen3 device-mapping cuda assert vs 3d_parallel execution failure." }, { "left": "issue:26413", "right": "issue:28530", "accept": false, - "reason": "Checkpoint resume failing on an empty epoch iterator and early stopping metric lookup are unrelated trainer bugs." + "reason": "Checkpoint-resume epoch_iterator error is unrelated to early-stopping metric resolution." }, { "left": "issue:44186", "right": "issue:45072", "accept": false, - "reason": "Tokenizer crashes on NER/batching and dtype mismatches in inference are unrelated." + "reason": "Tokenizer crash on NER/batched padding is not the same bug as dtype mismatch in bfloat16 inference." }, { "left": "issue:43873", "right": "issue:44756", "accept": false, - "reason": "Both touch memory/offloading, but one is quantization offloading and the other is mmap on Strix Halo; different fixes and failure paths." + "reason": "Both involve memory/offload concerns, but one is quantization offloading behavior and the other is mmap/OOM on Strix Halo." } ] }, @@ -4772,7 +4764,7 @@ { "label": "issue", "component_index": 1, - "component_count": 25, + "component_count": 26, "review_unit_index": 15, "review_unit_count": 61, "status": "reviewed", @@ -4845,8 +4837,8 @@ "issue:43404|issue:45357", "issue:44610|issue:44898", "issue:42915|issue:43317", - "issue:44423|issue:45406", "issue:44734|issue:46858", + "issue:44423|issue:45406", "issue:46821|issue:46869", "issue:44008|issue:45823", "issue:41720|issue:45161", @@ -4860,202 +4852,202 @@ "issue:44186|issue:44361", "issue:43425|issue:46821" ], - "prepared_review_unit_hash": "6dc154f8a69e3bd76b94d70624027051c015915c1fad3dfec3384fa15fc5e0a5", + "prepared_review_unit_hash": "f711175a2d63a874aeab4ccde3976ae6328aaa88444510913c89f4520332eed0", "component_budget": { "node_count": 505, "item_count": 505, "soft_pair_count": 1854, - "serialized_chars": 962354, - "estimated_input_tokens": 240589, - "estimated_eval_tokens": 481434 + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 }, "budget": { "node_count": 47, "item_count": 47, "soft_pair_count": 28, - "serialized_chars": 35675, - "estimated_input_tokens": 8919, - "estimated_eval_tokens": 18094 + "serialized_chars": 35677, + "estimated_input_tokens": 8920, + "estimated_eval_tokens": 18096 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "These items are largely unrelated issue reports across different models and regression classes. The soft pairs mostly share broad subsystem keywords (cache, loading, multimodal, device map) but not the same concrete bug, so none are strong duplicate candidates.", - "confidence": 0.84, - "canonical_issue_reason": "issue:43646 is the broadest, most central regression report in the set, describing a general Transformers 5.0 custom-model initialization breakage that is easier to use as an anchor than the narrower model-specific issues.", + "summary": "This cluster is not a duplicate set; it\u2019s a grab bag of unrelated Transformers issues. All soft-similarity pairs are false positives, mostly because they share broad subsystems (tokenizers, cache, loading, serving, quantization) but describe different concrete failures and code paths.", + "confidence": 0.98, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "issue:46665 is the most actionable and specific report among the set, with a concrete usage pattern (`from_pretrained` inside Python classes) that should be easier to reproduce and fix.", + "best_issue_reason": null, "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43526", "right": "issue:43883", "accept": false, - "reason": "Different bugs: one is a BEiT image processor label-reduction issue, the other is a missing `all_tied_weights_keys` attribute on Molmo CausalLM." + "reason": "Different failures: BeitImageProcessorFast label reduction vs MolmoForCausalLM missing tied-weights metadata." }, { "left": "issue:43881", "right": "issue:44991", "accept": false, - "reason": "Both are loading failures, but for different models and different failure points: GLM-4V model loading vs tokenizer loading for est-roberta." + "reason": "Unrelated model-specific loading problems: GLM-4V load failure vs tokenizer loading for a RoBERTa checkpoint." }, { "left": "issue:46612", "right": "issue:46858", "accept": false, - "reason": "Both involve cache logic, but one is skipped beam-search cache reordering in several models; the other is a prefill/static-cache arithmetic bug on non-CUDA backends." + "reason": "Both involve cache behavior, but one is missing reorder logic for several generation models and the other is a static-cache seq_length tensor/arithmetic bug under compile on non-CUDA backends." }, { "left": "issue:43577", "right": "issue:43708", "accept": false, - "reason": "Unrelated: one is dtype propagation when loading BLIP2, the other is trainer `max_steps` miscalculation after resume with batch-size changes." + "reason": "Distinct areas: dtype propagation in Blip2 loading vs Trainer checkpoint resume/max_steps calculation." }, { "left": "issue:43550", "right": "issue:43720", "accept": false, - "reason": "Different code paths and failures: torch.compile+SDPA in Bamba versus packed-weight unpacking during Accelerate loading for BitNet." + "reason": "Different code paths: torch.compile+SDPA in Bamba vs Accelerate loading of packed BitNet weights." }, { "left": "issue:43295", "right": "issue:45701", "accept": false, - "reason": "Both mention version-related behavior changes, but one is a processor/tokenizer regression and the other is a broader tokenization-change report with different symptoms." + "reason": "One is a regression in processor/tokenizer access and image passing; the other is a broader tokenization change report with different symptoms." }, { "left": "issue:45685", "right": "issue:46821", "accept": false, - "reason": "Completely different areas: MPS histogram support for Int tensors versus a TimesFM `window_size` attribute error." + "reason": "MPS histogram/int bug in MoE routing is unrelated to TimesFM's missing window_size attribute." }, { "left": "issue:43646", "right": "issue:46665", "accept": false, - "reason": "Related only at a high level; one is a broad Transformers 5.0 custom-model initialization regression, the other is a specific `from_pretrained` usage inconsistency inside Python classes." + "reason": "Custom model init breakage in Transformers 5.0 is not the same as inconsistent from_pretrained behavior inside Python classes." }, { "left": "issue:44186", "right": "issue:44610", "accept": false, - "reason": "Different model families and bugs: LayoutLMv2 tokenizer crashing on NER/padding versus OmDet-Turbo image size mismatch." + "reason": "Both are preprocessing issues, but LayoutLMv2 NER/padding crashes are unrelated to OmDet-Turbo's input-size mismatch." }, { "left": "issue:43550", "right": "issue:43606", "accept": false, - "reason": "Both are offload/device-related failures, but they affect different models and are likely different device-placement bugs." + "reason": "Different bugs: SDPA/torch.compile failure vs CPU-offload device mismatch in bark-small." }, { "left": "issue:36010", "right": "issue:42617", "accept": false, - "reason": "No shared root cause: one is an import error for `GenerationMixin`, the other is failure running `3d_parallel.py`." + "reason": "GenerationMixin import breakage is unrelated to a 3d_parallel.py runtime failure." }, { "left": "issue:43404", "right": "issue:45357", "accept": false, - "reason": "Both touch model save/load semantics, but one is tied weights in Mistral3 and the other is incorrect visual-encoder keys in Qwen3.5 save_pretrained." + "reason": "Different multimodal bugs: untied lm_head weights in Mistral3 vs incorrect visual encoder keys in Qwen3.5 save_pretrained." }, { "left": "issue:44610", "right": "issue:44898", "accept": false, - "reason": "Both concern vision input sizing, but they are different models and different expected resolutions." + "reason": "Both mention image sizing, but OmDet-Turbo processor output size and Perceiver positional interpolation are different failures." }, { "left": "issue:42915", "right": "issue:43317", "accept": false, - "reason": "Different failures: FineGrainedFP8Config on Qwen3Moe versus dequantized model loading with GPU+CPU offload." + "reason": "Quantization/FP8 failure in Qwen3Moe is separate from dequantized model offload loading on device_map=auto." }, { - "left": "issue:44423", - "right": "issue:45406", + "left": "issue:44734", + "right": "issue:46858", "accept": false, - "reason": "Both are serve/multimodal crashes, but the attribute errors and processor code paths are different." + "reason": "Serve KV-cache continuation indexing bug is different from the static-cache seq_length tensor issue in prefill/compile." }, { - "left": "issue:44734", - "right": "issue:46858", + "left": "issue:44423", + "right": "issue:45406", "accept": false, - "reason": "Both involve cache handling, but one is KV-cache continuation indexing in serve responses and the other is static-cache length arithmetic in prefill." + "reason": "Both affect transformers serve with multimodal models, but the concrete errors and processor failures are different (`str.to` vs missing `_tokenizer`)." }, { "left": "issue:46821", "right": "issue:46869", "accept": false, - "reason": "No overlap beyond generic device/model failure language; TimesFM window_size error is unrelated to YOLO multi-GPU device mapping." + "reason": "TimesFM's window_size AttributeError is unrelated to YOLO multi-GPU device_map loading." }, { "left": "issue:44008", "right": "issue:45823", "accept": false, - "reason": "Different Gemma variants and different bugs: a variable-name collision in forward versus a PLE device mismatch under `device_map=\"auto\"`." + "reason": "Different Gemma family bugs: variable-name collision in forward vs PLE device mismatch with auto device mapping." }, { "left": "issue:41720", "right": "issue:45161", "accept": false, - "reason": "Both are parallelism/device-map issues, but they affect different model stacks and failure modes (cuda assert vs tensor parallel not working)." + "reason": "Qwen3 auto device-map cudaErrorAssert and GPT-OSS MoE tensor-parallel failure are different model/runtime issues." }, { "left": "issue:28530", "right": "issue:33666", "accept": false, - "reason": "Unrelated: early stopping metric selection versus Qwen2-VL multi-GPU training." + "reason": "Early stopping metric selection has nothing to do with Qwen2-VL multi-GPU training." }, { "left": "issue:44464", "right": "issue:46858", "accept": false, - "reason": "Both mention compiled execution and generation/cache behavior, but the reported bugs are different enough to keep separate." + "reason": "Compiled forward chunked generation inconsistency is not the same bug as static-cache seq_length returning a device tensor." }, { "left": "issue:45632", "right": "issue:45684", "accept": false, - "reason": "Different save/load edge cases: cache-path collision for trust_remote_code versus read-only permission propagation during save_pretrained." + "reason": "Cache-path collision for local trust_remote_code models is unrelated to read-only permission propagation during save_pretrained." }, { "left": "issue:43531", "right": "issue:43572", "accept": false, - "reason": "Different model/config problems: Qwen3-MoE sliding-window behavior versus missing `pad_token_idx` in StableLmConfig." + "reason": "Qwen3-MoE sliding_window behavior and StableLmConfig pad_token_idx missing are separate config/model issues." }, { "left": "issue:44671", "right": "issue:44743", "accept": false, - "reason": "No shared underlying code path; one is CamemBERT masked-LM prediction drift, the other is Qwen3.5 recurrent-state reset with cache." + "reason": "CamemBERT masked-LM prediction regression and Qwen3.5 recurrent-state reset are unrelated." }, { - "left": "issue:42915", - "right": "issue:46291", + "left": "issue:46821", + "right": "issue:46869", "accept": false, - "reason": "Different subsystems: FP8 model loading versus an import failure in `transformers.pipelines` caused by kernels metadata requirements." + "reason": "TimesFM window_size attribute error does not match YOLO multi-GPU device_map auto failure." }, { - "left": "issue:36331", - "right": "issue:41720", + "left": "issue:44008", + "right": "issue:45823", "accept": false, - "reason": "Different trainer API mismatch versus model/device-map runtime assert." + "reason": "Same broader family, but different concrete failure modes and forward-path bugs." }, { - "left": "issue:44186", - "right": "issue:44361", + "left": "issue:463?", + "right": "issue:463?", "accept": false, - "reason": "Both are tokenizer-related, but they are different tokenizer classes and different failure scenarios." + "reason": "No additional soft edge provided." }, { "left": "issue:43425", "right": "issue:46821", "accept": false, - "reason": "Both are version-related, but one is Torch compatibility in general while the other is a model-specific `window_size` attribute error." + "reason": "Torch 2.10 compatibility concerns are unrelated to TimesFM's window_size attribute issue." } ] }, @@ -5069,7 +5061,7 @@ { "label": "issue", "component_index": 1, - "component_count": 25, + "component_count": 26, "review_unit_index": 16, "review_unit_count": 61, "status": "reviewed", @@ -5140,12 +5132,12 @@ "issue:30064|issue:34634", "issue:44534|issue:46665", "issue:43550|issue:43653", - "issue:43531|issue:45759", "issue:41720|issue:43746", + "issue:43531|issue:45759", "issue:43526|issue:43881", "issue:43606|issue:43720", - "issue:42490|issue:43329", "issue:44492|issue:44509", + "issue:42490|issue:43329", "issue:45406|issue:45464", "issue:43388|issue:45059", "issue:34567|issue:35707", @@ -5154,57 +5146,57 @@ "issue:43531|issue:46821", "issue:43540|issue:46858", "issue:45759|issue:45859", - "issue:44451|issue:44991", "issue:43531|issue:44589", + "issue:44451|issue:44991", "issue:45357|issue:46402", "issue:44077|issue:44625" ], - "prepared_review_unit_hash": "85a41635d870fb05fe805d1c3e697ba2eef22bce51996278f91034890a213746", + "prepared_review_unit_hash": "fb67cf9670f4c62cef4332a84b0c4830607ad4239bdd9452a24a449f1691bc90", "component_budget": { "node_count": 505, "item_count": 505, "soft_pair_count": 1854, - "serialized_chars": 962354, - "estimated_input_tokens": 240589, - "estimated_eval_tokens": 481434 + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 29, - "serialized_chars": 36368, - "estimated_input_tokens": 9092, - "estimated_eval_tokens": 18440 + "serialized_chars": 36369, + "estimated_input_tokens": 9093, + "estimated_eval_tokens": 18442 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "I don\u2019t see any true duplicate pairs here; the items span many unrelated Transformers bugs across tokenizers, configs, multimodal processors, offloading, and generation. I\u2019d keep the broad offloading/quantization report as the most representative issue, but I reject all soft-edge merges.", - "confidence": 0.79, - "canonical_issue_reason": "Issue 43873 is the broadest and most central report in the set, covering a generic offloading/quantization regression with multiple inbound references.", + "summary": "No soft pair looks like the same underlying bug/change. The items span unrelated subsystems (configs, tokenizers, vision, multimodal loading, offloading, docs), so this is not a real duplicate cluster.", + "confidence": 0.94, + "canonical_issue_reason": "issue:43873 is the least-specific, most hub-like report here (open, 6 inbound refs), so it is the best available anchor even though the set is not a true duplicate cluster.", "canonical_pr_reason": null, - "best_issue_reason": "43873 is the best standalone issue to anchor the cluster because it is broad, active, and plausibly covers the widest nearby surface area.", + "best_issue_reason": "issue:43873 is the best global issue candidate only as a routing anchor: it has the strongest ongoing signal and broadest symptom framing, but it still does not represent the rest of the items well.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43316", "right": "issue:44913", "accept": false, - "reason": "Both are config-related, but they describe different model/config fields and different failure modes; not the same bug." + "reason": "Different config bugs: Gemma3TextConfig API mismatch vs GPTNeoX rotary_pct not persisting on reload." }, { "left": "issue:28530", "right": "issue:30064", "accept": false, - "reason": "Early-stopping metric lookup and image-processor void segmentation maps are unrelated subsystems." + "reason": "Early stopping metric selection is unrelated to image processor void segmentation map handling." }, { "left": "issue:42915", "right": "issue:43322", "accept": false, - "reason": "Different model families and different errors: FP8 config failure vs Llava Next segfault." + "reason": "Different failures and code paths: Qwen3MoE FP8 config vs Llava Next loading segfault." }, { "left": "issue:33357", @@ -5216,151 +5208,157 @@ "left": "issue:43540", "right": "issue:43782", "accept": false, - "reason": "Both are Qwen multimodal loading/generation issues, but the concrete failures and code paths differ." + "reason": "Video input parsing error in Qwen3OmniMoe is not the same as Qwen3VL weight_only loading failure." }, { "left": "issue:29159", "right": "issue:36032", "accept": false, - "reason": "Tokenizer slow/fast inconsistency is not the same as T5 tokenizer class-name conflict on load." + "reason": "Tokenizer slow/fast inconsistency is a different bug from T5 add_special_tokens name conflict." }, { "left": "issue:43873", "right": "issue:45704", "accept": false, - "reason": "Offloading/quantization regression is different from a T5 apex RMSNorm memory leak." + "reason": "Quantization/offloading regression is unrelated to T5 apex.FusedRMSNorm memory leak." }, { "left": "issue:44387", "right": "issue:45701", "accept": false, - "reason": "Int4 quantization OOM and tokenization changes are separate regressions." + "reason": "Int4 CUDA reserved-memory/OOM regression is not the same as tokenization behavior changes." }, { "left": "issue:30064", "right": "issue:34634", "accept": false, - "reason": "Void segmentation-map processing and BarkProcessor voice_preset are unrelated image vs audio processor bugs." + "reason": "Void segmentation-map processing and BarkProcessor voice_preset are unrelated." }, { "left": "issue:44534", "right": "issue:46665", "accept": false, - "reason": "Non-persistent buffer junk in v5 and from_pretrained inside Python classes are different initialization behaviors." + "reason": "Non-persistent buffer initialization bug is different from from_pretrained behavior inside Python classes." }, { "left": "issue:43550", "right": "issue:43653", "accept": false, - "reason": "Bamba SDPA/torch.compile failure and BigBird mask-token registration are unrelated." + "reason": "torch.compile SDPA failure for Bamba is unrelated to BigBirdTokenizer special-token registration." }, { - "left": "issue:43531", - "right": "issue:45759", + "left": "issue:41720", + "right": "issue:43746", "accept": false, - "reason": "Sliding-window handling in Qwen3-MoE and config unwrapping for Qwen 3.5/3.6 are different bugs." + "reason": "Qwen3 auto device mapping CUDA assert is a different bug from GraniteSpeech PEFT local checkpoint loading." }, { - "left": "issue:41720", - "right": "issue:43746", + "left": "issue:43531", + "right": "issue:45759", "accept": false, - "reason": "Auto device-mapping cudaErrorAssert and GraniteSpeech PEFT checkpoint loading are not the same code path." + "reason": "Sliding-window Qwen3-MoE bug is unrelated to multimodal config unwrapping in AutoModelForCausalLM.from_config." }, { "left": "issue:43526", "right": "issue:43881", "accept": false, - "reason": "BeitImageProcessorFast label reduction and glm-4v-9b loading failure are unrelated." + "reason": "BeitImageProcessorFast label reduction bug is unrelated to glm-4v-9b loading failure." }, { "left": "issue:43606", "right": "issue:43720", "accept": false, - "reason": "Both involve loading/offload, but one is a device mismatch on CPU offload and the other is unpacking packed BitNet weights; not the same defect." + "reason": "CPU offload device mismatch in Bark is unrelated to BitNet packed-weight unpacking during accelerate loading." }, { - "left": "issue:42490", - "right": "issue:43329", + "left": "issue:44492", + "right": "issue:44509", "accept": false, - "reason": "Trailing-slash local model import warning is unrelated to undefined video-token counting variables." + "reason": "A cache-strategy typo is not the same as docs mentioning removed pipeline tasks." }, { - "left": "issue:44492", - "right": "issue:44509", + "left": "issue:42490", + "right": "issue:43329", "accept": false, - "reason": "A typo in cache strategies is not the same as outdated pipeline documentation." + "reason": "Trailing-slash local-path import warning is unrelated to undefined video-branch helpers in multimodal processors." }, { "left": "issue:45406", "right": "issue:45464", "accept": false, - "reason": "Gemma4Processor missing _tokenizer and Qwen3.5 streaming chat/completions failure are different API bugs." + "reason": "Gemma4Processor missing _tokenizer and Qwen3.5 streaming chat/completions failures are different issues." }, { "left": "issue:43388", "right": "issue:45059", "accept": false, - "reason": "Metric-gathering label truncation and SAM3 text/box behavior are unrelated." + "reason": "gather_for_metrics truncation of tuple labels is unrelated to SAM3 text/bbox behavior." }, { "left": "issue:34567", "right": "issue:35707", "accept": false, - "reason": "TrainerState token counter bug and progressive generation with inputs_embeds/past_key_values do not share a concrete failure path." + "reason": "TrainerState token counter not updating is not the same as progressive generation cache behavior." }, { "left": "issue:43606", "right": "issue:43653", "accept": false, - "reason": "CPU offload device mismatch and tokenizer special-token registration are unrelated." + "reason": "Bark CPU offload mismatch and BigBirdTokenizer decode issues do not share an underlying bug." }, { "left": "issue:34634", "right": "issue:41720", "accept": false, - "reason": "BarkProcessor voice_preset and Qwen3 auto device mapping are different components and errors." + "reason": "Bark voice_preset and Qwen3 auto device mapping failures are unrelated subsystems and symptoms." }, { "left": "issue:43531", "right": "issue:46821", "accept": false, - "reason": "Qwen3-MoE sliding_window and TimesFM window_size errors are superficially similar but not the same bug." + "reason": "Qwen3-MoE sliding-window bug is unrelated to TimesFM window_size attribute errors." }, { "left": "issue:43540", "right": "issue:46858", "accept": false, - "reason": "Qwen3OmniMoe video processing error and cache seq-length arithmetic on non-CUDA backends are unrelated." + "reason": "Qwen3OmniMoe video ValueError is unrelated to static-cache seq-length tensor arithmetic." }, { "left": "issue:45759", "right": "issue:45859", "accept": false, - "reason": "Qwen multimodal config unwrapping and CPU-offload rotary embedding behavior are separate issues." + "reason": "Config unwrapping for composite Qwen multimodal models is unrelated to CPU-offload incompatibility in rotary embedding forward." }, { - "left": "issue:44451", - "right": "issue:44991", + "left": "issue:43531", + "right": "issue:44589", "accept": false, - "reason": "Two different model/tokenizer loading failures; no shared concrete code-path is evident." + "reason": "Sliding-window handling and Float8 storage lookup failures are different code paths." }, { - "left": "issue:43531", - "right": "issue:44589", + "left": "issue:44451", + "right": "issue:44991", "accept": false, - "reason": "Sliding-window Qwen3-MoE behavior and float8 storage lookup errors are unrelated." + "reason": "ScandiBERT loading failure is not the same as est-roberta tokenizer changes under v5." }, { "left": "issue:45357", "right": "issue:46402", "accept": false, - "reason": "Wrong visual encoder keys on save and missing text encoder weights on load are different save/load regressions." + "reason": "Qwen3.5 visual encoder key saving regression is unrelated to Sam3 text encoder weights not loading." }, { "left": "issue:44077", "right": "issue:44625", "accept": false, - "reason": "patchtsmixer post_init policy and Qwen3.5 num_labels propagation are not the same underlying problem." + "reason": "patchtsmixer post_init policy and Qwen3.5 num_labels propagation are unrelated API problems." + }, + { + "left": "issue:44589", + "right": "issue:46821", + "accept": false, + "reason": "Float8 storage resolution failure is unrelated to TimesFM window_size AttributeError." } ] }, @@ -5374,7 +5372,7 @@ { "label": "issue", "component_index": 1, - "component_count": 25, + "component_count": 26, "review_unit_index": 17, "review_unit_count": 61, "status": "reviewed", @@ -5464,217 +5462,217 @@ "issue:44995|issue:46858", "issue:45356|issue:45907" ], - "prepared_review_unit_hash": "731cc84736856a5a11d77855f48770a6a433b66dd64d0d1447fc265b9634f269", + "prepared_review_unit_hash": "331e5ca6938ce89341c0dd167b71f655bfabe660adffd6f52c2b256d031bceaa", "component_budget": { "node_count": 505, "item_count": 505, "soft_pair_count": 1854, - "serialized_chars": 962354, - "estimated_input_tokens": 240589, - "estimated_eval_tokens": 481434 + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 29, - "serialized_chars": 36882, - "estimated_input_tokens": 9221, - "estimated_eval_tokens": 18698 + "serialized_chars": 36886, + "estimated_input_tokens": 9222, + "estimated_eval_tokens": 18700 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "Mostly a grab-bag of unrelated Transformers issues across tokenizers, vision, quantization, and training. The only near-duplicate-looking pair is 44514/44545, which share the same Qwen2_5_VLProcessor.apply_chat_template batched-input crash with padding=False.", - "confidence": 0.79, - "canonical_issue_reason": "issue:44514 is the best canonical issue for the Qwen2_5_VLProcessor.apply_chat_template padding=False crash because it is the earlier report and matches issue:44545 almost verbatim.", + "summary": "The cluster is mostly a set of unrelated Transformers issues with only one clear duplicate pair: the two Qwen2_5_VLProcessor batched padding=False crash reports. Most other soft-similarity pairs share only a broad subsystem or keyword, not the same concrete bug.", + "confidence": 0.87, + "canonical_issue_reason": "issue:44545 is the best canonical issue here: it is an exact duplicate of issue:44514, has the same concrete failure mode, and carries slightly more discussion/inbound reference signal.", "canonical_pr_reason": null, - "best_issue_reason": "issue:44514 is the clearest representative issue in this cluster because it is the strongest near-duplicate match and the most precise description of that shared crash.", + "best_issue_reason": "issue:44545 is the strongest standalone issue candidate in the cluster because it is the clearest exact duplicate, is highly specific, and appears most actionable for deduplication/triage.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43644", "right": "issue:46665", "accept": false, - "reason": "Different problems: non-persistent buffer initialization vs inconsistent from_pretrained behavior inside Python classes." + "reason": "Both mention model loading behavior, but one is about non-persistent buffers in Transformers 5.0.0 and the other is about from_pretrained behavior inside Python classes; different bugs." }, { "left": "issue:44062", "right": "issue:44451", "accept": false, - "reason": "Both involve loading, but one is an AddedToken TypeError and the other is a model-load failure for ScandiBERT." + "reason": "AddedToken constructor argument duplication and failing to load vesteinn/ScandiBERT are different failure modes with no shared concrete code path." }, { "left": "issue:44898", "right": "issue:45072", "accept": false, - "reason": "Unrelated subsystems: Perceiver image resizing vs dtype mismatches in SwitchTransformers/TimmWrapperModel." + "reason": "Perceiver resolution/interpolate_pos_encoding and bf16 dtype mismatches in SwitchTransformers/TimmWrapperModel are unrelated." }, { "left": "issue:44987", "right": "issue:45042", "accept": false, - "reason": "Different failures: a repo-specific load regression vs PIL image processors requiring torchvision." + "reason": "Loading a specific model package and an image-processor torchvision dependency regression are different issues." }, { "left": "issue:44625", "right": "issue:45479", "accept": false, - "reason": "Related to classification configs, but one is num_labels propagation in Qwen3.5 and the other is a zero-loss bug across multiple sequence-classification models." + "reason": "Both involve num_labels, but one is config propagation and the other is a degenerate loss bug in single-label classification; not the same bug." }, { "left": "issue:43381", "right": "issue:43844", "accept": false, - "reason": "Eval-mode gradient checkpointing behavior is distinct from the ZeRO-3 gradient explosion issue." + "reason": "Eval-mode gradient checkpointing restriction and ZeRO-3 gradient explosion are unrelated training-path problems." }, { "left": "issue:39692", "right": "issue:44610", "accept": false, - "reason": "Both touch vision examples/processors, but one is a SigLIP2 doc/quantization issue and the other is an OmDet-Turbo input-size mismatch." + "reason": "Both are model/processor mismatch reports, but they affect different models and different concrete processors." }, { "left": "issue:44368", "right": "issue:45357", "accept": false, - "reason": "A tie_word_embeddings warning in LoRA fine-tuning is not the same as incorrect visual-encoder keys saved by Qwen3.5." + "reason": "A tied-embeddings warning during LoRA fine-tuning is unrelated to incorrect visual-encoder keys saved by Qwen3.5." }, { "left": "issue:41720", "right": "issue:45704", "accept": false, - "reason": "Completely different bugs: CUDA assert during Qwen3 auto device mapping vs a T5/apex RMSNorm memory leak." + "reason": "CUDA assert during Qwen3 auto device mapping and a T5 RMSNorm memory leak are different bugs." }, { "left": "issue:43746", "right": "issue:46050", "accept": false, - "reason": "PEFT local checkpoint loading for GraniteSpeech is unrelated to sinq quantization save/load breakage." + "reason": "PEFT local-checkpoint loading and sinq quantized save/load breakage are different code paths." }, { "left": "issue:46655", "right": "issue:46693", "accept": false, - "reason": "Video processor padding bug vs FA2 long-context performance regression; different components and symptoms." + "reason": "Incorrect video padding in SmolVLM and FA2 long-context performance regression are not the same issue." }, { "left": "issue:43381", "right": "issue:45704", "accept": false, - "reason": "Eval-mode gradient checkpointing behavior is unrelated to the T5 RMSNorm leak." + "reason": "Gradient checkpointing eval-mode behavior and an apex RMSNorm leak are unrelated." }, { "left": "issue:43421", "right": "issue:43927", "accept": false, - "reason": "Runtime special-token post-processor updates are distinct from DiaConfig losing custom token IDs on save/load." + "reason": "Runtime post-processor updates for special tokens and DiaConfig losing custom token IDs after save/load are distinct tokenizer bugs." }, { "left": "issue:44918", "right": "issue:44936", "accept": false, - "reason": "Unpacking input embeddings in TRL SFT is a different failure than trainer.evaluate() after trainer.train()." + "reason": "TRL embedding-unpack failure and trainer.evaluate() failing after train() are different trainer issues." }, { "left": "issue:45704", "right": "issue:45923", "accept": false, - "reason": "Memory leak in T5/apex is unrelated to missing supports_gradient_checkpointing on a trust_remote_code model." + "reason": "T5 apex memory leak and a missing supports_gradient_checkpointing flag are unrelated." }, { "left": "issue:42886", "right": "issue:44843", "accept": false, - "reason": "Both hit HF_HUB_OFFLINE, but one is general cache loading and the other is a specific unconditional model_info() call in _patch_mistral_regex." + "reason": "Both are offline-mode related, but one is cache loading and the other is an unconditional model_info() call in _patch_mistral_regex; different root causes." }, { "left": "issue:45759", "right": "issue:45923", "accept": false, - "reason": "AutoModelForCausalLM config unwrapping for Qwen multimodal configs is not the same as a missing gradient-checkpointing flag." + "reason": "AutoModelForCausalLM config unwrapping and a missing gradient-checkpointing flag are unrelated." }, { "left": "issue:43746", "right": "issue:43873", "accept": false, - "reason": "Local checkpoint loading for PEFT adapters is not the same as quantization offloading behaving incorrectly." + "reason": "Local checkpoint loading with PEFT adapters and quantization/offloading behavior are different issues." }, { "left": "issue:44514", "right": "issue:44545", "accept": true, - "reason": "Near-verbatim titles and the same apply_chat_template batched-input crash with padding=False point to the same underlying bug." + "reason": "Exact same title and symptom: Qwen2_5_VLProcessor.apply_chat_template crashes on batched input when padding=False." }, { "left": "issue:43994", "right": "issue:44671", "accept": false, - "reason": "Different model-specific correctness bugs: SigLIP2 nonsensical outputs vs CamemBERT masked-LM predictions in v5." + "reason": "Both concern model outputs, but one is SigLIP2 and the other is CamemBERT masked-LM behavior; not the same bug." }, { "left": "issue:46032", "right": "issue:46463", "accept": false, - "reason": "Incorrect cache use in Mamba2Mixer is unrelated to a T5 decoder-input validation error message." + "reason": "Mamba2Mixer incorrect results with use_cache and a T5 decoder-input validation error message are different problems." }, { "left": "issue:39692", "right": "issue:43873", "accept": false, - "reason": "SigLIP2 documentation/example problems do not match quantization offloading behavior." + "reason": "SigLIP2 documentation/model mismatch and quantization offloading are unrelated." }, { "left": "issue:43421", "right": "issue:43653", "accept": false, - "reason": "Both concern tokenizers, but one is runtime post-processor updates and the other is BigBirdTokenizer special-token registration/decoding." + "reason": "Both touch tokenizer special-token handling, but one is runtime post-processor updates and the other is a BigBirdTokenizer mask-token registration bug." }, { "left": "issue:43782", "right": "issue:45381", "accept": false, - "reason": "Qwen3VL weight_only load failure is a different bug from Qwen2.5-VL video position ids." + "reason": "A from_pretrained weight_only error for Qwen3VL and wrong vision_position_ids in Qwen2.5-VL are different multimodal bugs." }, { "left": "issue:30064", "right": "issue:36010", "accept": false, - "reason": "Image processor void segmentation map handling is unrelated to a GenerationMixin import error." + "reason": "Image segmentation map processing and a GenerationMixin import error have no common code path." }, { "left": "issue:45081", "right": "issue:45356", "accept": false, - "reason": "Both involve mistral regex/tokenizer loading, but one is an AttributeError crash and the other is a Kimi-K2.5 codec-handling regression." + "reason": "Both involve tokenizer regex patching, but one is a crash in _patch_mistral_regex and the other is a codec/regression warning issue; not the same concrete bug." }, { "left": "issue:43421", "right": "issue:44186", "accept": false, - "reason": "Special-token post-processor updates are different from LayoutLMv2 tokenizer crashes on NER and batched padding/truncation." + "reason": "Special-token post-processor updates and LayoutLMv2 NER padding/truncation crashes are distinct tokenizer problems." }, { "left": "issue:44995", "right": "issue:46858", "accept": false, - "reason": "GlmMoeDsa stale indexer cache on second forward is unrelated to static-cache seq_length tensor arithmetic." + "reason": "Stale indexer cache on second forward pass and device-tensor seq-length arithmetic in static cache/torch.compile are different cache bugs." }, { "left": "issue:45356", "right": "issue:45907", "accept": false, - "reason": "Kimi-K2.5 tokenizer regression is unrelated to list_repo_files failing to forward revision in CI." + "reason": "Tokenizer regression in Kimi-K2.5 and list_repo_files missing revision forwarding are unrelated." } ] }, "evaluator_result": { - "accept": false, - "feedback": "The overall read is plausible, but it overstates the evidence in a few places. The packet does not show any explicit duplicate verification, so phrases like \"only verified duplicate target\" and \"more established report\" are stronger than the data supports. The 44514/44545 pair is a reasonable near-duplicate call based on the matching titles, but the rest of the summary should stay more conservative." + "accept": true, + "feedback": "The summary is grounded in the packet and stays conservative. The only positive duplicate callout, issue:44514 \u2194 issue:44545, is well supported by the identical title/symptom and the higher inbound-reference signal on issue:44545. The remaining soft-edge rejections do not overclaim shared bugs, and the cluster-level summary appropriately characterizes the rest as mostly unrelated." }, "evaluator_used": true, - "retried": true, + "retried": false, "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null @@ -5682,7 +5680,7 @@ { "label": "issue", "component_index": 1, - "component_count": 25, + "component_count": 26, "review_unit_index": 18, "review_unit_count": 61, "status": "reviewed", @@ -5756,14 +5754,14 @@ "issue:43596|issue:44220", "issue:43611|issue:46665", "issue:44186|issue:44898", - "issue:44008|issue:46858", "issue:43653|issue:43819", + "issue:44008|issue:46858", "issue:41720|issue:44991", "issue:45987|issue:46932", "issue:30064|issue:36331", - "issue:43643|issue:44877", "issue:44220|issue:44661", "issue:43618|issue:43697", + "issue:43643|issue:44877", "issue:43335|issue:43756", "issue:44843|issue:45081", "issue:43550|issue:43819", @@ -5774,20 +5772,20 @@ "issue:44186|issue:44442", "issue:43540|issue:44008" ], - "prepared_review_unit_hash": "a428c9dcba437a985e7945595562c1a56aca38503485f2ae936decc2303811c5", + "prepared_review_unit_hash": "6e0f8afc1c14e3c99e0befd2e30080671003a75f316495f0ae8bc5c75af696fd", "component_budget": { "node_count": 505, "item_count": 505, "soft_pair_count": 1854, - "serialized_chars": 962354, - "estimated_input_tokens": 240589, - "estimated_eval_tokens": 481434 + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 31, - "serialized_chars": 37230, + "serialized_chars": 37229, "estimated_input_tokens": 9308, "estimated_eval_tokens": 18872 }, @@ -5796,198 +5794,204 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "These are mostly unrelated Transformers bug reports rather than true duplicates. The only near-match is the Mistral regex/tokenizer pair, but they fail in different concrete ways and should stay separate.", - "confidence": 0.92, - "canonical_issue_reason": "issue:45081 is the most concrete, reproducible bug report in the set and a strong anchor for the Mistral tokenizer-regex loading path, with an explicit stack trace and clear trigger.", + "summary": "All proposed soft pairs are false positives: they share broad wording like loading, tokenizer, config, or vision/video, but each points to a different concrete bug or code path. This cluster is too heterogeneous to have a single canonical duplicate.", + "confidence": 0.96, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "issue:45081 is the clearest single canonical candidate here: it names the exact failing helper, has a precise repro, and is better documented than the other nearby tokenizer-loading reports.", + "best_issue_reason": null, "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43720", "right": "issue:43819", "accept": false, - "reason": "Different bugs: BitNet accelerate-loading unpacking vs DAC latent conversion/STE mismatch." + "reason": "Different subsystems and failures: BitNet accelerate weight unpacking vs DAC latent reconstruction/STE behavior." }, { "left": "issue:46829", "right": "issue:46832", "accept": false, - "reason": "Different models and code paths: VideoPrism output field wiring vs RT-DETR feature-level handling." + "reason": "Both are vision-model issues, but one is an output-field mismatch and the other is an RT-DETR feature-level shape/type error." }, { "left": "issue:43653", "right": "issue:43720", "accept": false, - "reason": "Tokenizer special-token registration bug is unrelated to BitNet packed-weight loading." + "reason": "Tokenizer special-token registration bug vs quantized linear loading/unpacking bug; unrelated code paths." }, { "left": "issue:45987", "right": "issue:46829", "accept": false, - "reason": "StaticCache seq-length contract issue is unrelated to VideoPrism output labeling." + "reason": "Cache sequence-length return type issue vs VideoPrism output tensor assignment; not the same underlying bug." }, { "left": "issue:43295", "right": "issue:45834", "accept": false, - "reason": "Processor/tokenizer regression and Kosmos2.5 long-OCR index error are different failures." + "reason": "Processor/tokenizer regression in v4.57.5 vs Kosmos2.5 OCR long-input indexing; different models and failure modes." }, { "left": "issue:44062", "right": "issue:44488", "accept": false, - "reason": "AddedToken keyword collision is not the same as loading a specific tokenizer model." + "reason": "AddedToken keyword collision vs loading a specific tokenizer/model; the reported errors do not share a concrete code path." }, { "left": "issue:43550", "right": "issue:43927", "accept": false, - "reason": "Bamba torch.compile/SDPA failure is unrelated to DiaConfig save/load custom token IDs." + "reason": "torch.compile/SDPA issue in Bamba vs DiaConfig losing custom token IDs on save/load." }, { "left": "issue:43550", "right": "issue:45072", "accept": false, - "reason": "Both are model/runtime issues, but SDPA compile failure and dtype mismatch are different bugs." + "reason": "Both involve runtime/dtype behavior, but the concrete bugs are different models and different failure mechanisms." }, { "left": "issue:43540", "right": "issue:46294", "accept": false, - "reason": "Both involve video inputs, but Qwen3OmniMoe processing and Qwen3.5 chat-template token/feature alignment are different code paths." + "reason": "Both are video-related, but one is a Qwen3OmniMoe video-processing ValueError and the other is a chat-template token/feature mismatch." }, { "left": "issue:45381", "right": "issue:46693", "accept": false, - "reason": "Qwen2.5-VL video position IDs and FA2 long-context performance regression are unrelated." + "reason": "Qwen2.5-VL video position IDs bug vs FA2 long-context performance regression; unrelated." }, { "left": "issue:43531", "right": "issue:43643", "accept": false, - "reason": "Sliding-window MoE behavior and trust_remote_code config field loss are unrelated." + "reason": "Sliding-window handling in Qwen3-MoE vs AutoConfig trust_remote_code field loss; different layers of the stack." }, { "left": "issue:43596", "right": "issue:44220", "accept": false, - "reason": "Deepspeed zero3/BertModel init failure is unrelated to fbank feature extraction." + "reason": "DeepSpeed ZeRO-3/BertModel init failure vs audio fbank feature extraction bug." }, { "left": "issue:43611", "right": "issue:46665", "accept": false, - "reason": "Both concern model loading semantics, but base_model_prefix regression and from_pretrained-in-classes inconsistency are different issues." + "reason": "Both touch loading, but one is base_model_prefix compatibility in Transformers 5 and the other is from_pretrained behavior inside Python classes." }, { "left": "issue:44186", "right": "issue:44898", "accept": false, - "reason": "LayoutLMv2 tokenizer padding/NER crash and Perceiver interpolation failure are unrelated." + "reason": "Tokenizer crash on NER/padding vs Perceiver image classification resolution failure; different model families and code paths." }, { - "left": "issue:44008", - "right": "issue:46858", + "left": "issue:43653", + "right": "issue:43819", "accept": false, - "reason": "Gemma3n variable collision in forward() is unrelated to StaticCache seq-length tensor arithmetic." + "reason": "BigBird tokenizer special-token registration vs DAC.from_latents mismatch; no shared underlying defect." }, { - "left": "issue:43653", - "right": "issue:43819", + "left": "issue:44008", + "right": "issue:46858", "accept": false, - "reason": "BigBirdTokenizer special-token decode bug is unrelated to DAC from_latents/STE behavior." + "reason": "Gemma3n forward variable collision vs cache.get_seq_length returning a device tensor; unrelated bugs." }, { "left": "issue:41720", "right": "issue:44991", "accept": false, - "reason": "Qwen3 auto device mapping CUDA assert and est-roberta tokenizer loading are unrelated." + "reason": "Qwen3 auto device mapping CUDA assert vs tokenizer loading failure for a specific model." }, { "left": "issue:45987", "right": "issue:46932", "accept": false, - "reason": "StaticCache return-type bug is unrelated to Gemma4 quantized embedder dtype casting." + "reason": "StaticCache return-type contract bug vs Gemma4 quantized embedder dtype casting; different components." }, { "left": "issue:30064", "right": "issue:36331", "accept": false, - "reason": "Void segmentation-map processing and CustomTrainer compute_loss signature mismatch are unrelated." - }, - { - "left": "issue:43643", - "right": "issue:44877", - "accept": false, - "reason": "Both are config-loading issues, but missing fields from trust_remote_code and strict granite_speech config validation are different bugs." + "reason": "ImageProcessor segmentation-map processing vs CustomTrainer.compute_loss signature change; no overlap in code path." }, { "left": "issue:44220", "right": "issue:44661", "accept": false, - "reason": "fbank feature extraction and TOKENIZER_MAPPING_NAMES model-addition failure are unrelated." + "reason": "Audio feature extraction helper vs auto-model registration failure in TOKENIZER_MAPPING_NAMES." }, { "left": "issue:43618", "right": "issue:43697", "accept": false, - "reason": "CLIPOutput attentions assignment and RTDetrV2 output differences are not the same failure." + "reason": "CLIPOutput attentions field regression vs RTDetrV2 output differences; both are v5 regressions but not the same bug." + }, + { + "left": "issue:43643", + "right": "issue:44877", + "accept": false, + "reason": "AutoConfig trust_remote_code missing fields vs strict config blocking granite_speech loading; different symptoms and fixes." }, { "left": "issue:43335", "right": "issue:43756", "accept": false, - "reason": "SwitchTransformers sparse-layer construction and Smollm3 RoPE-layer counting are different architecture/config bugs." + "reason": "SwitchTransformers sparse-layer construction bug vs Smollm3 RoPE-layer dropping bug." }, { "left": "issue:44843", "right": "issue:45081", "accept": false, - "reason": "Both hit _patch_mistral_regex, but one is an offline model_info call and the other is a backend_tokenizer attribute error; different triggers and fixes." + "reason": "Both involve _patch_mistral_regex, but one is an offline model_info() network call and the other is a backend_tokenizer AttributeError." }, { "left": "issue:43550", "right": "issue:43819", "accept": false, - "reason": "Bamba compile/SDPA and DAC latent reconstruction are unrelated model bugs." + "reason": "Same as above pair: different model/runtime bug vs DAC latent reconstruction bug." }, { "left": "issue:39692", "right": "issue:45834", "accept": false, - "reason": "SigLIP2 doc-example issues and Kosmos2.5 OCR index errors are unrelated." + "reason": "Documentation/example errors for SigLIP2 vs Kosmos2.5 OCR long-input indexing; not the same issue." }, { "left": "issue:44821", "right": "issue:45084", "accept": false, - "reason": "AutoImageProcessor URL loading and template-node compilation errors are unrelated." + "reason": "AutoImageProcessor URL loading vs a template compilation TypeError; unrelated functionality." }, { "left": "issue:44448", "right": "issue:44671", "accept": false, - "reason": "Pegasus v4/v5 output differences and CamemBERT masked-LM prediction errors are different bugs." + "reason": "Different model regressions in v5: Pegasus output change vs CamemBERT masked-LM prediction change." }, { "left": "issue:34634", "right": "issue:36010", "accept": false, - "reason": "BarkProcessor voice_preset and GenerationMixin import failure are unrelated." + "reason": "BarkProcessor voice_preset bug vs GenerationMixin import failure; unrelated." }, { "left": "issue:44186", "right": "issue:44442", "accept": false, - "reason": "LayoutLMv2 tokenizer crash and FastSpeech2Conformer AutoTokenizer load failure are unrelated." + "reason": "LayoutLMv2 tokenizer padding/NER crash vs AutoTokenizer failing to load a FastSpeech2Conformer tokenizer." }, { "left": "issue:43540", "right": "issue:44008", "accept": false, - "reason": "Qwen3OmniMoe video processing and Gemma3n audio_mel_mask collision are unrelated." + "reason": "Qwen3OmniMoe video-input handling vs Gemma3n variable-name collision in forward(); different models and bugs." + }, + { + "left": "issue:45381", + "right": "issue:46693", + "accept": false, + "reason": "Qwen video-position-id bug vs FA2 long-context performance regression; no shared code-path." } ] }, @@ -6001,7 +6005,7 @@ { "label": "issue", "component_index": 1, - "component_count": 25, + "component_count": 26, "review_unit_index": 19, "review_unit_count": 61, "status": "reviewed", @@ -6062,17 +6066,17 @@ "soft_pairs": [ "issue:45584|issue:45834", "issue:45183|issue:45479", - "issue:43644|issue:43749", "issue:43322|issue:45685", + "issue:43644|issue:43749", "issue:43606|issue:45072", "issue:43782|issue:44977", "issue:43421|issue:44811", - "issue:43577|issue:43818", - "issue:44479|issue:45834", "issue:43582|issue:46512", + "issue:44479|issue:45834", + "issue:43577|issue:43818", "issue:43388|issue:46032", - "issue:43825|issue:43994", "issue:44561|issue:45020", + "issue:43825|issue:43994", "issue:44991|issue:46291", "issue:46294|issue:46463", "issue:43716|issue:44008", @@ -6083,25 +6087,25 @@ "issue:43749|issue:44291", "issue:28218|issue:29159", "issue:43720|issue:45072", - "issue:43825|issue:43827", "issue:43645|issue:46665", + "issue:43825|issue:43827", "issue:44671|issue:45701", "issue:43577|issue:43582" ], - "prepared_review_unit_hash": "008c7532b4c83256d6f3bcb9e0d6f452e78b49e81d7c88cc58262c680d3d6b5d", + "prepared_review_unit_hash": "f8a68d1e059e85628e4cd4b1f63b0417852e15d58f5048bce1d1e60c3b281cfe", "component_budget": { "node_count": 505, "item_count": 505, "soft_pair_count": 1854, - "serialized_chars": 962354, - "estimated_input_tokens": 240589, - "estimated_eval_tokens": 481434 + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 }, "budget": { "node_count": 47, "item_count": 47, "soft_pair_count": 27, - "serialized_chars": 35591, + "serialized_chars": 35589, "estimated_input_tokens": 8898, "estimated_eval_tokens": 18052 }, @@ -6110,180 +6114,180 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This set is mostly heterogeneous regression reports rather than one tight duplicate cluster. The only strong duplicate-looking pair is the tokenizer spacing issue vs the broader slow/fast tokenizer inconsistency report; the rest are related only at a high level (tokenization, dtype, loading, video inputs, etc.) and should stay separate.", - "confidence": 0.74, - "canonical_issue_reason": "issue:29159 is the broader tokenizer-regression report and best matches the narrower added-token spacing symptom from issue:28218.", + "summary": "This cluster is mostly a grab bag of unrelated tokenizer, loading, video, and dtype issues. The only likely duplicate pair is the tokenizer added-token spacing report and the slow/fast tokenizer inconsistency report.", + "confidence": 0.86, + "canonical_issue_reason": "issue:28218 is the most specific statement of the tokenizer spacing bug and likely captures the underlying symptom behind the broader tokenizer inconsistency report.", "canonical_pr_reason": null, - "best_issue_reason": "issue:29159 is the best dedupe anchor here because it is broad enough to subsume the accepted tokenizer duplicate and is more representative than the narrower symptom report.", + "best_issue_reason": "issue:28218 is the best representative for the only clear duplicate subcluster in this set; the remaining issues are not close enough to share a single root cause.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:45584", "right": "issue:45834", "accept": false, - "reason": "Both are input/edge-case errors, but one is Whisper empty transcription handling and the other is Kosmos2.5 long OCR indexing; different bug paths." + "reason": "Both are input/decoding failures, but they involve different models and different code paths (Whisper empty transcription vs Kosmos2.5 OCR length/indexing)." }, { "left": "issue:45183", "right": "issue:45479", "accept": false, - "reason": "Input validation XOR messaging is unrelated to the single-label-classification zero-loss bug." + "reason": "One is an input validation error-message problem; the other is a degenerate loss bug in single-label classification. Not the same bug." }, { - "left": "issue:43644", - "right": "issue:43749", + "left": "issue:43322", + "right": "issue:45685", "accept": false, - "reason": "Different loading regressions: non-persistent buffer initialization vs broken FSDP CPU RAM-efficient loading." + "reason": "Segfault loading Llava Next and an MPS histogram limitation are unrelated failures." }, { - "left": "issue:43322", - "right": "issue:45685", + "left": "issue:43644", + "right": "issue:43749", "accept": false, - "reason": "Completely different models and failures: Llava Next loading segfault vs MPS histogram/int type error." + "reason": "Both concern loading, but one is junk-filled non-persistent buffers and the other is broken FSDP CPU RAM efficient loading; different root causes." }, { "left": "issue:43606", "right": "issue:45072", "accept": false, - "reason": "Both mention mismatches, but one is CPU offload device placement and the other is bfloat16 dtype mismatch in different model paths." + "reason": "CPU offload device mismatch in bark-small is unrelated to bfloat16 dtype mismatches in other models." }, { "left": "issue:43782", "right": "issue:44977", "accept": false, - "reason": "Qwen3VL weight_only loading error and Qwen3.5 flash-attention generation failure are separate code paths." + "reason": "Qwen3VL weight_only loading error and Qwen3.5 flash-attention generation failure are different concrete failures." }, { "left": "issue:43421", "right": "issue:44811", "accept": false, - "reason": "Special-token post-processor updates and Whisper batch_decode skip_special_tokens behavior are different tokenizer/processor bugs." + "reason": "Both touch tokenizers, but runtime special-token post-processor updates and Whisper batch_decode skip_special_tokens are different behaviors." }, { - "left": "issue:43577", - "right": "issue:43818", + "left": "issue:43582", + "right": "issue:46512", "accept": false, - "reason": "Blip2 dtype loading and Video-LLaVA temporal-attention/weight-sharing issues do not describe the same defect." + "reason": "Apple Silicon warmup TypeError and stale config.dtype after from_config are different dtype/config issues." }, { "left": "issue:44479", "right": "issue:45834", "accept": false, - "reason": "Both are video/OCR-related regressions, but they affect different models and failure modes." + "reason": "Different regressions: multi-model video-input processing versus long OCR indexing on Kosmos2.5." }, { - "left": "issue:43582", - "right": "issue:46512", + "left": "issue:43577", + "right": "issue:43818", "accept": false, - "reason": "Apple Silicon warmup TypeError is unrelated to stale config.dtype after from_config(dtype=...)." + "reason": "Blip2 dtype retention and Video-LLaVA missing temporal attention are unrelated model-loading/model-architecture bugs." }, { "left": "issue:43388", "right": "issue:46032", "accept": false, - "reason": "Metric gathering on tuple labels and Mamba2Mixer use_cache correctness are unrelated bugs." + "reason": "One is metrics gathering dropping tuple labels; the other is Mamba2 use_cache correctness. Not the same issue." }, { - "left": "issue:43825", - "right": "issue:43994", + "left": "issue:44561", + "right": "issue:45020", "accept": false, - "reason": "Pipeline translation error messaging and SigLIP2 nonsensical outputs are not the same underlying issue." + "reason": "Both mention v5/remote_code breakages, but one is specifically is_torch_fx_available removal and the other is a broad umbrella report." }, { - "left": "issue:44561", - "right": "issue:45020", + "left": "issue:43825", + "right": "issue:43994", "accept": false, - "reason": "The removed is_torch_fx_available symbol is a specific breakage; the other is a broader remote_code regression without enough evidence of the same root cause." + "reason": "An error-message wording issue and a model output correctness issue are not the same bug." }, { "left": "issue:44991", "right": "issue:46291", "accept": false, - "reason": "Tokenizer loading failure for one model and pipeline import failure due to kernels metadata are separate problems." + "reason": "Tokenizer loading failure for one model and kernels dependency/version import failure are unrelated." }, { "left": "issue:46294", "right": "issue:46463", "accept": false, - "reason": "Video chat-template feature/token mismatch is unrelated to T5 decoder input validation messaging." + "reason": "Qwen3.5 video chat template mismatch and T5 decoder-input validation messaging are different models and code paths." }, { "left": "issue:43716", "right": "issue:44008", "accept": false, - "reason": "Image preprocessor dtype mismatch and Gemma3n audio_mel_mask variable collision are distinct model bugs." + "reason": "dtype mismatch in Mistral-3 preprocessing/model loading is unrelated to Gemma3n's variable-name collision." }, { "left": "issue:43931", "right": "issue:46846", "accept": false, - "reason": "A Qwen3-VL weight-shape load error is not the same as missing linear_attn entries in a TP plan." + "reason": "Qwen3-VL weight-shape mismatch on load is not the same as missing tensor-parallel plan entries causing OOM/shape errors." }, { "left": "issue:44805", "right": "issue:44936", "accept": false, - "reason": "Mask/index shape mismatch and trainer.evaluate() failing after train() are too generic to treat as the same bug." + "reason": "A tensor-mask shape IndexError and trainer.evaluate() failing after train() are different failure modes." }, { "left": "issue:43540", "right": "issue:45381", "accept": false, - "reason": "Both involve video input, but Qwen3OmniMoe video processing and Qwen2.5-VL vision_position_ids are different regressions." + "reason": "Both concern video inputs, but Qwen3OmniMoe processing ValueError and qwen2.5-vl vision_position_ids regression are distinct." }, { "left": "issue:44610", "right": "issue:45072", "accept": false, - "reason": "OmDet-Turbo image-size mismatch is unrelated to the dtype mismatch issue." + "reason": "Processor image-size mismatch and dtype mismatch are different bugs." }, { "left": "issue:43749", "right": "issue:44291", "accept": false, - "reason": "Broken FSDP CPU RAM-efficient loading and init_empty_weights TypeError are separate model-loading failures." + "reason": "Broken FSDP CPU RAM efficient loading and an init_empty_weights unexpected-argument TypeError are different loading-path issues." }, { "left": "issue:28218", "right": "issue:29159", "accept": true, - "reason": "These describe the same tokenizer inconsistency: added tokens getting extra spacing fits the broader slow-vs-fast tokenizer behavior difference." + "reason": "Both describe the same tokenizer symptom: added-token handling differs between tokenizer implementations and introduces an extra/incorrect space." }, { "left": "issue:43720", "right": "issue:45072", "accept": false, - "reason": "BitNet packed-weight unpacking during accelerate loading is unrelated to dtype mismatches in other models." + "reason": "BitNet packed-weight unpacking during accelerate loading is unrelated to bfloat16 dtype mismatches." }, { - "left": "issue:43825", - "right": "issue:43827", + "left": "issue:43645", + "right": "issue:46665", "accept": false, - "reason": "One is an error-message bug, the other is outdated docs; they are related to v5 pipeline removal but not the same defect." + "reason": "Custom model initialization in notebooks and from_pretrained behavior inside Python classes are different workflow bugs." }, { - "left": "issue:43645", - "right": "issue:46665", + "left": "issue:43825", + "right": "issue:43827", "accept": false, - "reason": "Jupyter custom-model initialization breakage and from_pretrained-in-class inconsistency are different initialization scenarios." + "reason": "One is an incorrect pipeline error message; the other is stale documentation. Same theme, not the same underlying bug." }, { "left": "issue:44671", "right": "issue:45701", "accept": false, - "reason": "CamemBERT MLM prediction regression and a general tokenization-version change are not the same bug." + "reason": "CamemBERT masked-LM prediction regression and a version-dependent tokenization change are not clearly the same code-path defect." }, { "left": "issue:43577", "right": "issue:43582", "accept": false, - "reason": "dtype propagation on Blip2 loading and Apple Silicon caching_allocator_warmup TypeError are unrelated." + "reason": "A model dtype problem and an Apple Silicon warmup TypeError are unrelated." } ] }, "evaluator_result": { "accept": true, - "feedback": "The summary is grounded in the packet and stays conservative. The only accepted duplicate pair, issue:28218 and issue:29159, is plausibly supported by the titles and the broader/narrower framing. The other pair verdicts appropriately avoid overclaiming cross-issue duplication and mostly explain why the symptoms or models differ." + "feedback": "The summary is grounded in the packet and stays conservative. The only clearly supported duplicate pair is issue:28218 / issue:29159, and the other soft-edge rejections do not overstate sameness of bug. The canonical/best issue choice is reasonable given the evidence." }, "evaluator_used": true, "retried": false, @@ -6294,7 +6298,7 @@ { "label": "issue", "component_index": 1, - "component_count": 25, + "component_count": 26, "review_unit_index": 20, "review_unit_count": 61, "status": "reviewed", @@ -6308,7 +6312,6 @@ "issue:36010", "issue:36331", "issue:42175", - "issue:42490", "issue:43257", "issue:43262", "issue:43329", @@ -6370,9 +6373,9 @@ "issue:43550|issue:44912", "issue:43452|issue:44991", "issue:43257|issue:43329", - "issue:43526|issue:43697", "issue:44060|issue:44368", "issue:36010|issue:43645", + "issue:43526|issue:43697", "issue:43906|issue:46209", "issue:36010|issue:36331", "issue:43329|issue:45600", @@ -6381,211 +6384,204 @@ "issue:43819|issue:43927", "issue:43531|issue:45042", "issue:43638|issue:44936", - "issue:42175|issue:46291", - "issue:42490|issue:43257" + "issue:42175|issue:46291" ], - "prepared_review_unit_hash": "0e090d9b94e78b0fdc117d8034608c6c7d3358fc88accf4b02ebb4e88f618c2b", + "prepared_review_unit_hash": "a50d7c4635a2c41f1989a1d3a1e08e448151015bc7f8c255a01df1e42644822c", "component_budget": { "node_count": 505, "item_count": 505, "soft_pair_count": 1854, - "serialized_chars": 962354, - "estimated_input_tokens": 240589, - "estimated_eval_tokens": 481434 + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 }, "budget": { - "node_count": 48, - "item_count": 48, - "soft_pair_count": 29, - "serialized_chars": 36862, - "estimated_input_tokens": 9216, - "estimated_eval_tokens": 18688 + "node_count": 47, + "item_count": 47, + "soft_pair_count": 28, + "serialized_chars": 35990, + "estimated_input_tokens": 8998, + "estimated_eval_tokens": 18252 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly a set of unrelated Transformer bugs/regressions rather than true duplicates. The proposed soft edges are largely broad-subsystem or same-family similarities, but not the same concrete code-path problem.", - "confidence": 0.93, - "canonical_issue_reason": "No true canonical duplicate emerges from this heterogeneous set; issue:45020 is the broadest umbrella regression and the best single representative if one must be chosen.", + "summary": "These items are mostly unrelated model-loading, processor, trainer, and backend regressions. No soft pair looks like the same concrete bug/change, so none should be merged as duplicates. If a single issue must anchor the set, the broadest umbrella is the remote_code loading regression.", + "confidence": 0.79, + "canonical_issue_reason": "issue:45020 is the broadest, most umbrella-like regression in the list and can plausibly subsume several model-loading breakages better than the highly specific issues.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45020 has the widest user impact and the strongest umbrella framing among the items, so it is the least bad representative of the cluster.", + "best_issue_reason": "issue:45020 is the best representative because it describes a wide loader regression rather than a model-specific symptom, making it the most plausible cluster anchor.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:34634", "right": "issue:36331", "accept": false, - "reason": "Different subsystems and failures: Bark voice preset handling vs Trainer compute_loss signature mismatch." + "reason": "Different subsystems and failure modes: Bark voice preset handling vs CustomTrainer loss signature mismatch." }, { "left": "issue:42175", "right": "issue:45081", "accept": false, - "reason": "Packaging/backend dependency issue vs Mistral tokenizer regex crash; no shared code-path." + "reason": "Unrelated packaging/backend dependency issue vs Mistral tokenizer regex crash." }, { "left": "issue:46175", "right": "issue:46932", "accept": false, - "reason": "Both Gemma4-related, but one is a huge vision patch-embedder memory issue and the other is quantized embedder dtype casting; different bugs." + "reason": "Both are Gemma4-related, but one is a large vision patch embedding memory issue and the other is dtype casting for quantized embedders; not the same bug." }, { "left": "issue:46697", "right": "issue:46728", "accept": false, - "reason": "Both involve image sizing/padding, but they affect different models and different logic paths." + "reason": "Different image processors and different padding/counting logic; no shared concrete code path." }, { "left": "issue:45003", "right": "issue:46285", "accept": false, - "reason": "Unsafe sys.modules access is unrelated to CLIP text-model structure inconsistency." + "reason": "sys.modules access bug in modeling_utils is not the same as CLIP text-model structure divergence." }, { "left": "issue:43782", "right": "issue:45325", "accept": false, - "reason": "Different Qwen/VL bugs: weight_only loading vs rope-index scaling." + "reason": "Different Qwen VL generations and different failure mechanisms (weight_only loading vs rope index scaling)." }, { "left": "issue:43653", "right": "issue:44912", "accept": false, - "reason": "Tokenizer special-token registration vs MXFP4 quantization fallback are unrelated." + "reason": "Tokenizer special-token registration bug vs MXFP4 quantization fallback; unrelated." }, { "left": "issue:43262", "right": "issue:43819", "accept": false, - "reason": "Audio chat-template sampling-rate default vs DAC latent reconstruction mismatch are different bugs." + "reason": "Audio chat-template sampling-rate default bug vs DAC latent/STE mismatch; different code paths." }, { "left": "issue:43582", "right": "issue:44857", "accept": false, - "reason": "Apple Silicon warmup TypeError vs CUDA AMP loss crash; unrelated code paths." + "reason": "Apple Silicon allocator TypeError is unrelated to CUDA AMP loss crash." }, { "left": "issue:45372", "right": "issue:46326", "accept": false, - "reason": "Both mention mistral_common/Gemma, but one is an import dependency break and the other is continue_final_message behavior." + "reason": "Gemma4 processor import issue from mistral_common is different from continue_final_message behavior in mistral_common tokenizers." }, { "left": "issue:45081", "right": "issue:45406", "accept": false, - "reason": "Tokenizer regex patch crash and Gemma4Processor missing _tokenizer are different failures." + "reason": "Mistral tokenizer regex patch crash and Gemma4Processor missing _tokenizer are separate processor/tokenizer bugs." }, { "left": "issue:43742", "right": "issue:45020", "accept": false, - "reason": "A specific model loading KeyError is not the same as the broad remote_code breakage." + "reason": "A specific MobileLLM key error is not the same as the broader remote_code regression, even if both are loading-related." }, { "left": "issue:44589", "right": "issue:45042", "accept": false, - "reason": "Float8 storage lookup failure vs PIL backend torchvision dependency issue are unrelated." + "reason": "Float8 storage lookup error and PIL/torchvision backend requirement are unrelated." }, { "left": "issue:43550", "right": "issue:44912", "accept": false, - "reason": "Bamba torch.compile SDPA failure is unrelated to MXFP4 quantization fallback." + "reason": "torch.compile+SDPA failure in Bamba is unrelated to MXFP4 quantization fallback." }, { "left": "issue:43452", "right": "issue:44991", "accept": false, - "reason": "Both are loading regressions, but they affect different entry points and have different root causes." + "reason": "Different tokenizer-loading breakages with different triggers; not the same concrete bug." }, { "left": "issue:43257", "right": "issue:43329", "accept": false, - "reason": "Qwen3 MoE weight conversion under deepspeed is unrelated to multimodal token counting undefined-name bugs." - }, - { - "left": "issue:43526", - "right": "issue:43697", - "accept": false, - "reason": "BEiT reduce_labels and RTDetr output divergence are separate vision-model issues." + "reason": "Qwen3 MoE weight conversion under accelerate/deepspeed is unrelated to an undefined video-token helper path." }, { "left": "issue:44060", "right": "issue:44368", "accept": false, - "reason": "Both are warning-related and Qwen-family, but they describe different models and different warning mechanisms." + "reason": "Both mention tied-weights warnings, but one is a model wiring bug and the other is a config warning during LoRA fine-tuning; not the same underlying issue." }, { "left": "issue:36010", "right": "issue:43645", "accept": false, - "reason": "GenerationMixin import error and custom-model Jupyter init breakage are different regressions." + "reason": "GenerationMixin import-path breakage is not the same as Jupyter custom-model initialization failures." + }, + { + "left": "issue:43526", + "right": "issue:43697", + "accept": false, + "reason": "BEiT label reduction bug and RTDetrV2 output-difference regression are different model/processor issues." }, { "left": "issue:43906", "right": "issue:46209", "accept": false, - "reason": "An isolated reproduction of an older issue is unrelated to FP8 inference failure on B200." + "reason": "A reproduction of an old issue is not the same as FP8 inference failure on B200." }, { "left": "issue:36010", "right": "issue:36331", "accept": false, - "reason": "ImportError from transformers.generation is unrelated to Trainer compute_loss argument changes." + "reason": "Import path breakage for GenerationMixin and Trainer compute_loss signature mismatch are unrelated." }, { "left": "issue:43329", "right": "issue:45600", "accept": false, - "reason": "Undefined video-branch helpers and stale auto_mappings config references are different CI/runtime failures." + "reason": "Undefined multimodal-token helpers and stale auto_mappings references are different maintenance bugs." }, { "left": "issue:43526", "right": "issue:44554", "accept": false, - "reason": "BEiT label reduction bug is unrelated to the MPS attention correctness issue." + "reason": "Image label reduction and MPS attention correctness are unrelated." }, { "left": "issue:45084", "right": "issue:45538", "accept": false, - "reason": "Template compilation error and CLIPTokenizer max-length default are not the same bug." + "reason": "Template compilation TypeError and CLIPTokenizer max_length constant are different problems." }, { "left": "issue:43819", "right": "issue:43927", "accept": false, - "reason": "DAC forward-vs-from_latents mismatch and DiaConfig save/load token-ID loss are unrelated audio-model problems." + "reason": "DAC latent reconstruction mismatch and DiaConfig token-ID persistence are unrelated." }, { "left": "issue:43531", "right": "issue:45042", "accept": false, - "reason": "Qwen3-MoE sliding-window behavior is unrelated to PIL backend torchvision requirements." + "reason": "Qwen3-MoE sliding-window behavior is unrelated to PIL backend dependency requirements." }, { "left": "issue:43638", "right": "issue:44936", "accept": false, - "reason": "Zero3 empty-index error during training and train/evaluate lifecycle breakage are different trainer bugs." + "reason": "Zero3/non-pretrained BERT index error and evaluate-after-train failure are not the same code-path bug." }, { "left": "issue:42175", "right": "issue:46291", "accept": false, - "reason": "Torch backend install omission is unrelated to kernels version/revision import failure." - }, - { - "left": "issue:42490", - "right": "issue:43257", - "accept": false, - "reason": "Trailing-slash local path warning is unrelated to Qwen3 MoE weight conversion." + "reason": "Packaging extras missing TensorFlow is unrelated to kernels version/revision import failures." } ] }, @@ -6599,7 +6595,7 @@ { "label": "issue", "component_index": 1, - "component_count": 25, + "component_count": 26, "review_unit_index": 21, "review_unit_count": 61, "status": "reviewed", @@ -6616,6 +6612,7 @@ "issue:38175", "issue:39692", "issue:40990", + "issue:42490", "issue:42915", "issue:43257", "issue:43262", @@ -6659,6 +6656,7 @@ ], "soft_pairs": [ "issue:43746|issue:44811", + "issue:42490|issue:43257", "issue:43854|issue:44977", "issue:43643|issue:45759", "issue:39692|issue:44912", @@ -6671,13 +6669,13 @@ "issue:43329|issue:43450", "issue:43329|issue:44479", "issue:42915|issue:43742", - "issue:45459|issue:45803", "issue:34567|issue:40990", + "issue:45459|issue:45803", "issue:46421|issue:46424", "issue:44291|issue:45042", "issue:43749|issue:44756", - "issue:43650|issue:43906", "issue:29159|issue:34567", + "issue:43650|issue:43906", "issue:42915|issue:44821", "issue:43606|issue:45859", "issue:43540|issue:46701", @@ -6686,196 +6684,202 @@ "issue:43927|issue:44912", "issue:44568|issue:45356" ], - "prepared_review_unit_hash": "21df2914e55aef8f08d34cd602e20848c1513a32ff147a82c1578948833421c6", + "prepared_review_unit_hash": "1d6bcb0c560a453e33607d27cecb4c73a9b335f3fd7ee7da87f3d439308ff89c", "component_budget": { "node_count": 505, "item_count": 505, "soft_pair_count": 1854, - "serialized_chars": 962354, - "estimated_input_tokens": 240589, - "estimated_eval_tokens": 481434 + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 }, "budget": { - "node_count": 47, - "item_count": 47, - "soft_pair_count": 27, - "serialized_chars": 35443, - "estimated_input_tokens": 8861, - "estimated_eval_tokens": 17978 + "node_count": 48, + "item_count": 48, + "soft_pair_count": 28, + "serialized_chars": 36313, + "estimated_input_tokens": 9079, + "estimated_eval_tokens": 18414 }, "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 vocabulary (tokenizer, loading, OOM, video, cache, etc.) but not the same underlying bug or fix path. I would reject all soft-edge duplicates and treat this as a set of unrelated issues rather than one duplicate cluster. No PRs are present.", - "confidence": 0.93, - "canonical_issue_reason": "No single canonical issue stands out; the items span many unrelated subsystems and the soft similarities are only topical, not true duplicates.", + "summary": "The items are broadly related to Transformers/model-loading topics, but the soft pairs do not describe the same underlying bug or change. They span distinct failures in tokenizers, multimodal processors, quantization, offload, cache behavior, and training-state tracking.", + "confidence": 0.96, + "canonical_issue_reason": "No clear canonical issue: the set is heterogeneous, and the candidate pairs point to different root causes rather than one duplicate defect.", "canonical_pr_reason": null, - "best_issue_reason": "No strong global best issue for deduping: none of the issues clearly represents the same underlying bug as the others.", + "best_issue_reason": "No single issue is a strong representative because the cluster does not collapse into one bug; any pick would be arbitrary.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43746", "right": "issue:44811", "accept": false, - "reason": "Both involve processors/tokenization, but one is a PEFT/local-checkpoint loading problem and the other is Whisper batch_decode behavior; different code paths." + "reason": "Local PEFT checkpoint loading vs Whisper batch_decode special-token handling; different bugs." + }, + { + "left": "issue:42490", + "right": "issue:43257", + "accept": false, + "reason": "Trailing-slash import warning vs Qwen3 MoE weight-conversion failure; unrelated." }, { "left": "issue:43854", "right": "issue:44977", "accept": false, - "reason": "Different model families and failures: GLM-4.7 test loading vs Qwen3.5 flash-attention generation." + "reason": "GLM-4.7-Flash load/test issue vs Qwen3.5 flash-attention generation regression; different code paths." }, { "left": "issue:43643", "right": "issue:45759", "accept": false, - "reason": "Both touch config/model loading, but one is missing fields from trust_remote_code AutoConfig and the other is incorrect text_config unwrapping in composite multimodal configs." + "reason": "Missing fields with trust_remote_code vs text_config unwrapping for composite multimodal configs; not the same defect." }, { "left": "issue:39692", "right": "issue:44912", "accept": false, - "reason": "SigLIP2 documentation/example issues vs GPT-oss MXFP4 quantization fallback; unrelated bugs." + "reason": "Documentation/example errors vs MXFP4 quantization fallback bug; different symptoms and causes." }, { "left": "issue:43720", "right": "issue:44610", "accept": false, - "reason": "BitNet packed-weight loading vs OmDet-Turbo image size mismatch; not the same problem." + "reason": "Packed-weight accelerate loading bug vs processor image-size mismatch; unrelated." }, { "left": "issue:38175", "right": "issue:44610", "accept": false, - "reason": "Unexpected zero probabilities in SigLIP2 is unrelated to the OmDet processor/model resolution mismatch." + "reason": "Unexpected zero probabilities vs image preprocessing size mismatch; different issues." }, { "left": "issue:43257", "right": "issue:43696", "accept": false, - "reason": "Qwen3 MoE conversion/loading with accelerate+deepspeed is unrelated to a GPT-oss CUDA OOM." + "reason": "MoE weight conversion/loading bug vs CUDA OOM on GPT-oss; not a duplicate." }, { "left": "issue:43262", "right": "issue:44811", "accept": false, - "reason": "Both involve audio processors, but one is apply_chat_template sampling-rate handling and the other is Whisper batch_decode skip_special_tokens behavior." + "reason": "Audio chat-template sample-rate default vs Whisper decode special-token behavior; different components." }, { "left": "issue:44387", "right": "issue:44617", "accept": false, - "reason": "Both mention OOM, but the root causes and code paths differ: int4 quantization memory reservation vs Sam3Video inference memory usage." + "reason": "General int4 memory regression vs Sam3Video OOM; different models and root causes." }, { "left": "issue:43257", "right": "issue:43931", "accept": false, - "reason": "Different Qwen loading failures: missing MoE weight conversion with accelerate+deepspeed vs weight-shape mismatch for Qwen3-VL." + "reason": "Qwen3 MoE conversion issue vs Qwen3-VL weight-shape mismatch; unrelated load failures." }, { "left": "issue:43329", "right": "issue:43450", "accept": false, - "reason": "Both are video-related, but one is undefined variables in multimodal token counting and the other is incorrect batched processor output shape." + "reason": "Undefined multimodal-token helper vars vs batched video shape mismatch; different bugs." }, { "left": "issue:43329", "right": "issue:44479", "accept": false, - "reason": "The latter is a broader video-input regression across models, while the former is a specific undefined-variable bug in token counting; not the same fix." + "reason": "Specific helper-function undefined-name bug vs broader video-input regression; too different to merge." }, { "left": "issue:42915", "right": "issue:43742", "accept": false, - "reason": "Qwen3Moe FineGrainedFP8Config failure and MobileLLM key error are unrelated loading issues." - }, - { - "left": "issue:45459", - "right": "issue:45803", - "accept": false, - "reason": "Both concern exception handling, but they are different modules and different bugs; not mergeable into one fix." + "reason": "FineGrainedFP8Config failure in Qwen3 MoE vs key error loading MobileLLM-125M; unrelated." }, { "left": "issue:34567", "right": "issue:40990", "accept": false, - "reason": "TrainerState token counting bug and GPT-oss perplexity issue do not describe the same underlying defect." + "reason": "TrainerState token counter bug vs high perplexity on GPT-oss; no shared root cause." }, { - "left": "issue:46421", + "left": "issue:45459", + "right": "issue:45803", + "accept": false, + "reason": "Protobuf import error masking tokenizer failures vs bare except swallowing KeyboardInterrupt; both exception-related but distinct." + }, + { + "left": "issue:46421", "right": "issue:46424", "accept": false, - "reason": "Both are cache-related, but one is spurious recompilation from lazy init and the other is max_cache_length handling for quantized cache only." + "reason": "Lazy cache init causing recompilation vs max_cache_length only honored for quantized cache; different cache bugs." }, { "left": "issue:44291", "right": "issue:45042", "accept": false, - "reason": "init_empty_weights/_is_hf_initialized and PIL-image-processor torchvision dependency are unrelated." + "reason": "init_empty_weights argument mismatch vs PIL backend requiring torchvision; unrelated." }, { "left": "issue:43749", "right": "issue:44756", "accept": false, - "reason": "Both involve memory/loading, but one is FSDP CPU RAM-efficient loading and the other is mmap on Strix Halo causing OOM; different scenarios." + "reason": "FSDP CPU RAM efficient loading bug vs mmap-on-Strix-Halo OOM; different mechanisms." }, { - "left": "issue:43650", - "right": "issue:43906", + "left": "issue:29159", + "right": "issue:34567", "accept": false, - "reason": "Placeholder/irrelevant issue title versus an isolated reproduction of a specific upstream issue; no duplicate relationship." + "reason": "Slow/fast tokenizer inconsistency vs TrainerState input-token counting; unrelated." }, { - "left": "issue:29159", - "right": "issue:34567", + "left": "issue:43650", + "right": "issue:43906", "accept": false, - "reason": "Slow-vs-fast tokenizer inconsistency is unrelated to TrainerState.num_input_tokens_seen not updating." + "reason": "Placeholder content issue vs isolated reproduction of another issue; not the same bug." }, { "left": "issue:42915", "right": "issue:44821", "accept": false, - "reason": "Qwen3Moe FP8 loading failure and AutoImageProcessor URL loading are unrelated." + "reason": "Qwen3 MoE FP8 loading failure vs AutoImageProcessor URL loading; different areas." }, { "left": "issue:43606", "right": "issue:45859", "accept": false, - "reason": "Both mention CPU offload, but they affect different models and different code paths; not the same bug." + "reason": "Both mention CPU offload, but one is a device mismatch and the other a rotary-embedding incompatibility; not the same change." }, { "left": "issue:43540", "right": "issue:46701", "accept": false, - "reason": "Both are multimodal/video generation issues, but they involve different models, APIs, and failure modes." + "reason": "Qwen3OmniMoe video-input ValueError vs DiffusionGemma image-text-to-text pipeline failure; different pipelines." }, { "left": "issue:43454", "right": "issue:43577", "accept": false, - "reason": "AyaVision lm_head tying and Blip2 dtype propagation are separate model-initialization bugs." + "reason": "AyaVision weight-tying bug vs BLIP2 dtype propagation bug; unrelated." }, { "left": "issue:28282", "right": "issue:29870", "accept": false, - "reason": "PyTorch missing ImportError and Bart generate decoder_attention_mask unbound variable are unrelated." + "reason": "Missing PyTorch ImportError vs Bart generate UnboundLocalError; different failure layers." }, { "left": "issue:43927", "right": "issue:44912", "accept": false, - "reason": "DiaConfig save/load token IDs issue is unrelated to MXFP4 quantization fallback." + "reason": "DiaConfig save/load token-ID loss vs MXFP4 quantization fallback; different bugs." }, { "left": "issue:44568", "right": "issue:45356", "accept": false, - "reason": "Both are tokenizer regressions, but they concern different models and different failure mechanisms; too broad to treat as duplicates." + "reason": "mDeBERTa special-token regression vs Kimi-K2.5 tokenizer codec/regex regression; unrelated tokenizer issues." } ] }, @@ -6889,7 +6893,7 @@ { "label": "issue", "component_index": 1, - "component_count": 25, + "component_count": 26, "review_unit_index": 22, "review_unit_count": 61, "status": "reviewed", @@ -6953,40 +6957,40 @@ "issue:28282|issue:34689", "issue:33357|issue:34567", "issue:43425|issue:46291", - "issue:43697|issue:45357", "issue:44857|issue:46832", + "issue:43697|issue:45357", "issue:43404|issue:45127", + "issue:45137|issue:45405", "issue:42175|issue:43824", "issue:45538|issue:46869", - "issue:45137|issue:45405", - "issue:28282|issue:29127", "issue:42915|issue:45538", + "issue:28282|issue:29127", "issue:44479|issue:46693", "issue:44062|issue:44991", "issue:43742|issue:44291", "issue:44936|issue:46821", - "issue:42915|issue:43749", "issue:43452|issue:43582", - "issue:43577|issue:46512", + "issue:42915|issue:43749", "issue:24643|issue:29159", + "issue:43577|issue:46512", "issue:44849|issue:45701", "issue:27957|issue:28282", "issue:44987|issue:45907", "issue:41720|issue:43540", "issue:43540|issue:44560", "issue:44479|issue:45381", - "issue:42175|issue:44488", "issue:43540|issue:44479", + "issue:42175|issue:44488", "issue:34567|issue:36331" ], - "prepared_review_unit_hash": "6a1529f248422c8f80976ec0ac992358ee678a36a7d6381729e886e750fff09b", + "prepared_review_unit_hash": "67b34d0821c2a553656c901f2677f76d7f7e84d07d8cb2e3ca0e589049b2ddfa", "component_budget": { "node_count": 505, "item_count": 505, "soft_pair_count": 1854, - "serialized_chars": 962354, - "estimated_input_tokens": 240589, - "estimated_eval_tokens": 481434 + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 }, "budget": { "node_count": 47, @@ -7001,198 +7005,192 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "Mostly unrelated issues. The only soft-edge pair that looks like the same underlying bug is the Qwen2.5-VL video-input regression pair (44479/45381).", - "confidence": 0.79, - "canonical_issue_reason": "issue:44479 is the broader, more general report of the Qwen video-input regression across several models; it better represents the shared bug than the narrower follow-up symptom reports.", + "summary": "The cluster is mostly heterogeneous; most soft pairs are false positives. Only two pairs look like true duplicates: the AddedToken tokenizer-load failure and the Qwen2.5-VL video position-id regression.", + "confidence": 0.62, + "canonical_issue_reason": "No single duplicate theme dominates the set; issue 43749 is the best anchor because it is open, heavily discussed, and has inbound references, making it the strongest triage target.", "canonical_pr_reason": null, - "best_issue_reason": "issue:44479 is the best single issue to keep as the representative ticket for the accepted Qwen video regression cluster.", + "best_issue_reason": "Issue 43749 is the strongest overall issue to anchor the cluster: concrete regression, active discussion, and external references.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43525", "right": "issue:46291", "accept": false, - "reason": "Different subsystems and failure modes: Llama4 config/pad token vs kernels revision/version requirements." + "reason": "Different subsystems: missing config attribute for Llama4 vs kernels revision/version requirement in pipelines import." }, { "left": "issue:44987", "right": "issue:45538", "accept": false, - "reason": "Different bugs: a model-specific tokenizer load failure vs an unrelated CLIPTokenizer max-length issue." + "reason": "Different bugs: model loading regression vs CLIPTokenizer max_length constant." }, { "left": "issue:28282", "right": "issue:34689", "accept": false, - "reason": "Both are loading failures, but they concern different causes and code paths." + "reason": "Both are loading failures, but one is missing PyTorch and the other is a specific Llama 3.2 Vision regression." }, { "left": "issue:33357", "right": "issue:34567", "accept": false, - "reason": "Unrelated: MacOS bus error during CLIP loading vs trainer state token-count tracking." + "reason": "Unrelated: Mac bus error on CLIP model vs TrainerState token counter update issue." }, { "left": "issue:43425", "right": "issue:46291", "accept": false, - "reason": "Torch compatibility and kernels revision/version errors are unrelated." + "reason": "Torch compatibility problem vs kernels revision-forwarding issue; not the same code path." }, { - "left": "issue:43697", - "right": "issue:45357", + "left": "issue:44857", + "right": "issue:46832", "accept": false, - "reason": "Different model families and different problems: RT-DETR output mismatch vs save_pretrained key regression." + "reason": "Both touch detection models, but the crashes are different and not the same underlying bug." }, { - "left": "issue:44857", - "right": "issue:46832", + "left": "issue:43697", + "right": "issue:45357", "accept": false, - "reason": "Both involve detection models, but the concrete bugs are different code paths and not the same fix." + "reason": "Different regressions: RT-DETR output mismatch vs save_pretrained key serialization." }, { "left": "issue:43404", "right": "issue:45127", "accept": false, - "reason": "Both mention weight tying, but they affect different models and different symptoms; not the same underlying bug." + "reason": "Both involve weight tying, but one is a model implementation bug and the other is LoRA merge behavior with extended vocab." + }, + { + "left": "issue:45137", + "right": "issue:45405", + "accept": false, + "reason": "DeepSpeed ZeRO3 deque error is unrelated to the PEFT version bump." }, { "left": "issue:42175", "right": "issue:43824", "accept": false, - "reason": "Install/backend packaging issue vs missing model class import; unrelated." + "reason": "Packaging/backend dependency issue vs a missing model class import." }, { "left": "issue:45538", "right": "issue:46869", "accept": false, - "reason": "CLIP tokenizer max-length bug is unrelated to YOLO multi-GPU device-map failure." + "reason": "Tokenizer max-length constant is unrelated to YOLO multi-GPU device-map failure." }, { - "left": "issue:45137", - "right": "issue:45405", + "left": "issue:42915", + "right": "issue:45538", "accept": false, - "reason": "DeepSpeed ZeRO3 deque error is unrelated to the PEFT minimum-version bump." + "reason": "Qwen3Moe FP8 config failure vs CLIPTokenizer max_length bug; different components." }, { "left": "issue:28282", "right": "issue:29127", "accept": false, - "reason": "A missing PyTorch dependency error is unrelated to a LayoutLMv3 validation-message clarity issue." - }, - { - "left": "issue:42915", - "right": "issue:45538", - "accept": false, - "reason": "Qwen3Moe FP8 failure and CLIP tokenizer max-length are unrelated." + "reason": "Model import error vs LayoutLMv3 validation/error-message clarity issue." }, { "left": "issue:44479", "right": "issue:46693", "accept": false, - "reason": "Both are regressions, but one is video-input parsing and the other is long-context FA2 performance; different bugs." + "reason": "Different regressions: multimodal video input bug vs FA2 long-context performance regression." }, { "left": "issue:44062", "right": "issue:44991", - "accept": false, - "reason": "Likely tokenizer-loading related, but not enough evidence they are the same concrete bug or fix." + "accept": true, + "reason": "Likely the same tokenizer deserialization bug: AddedToken special-argument handling breaks loading for a specific tokenizer/model." }, { "left": "issue:43742", "right": "issue:44291", "accept": false, - "reason": "Different model-loading failures with different triggers and code paths." + "reason": "Both are load-time errors, but one is a key error for a model and the other is an init_empty_weights argument mismatch." }, { "left": "issue:44936", "right": "issue:46821", "accept": false, - "reason": "Trainer lifecycle bug vs TimesFM window_size attribute error; unrelated." - }, - { - "left": "issue:42915", - "right": "issue:43749", - "accept": false, - "reason": "Different training/loading issues and different models; only broad ML-framework context overlaps." + "reason": "Trainer train/evaluate state issue vs TimesFM window_size AttributeError; unrelated." }, { "left": "issue:43452", "right": "issue:43582", "accept": false, - "reason": "GGUF loading failure and Apple Silicon allocator TypeError are unrelated." + "reason": "gguf_file loading breakage is unrelated to Apple Silicon caching_allocator_warmup TypeError." }, { - "left": "issue:43577", - "right": "issue:46512", + "left": "issue:42915", + "right": "issue:43749", "accept": false, - "reason": "Both are dtype-related, but one is a model-component loading issue and the other is a stale config dtype issue; not the same bug." + "reason": "Different domains: Qwen3Moe FP8 config vs FSDP CPU RAM efficient loading." }, { "left": "issue:24643", "right": "issue:29159", "accept": false, - "reason": "DeepSpeed training runtime error vs tokenizer inconsistency; unrelated." + "reason": "DeepSpeed weight-dimension error vs tokenizer slow/fast inconsistency; unrelated." + }, + { + "left": "issue:43577", + "right": "issue:46512", + "accept": false, + "reason": "Both concern dtype propagation, but one is model submodule dtype on load and the other is stale config.dtype from from_config." }, { "left": "issue:44849", "right": "issue:45701", "accept": false, - "reason": "Different issues: Qwen3.5 hidden-states behavior vs version-dependent tokenization changes." + "reason": "Qwen3.5 hidden-states bug vs tokenizer behavior change across versions; not the same bug." }, { "left": "issue:27957", "right": "issue:28282", "accept": false, - "reason": "Flash Attention 2 behavior for XLM-Roberta is unrelated to missing PyTorch detection." + "reason": "Flash Attention 2 support issue vs missing PyTorch environment dependency." }, { "left": "issue:44987", "right": "issue:45907", "accept": false, - "reason": "Specific tokenizer load failure vs list_repo_files not forwarding revision; unrelated." + "reason": "Model loading regression vs list_repo_files revision forwarding; different APIs." }, { "left": "issue:41720", "right": "issue:43540", "accept": false, - "reason": "Both are Qwen issues, but one is auto device-map CUDA assertion and the other is video-input processing; different bugs." + "reason": "Auto device-map CUDA assert vs video-input processing error in Qwen3OmniMoe; different failure modes." }, { "left": "issue:43540", "right": "issue:44560", "accept": false, - "reason": "Both concern video inputs, but the model families and reported failures differ enough that they do not look like one concrete bug." + "reason": "Both are video-input bugs, but they affect different models and likely different code paths." }, { "left": "issue:44479", "right": "issue:45381", "accept": true, - "reason": "Both describe the same Qwen2.5-VL video-input regression in transformers 5.3.0; 45381 is a narrower symptom of the broader 44479 report." + "reason": "Same underlying Qwen2.5-VL video regression: incorrect vision_position_ids in the video path." }, { "left": "issue:42175", "right": "issue:44488", "accept": false, - "reason": "Packaging/backend install issue vs a specific model-loading failure; unrelated." - }, - { - "left": "issue:43540", - "right": "issue:44479", - "accept": false, - "reason": "Related video-input theme, but different model families and not clearly the same concrete failure path." + "reason": "TensorFlow backend packaging issue vs loading a specific model; unrelated." }, { "left": "issue:34567", "right": "issue:36331", "accept": false, - "reason": "TrainerState token counting and custom compute_loss signature changes are different trainer API issues." + "reason": "TrainerState token accounting bug vs CustomTrainer compute_loss signature mismatch." } ] }, "evaluator_result": { "accept": true, - "feedback": "The summary is grounded in the packet and stays conservative. The representative issue choice for 44479 over 45381 is justified by the titles, and the soft-edge verdicts consistently avoid overclaiming identity where only broad thematic overlap exists. The only accepted pair (44479/45381) is the one that clearly looks like the same Qwen2.5-VL video-input regression." + "feedback": "The summary is grounded in the packet: issue 43749 is a reasonable anchor given its open state, discussion, and inbound references, and the soft-edge judgments stay mostly conservative. The two accepted pairs are plausible duplicates from the issue titles, and the rejected pairs are not overstated." }, "evaluator_used": true, "retried": false, @@ -7203,7 +7201,7 @@ { "label": "issue", "component_index": 1, - "component_count": 25, + "component_count": 26, "review_unit_index": 23, "review_unit_count": 61, "status": "reviewed", @@ -7264,28 +7262,28 @@ ], "soft_pairs": [ "issue:38175|issue:44898", - "issue:44568|issue:45488", "issue:44792|issue:44964", + "issue:44568|issue:45488", "issue:44062|issue:45081", - "issue:43257|issue:43756", "issue:44291|issue:44821", + "issue:43257|issue:43756", "issue:43582|issue:46821", "issue:45042|issue:45081", - "issue:43756|issue:46399", "issue:43742|issue:44589", + "issue:43756|issue:46399", "issue:43756|issue:45325", "issue:43716|issue:46701", - "issue:44568|issue:46491", "issue:43577|issue:45072", - "issue:45405|issue:45600", + "issue:44568|issue:46491", "issue:44737|issue:46463", + "issue:45405|issue:45600", "issue:43531|issue:44821", "issue:45672|issue:45674", "issue:45823|issue:46463", "issue:43906|issue:46693", "issue:24643|issue:29870", - "issue:43906|issue:46468", "issue:43452|issue:44488", + "issue:43906|issue:46468", "issue:44625|issue:44877", "issue:43540|issue:46693", "issue:45663|issue:46846", @@ -7293,14 +7291,14 @@ "issue:44737|issue:45588", "issue:46463|issue:46899" ], - "prepared_review_unit_hash": "367fa94095b036268d808abb6a818db77d2cca9d97e50850d638eb2dc40f4921", + "prepared_review_unit_hash": "0f1e931427547e4bf150ec8b338ff816e756a8f37e6ff6be15b15efc70dc3335", "component_budget": { "node_count": 505, "item_count": 505, "soft_pair_count": 1854, - "serialized_chars": 962354, - "estimated_input_tokens": 240589, - "estimated_eval_tokens": 481434 + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 }, "budget": { "node_count": 48, @@ -7315,186 +7313,186 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "These issues are mostly unrelated, model-specific bugs and regressions rather than one duplicate cluster. I found no soft-edge pair that looks like the same underlying bug/change, so all candidate links should be rejected.", - "confidence": 0.91, - "canonical_issue_reason": null, + "summary": "This cluster is a heterogeneous mix of unrelated bug reports across tokenizers, multimodal models, quantization, distributed training, and config loading. None of the soft pairs look like the same underlying bug or change, so all soft edges are rejected.", + "confidence": 0.94, + "canonical_issue_reason": "If a single issue must anchor the cluster, issue 45488 is the strongest candidate because it is the most detailed and broadly actionable report, but there is no real duplicate center in this set.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45488 has the strongest discussion/inbound signal and a broad, user-visible regression, but the set overall does not share one underlying root cause.", + "best_issue_reason": "Issue 45488 is the best representative only as a high-signal, well-described issue; it should not be treated as a duplicate hub for the other items.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:38175", "right": "issue:44898", "accept": false, - "reason": "Different models and failure modes: SigLIP2 zero probabilities vs Perceiver non-default-resolution failure." + "reason": "Both are vision-related regressions, but they involve different models, different symptoms, and different code paths." }, { - "left": "issue:44568", - "right": "issue:45488", + "left": "issue:44792", + "right": "issue:44964", "accept": false, - "reason": "Tokenizer regressions, but in different code paths and families: mDeBERTa special-token handling vs LlamaTokenizer overriding tokenizer.json pre-tokenizer behavior." + "reason": "Different multimodal models with unrelated failures: a Janus test regression versus Phi-4 model loading." }, { - "left": "issue:44792", - "right": "issue:44964", + "left": "issue:44568", + "right": "issue:45488", "accept": false, - "reason": "Different multimodal issues: a failed Janus generation test vs loading Phi-4-multimodal-instruct." + "reason": "Both are tokenizer regressions, but they affect different tokenizers and different breakages (special tokens handling vs pre-tokenizer override)." }, { "left": "issue:44062", "right": "issue:45081", "accept": false, - "reason": "Unrelated tokenizer bugs: AddedToken special-argument collision vs Mistral regex patch crashing on backend_tokenizer access." + "reason": "Separate tokenizer bugs: AddedToken argument duplication versus a Mistral regex patch crash." }, { - "left": "issue:43257", - "right": "issue:43756", + "left": "issue:44291", + "right": "issue:44821", "accept": false, - "reason": "Both involve Qwen/SmolLM model internals, but the bugs differ: MoE weight conversion under accelerate+deepspeed vs dropped RoPE layers in Smollm3." + "reason": "One is about init_empty_weights unexpected args; the other is AutoImageProcessor URL loading. No shared bug." }, { - "left": "issue:44291", - "right": "issue:44821", + "left": "issue:43257", + "right": "issue:43756", "accept": false, - "reason": "Different areas entirely: init_empty_weights handling vs loading AutoImageProcessor from URL." + "reason": "Different Qwen/Smollm model issues with unrelated root causes (MoE weight conversion vs RoPE layer dropping)." }, { "left": "issue:43582", "right": "issue:46821", "accept": false, - "reason": "Different runtime errors in different subsystems: Apple Silicon allocator warmup TypeError vs TimesFM window_size AttributeError." + "reason": "Completely different runtime failures in unrelated components: Apple Silicon allocator warmup versus TimesFM window_size handling." }, { "left": "issue:45042", "right": "issue:45081", "accept": false, - "reason": "PIL backend torchvision dependency regression is unrelated to the Mistral tokenizer regex crash." + "reason": "Image processor torchvision dependency regression is unrelated to the Mistral tokenizer regex crash." }, { - "left": "issue:43756", - "right": "issue:46399", + "left": "issue:43742", + "right": "issue:44589", "accept": false, - "reason": "Both mention rotary-related internals, but they are distinct model-specific defects: Smollm3 RoPE layer count vs Qwen3.5 kernel registration." + "reason": "Different load-time errors with different mechanisms: config/key loading versus Float8 storage lookup." }, { - "left": "issue:43742", - "right": "issue:44589", + "left": "issue:43756", + "right": "issue:46399", "accept": false, - "reason": "Different loading failures with different causes: MobileLLM key error vs missing Float8 storage type." + "reason": "Both touch Qwen/attention internals, but one is a RoPE configuration bug and the other is a kernel registration crash." }, { "left": "issue:43756", "right": "issue:45325", "accept": false, - "reason": "Both touch RoPE logic, but one is Smollm3 layer dropping while the other is Qwen2.5-VL still-image temporal scaling." + "reason": "Different Qwen model families and different rotary-position bugs; not the same underlying issue." }, { "left": "issue:43716", "right": "issue:46701", "accept": false, - "reason": "Different multimodal regressions: dtype mismatch in Mistral-3 preprocessing vs DiffusionGemma image-text-to-text pipeline failure." - }, - { - "left": "issue:44568", - "right": "issue:46491", - "accept": false, - "reason": "Both are tokenizer regressions, but one is BOS/EOS special-token addition and the other is CodeLlama leading-space round-trip behavior." + "reason": "Unrelated multimodal bugs: image preprocessor dtype mismatch versus a pipeline/output wiring issue." }, { "left": "issue:43577", "right": "issue:45072", "accept": false, - "reason": "Both are dtype-mismatch reports, but they involve different models and different parts of the inference stack." + "reason": "Both mention dtype mismatches, but they occur in different models and different execution paths." }, { - "left": "issue:45405", - "right": "issue:45600", + "left": "issue:44568", + "right": "issue:46491", "accept": false, - "reason": "Completely different issues: dependency version bump vs stale auto_mappings references to removed configs." + "reason": "Two distinct tokenizer regressions affecting different model families and different round-trip behaviors." }, { "left": "issue:44737", "right": "issue:46463", "accept": false, - "reason": "XLNet CPU placement bug is unrelated to T5 decoder-input validation error messaging." + "reason": "XLNet CPU placement bug is unrelated to T5 decoder input validation messaging." + }, + { + "left": "issue:45405", + "right": "issue:45600", + "accept": false, + "reason": "A dependency version bump problem is not the same as broken auto_mappings references to removed configs." }, { "left": "issue:43531", "right": "issue:44821", "accept": false, - "reason": "Qwen3-MoE sliding_window behavior is unrelated to AutoImageProcessor URL loading." + "reason": "Sliding-window behavior in Qwen3-MoE is unrelated to loading AutoImageProcessor from a URL." }, { "left": "issue:45672", "right": "issue:45674", "accept": false, - "reason": "Both affect 4-bit inference, but one is a uint8/torch.finfo bug in Gemma4 audio modules and the other is a BitsAndBytes skip-modules exclusion bug." + "reason": "Both involve 4-bit inference, but the root causes differ: unsafe dtype cast in Gemma4 audio/vision modules versus BitsAndBytes skip-module exclusion logic." }, { "left": "issue:45823", "right": "issue:46463", "accept": false, - "reason": "Gemma4 device_map auto placement issue is unrelated to the T5 decoder input validation error." + "reason": "Gemma4 device placement during forward is unrelated to the T5 decoder-input error message bug." }, { "left": "issue:43906", "right": "issue:46693", "accept": false, - "reason": "The reproduction of issue 38071 is not the same as the FA2 long-context performance regression." + "reason": "The titles indicate different problems: an isolated repro of an older issue versus a new FA2 long-context performance regression." }, { "left": "issue:24643", "right": "issue:29870", "accept": false, - "reason": "Different errors and subsystems: DeepSpeed training weight shape RuntimeError vs Bart generate() unbound local decoder_attention_mask." + "reason": "DeepSpeed training weight-shape failure and BART generate decoder_attention_mask bug are distinct issues." }, { - "left": "issue:43906", - "right": "issue:46468", + "left": "issue:43452", + "right": "issue:44488", "accept": false, - "reason": "An isolated reproduction of another issue is not the same as the broken Nemotron example report." + "reason": "A gguf_file-related from_pretrained breakage is not the same as a specific model failing to load." }, { - "left": "issue:43452", - "right": "issue:44488", + "left": "issue:43906", + "right": "issue:46468", "accept": false, - "reason": "Different model-loading problems: gguf_file compatibility vs loading cjvt/sleng-bert." + "reason": "An isolated repro of a prior issue and a broken Nemotron example are different failures with no clear shared code path." }, { "left": "issue:44625", "right": "issue:44877", "accept": false, - "reason": "Qwen3.5 num_labels propagation bug is unrelated to strict config blocking granite_speech loading." + "reason": "Qwen3.5 label propagation and granite_speech strict-config loading are unrelated config bugs." }, { "left": "issue:43540", "right": "issue:46693", "accept": false, - "reason": "Different Qwen3OmniMoe video-input validation bug vs a FA2 long-context generation performance regression." + "reason": "Video-input processing in Qwen3OmniMoe is unrelated to FA2 long-context performance regression." }, { "left": "issue:45663", "right": "issue:46846", "accept": false, - "reason": "Gemma4 FSDP2 shared_kv_states rebuild bug is unrelated to Qwen3.5 tensor-parallel plan omissions." + "reason": "Different distributed-training failures in different models: Gemma4 FSDP2 KeyError versus Qwen3.5 TP plan omissions." }, { "left": "issue:43697", "right": "issue:46829", "accept": false, - "reason": "Different RTDetrV2 output-difference regression vs VideoPrism hidden_states field mix-up." + "reason": "Different model outputs/outputs structure bugs in different vision models, not the same defect." }, { "left": "issue:44737", "right": "issue:45588", "accept": false, - "reason": "XLNet CPU arange bug is unrelated to the flash_attention s_aux=None AttributeError." + "reason": "XLNet CPU allocation bug and flash_attention s_aux=None crash are unrelated." }, { "left": "issue:46463", "right": "issue:46899", "accept": false, - "reason": "T5 decoder input validation messaging is unrelated to Gemma4 4-bit multimodal dtype cast failures." + "reason": "T5 decoder validation messaging has nothing to do with Gemma4 multimodal 4-bit quantization dtype casting." } ] }, @@ -7508,7 +7506,7 @@ { "label": "issue", "component_index": 1, - "component_count": 25, + "component_count": 26, "review_unit_index": 24, "review_unit_count": 61, "status": "reviewed", @@ -7572,9 +7570,9 @@ "issue:44913|issue:46121", "issue:43526|issue:45479", "issue:43697|issue:45216", - "issue:43643|issue:45375", - "issue:44466|issue:45005", "issue:43720|issue:44898", + "issue:44466|issue:45005", + "issue:43643|issue:45375", "issue:45200|issue:46463", "issue:44493|issue:46050", "issue:41762|issue:44936", @@ -7593,198 +7591,201 @@ "issue:43531|issue:43742", "issue:43295|issue:45356" ], - "prepared_review_unit_hash": "c39b13555f605ca427165bb411c3fb8f6ea1ed4badbc2a7cbf542cdd2fe72873", + "prepared_review_unit_hash": "d27eac30ff96a244a85f0cbd37db6fcf44b24a3e191c7dbaf3cd8f88fe57a71e", "component_budget": { "node_count": 505, "item_count": 505, "soft_pair_count": 1854, - "serialized_chars": 962354, - "estimated_input_tokens": 240589, - "estimated_eval_tokens": 481434 + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 }, "budget": { "node_count": 47, "item_count": 47, "soft_pair_count": 25, - "serialized_chars": 35140, - "estimated_input_tokens": 8785, - "estimated_eval_tokens": 17826 + "serialized_chars": 35141, + "estimated_input_tokens": 8786, + "estimated_eval_tokens": 17828 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is highly heterogeneous: most items are unrelated bug reports spanning tokenizers, configs, generation outputs, training, and model-specific regressions. The soft pairs mostly share broad subsystem vocabulary but do not look like the same concrete bug or fix path.", - "confidence": 0.68, - "canonical_issue_reason": "issue:45920 is the broadest umbrella-style report in the set: a v5 AutoTokenizer regression affecting multiple model families with a clear, repeated symptom and high likelihood of attracting related duplicates.", + "summary": "Mostly heterogeneous issue cluster with one likely duplicate pair around tied-weights serialization in v5. Most other pairs are different model/component failures and should stay separate.", + "confidence": 0.84, + "canonical_issue_reason": "issue:45005 is the broadest and clearest representative of the tied-weights v5 regression, and it plausibly subsumes the narrower lm_head.weight serialization report in issue:44466.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45920 is the strongest representative candidate because it is broad, well-scoped, and impacts several models rather than a single narrow edge case.", + "best_issue_reason": "issue:45005 best captures the underlying bug class and is the most suitable anchor for deduping the tied-weights serialization reports.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:45357", "right": "issue:45920", "accept": false, - "reason": "Both concern v5 serialization/tokenization regressions, but one is a Qwen3.5 save_pretrained visual-key issue and the other is wrong token IDs across many models; different code paths and symptoms." + "reason": "Different regressions: Qwen3.5 visual encoder key serialization vs broad tokenizer token-id corruption across many models." }, { "left": "issue:28282", "right": "issue:32090", "accept": false, - "reason": "Completely different failures: missing PyTorch import for AutoModel vs a Trainer GPU broadcast TypeError." + "reason": "Unrelated failures: missing PyTorch import for AutoModel vs Trainer GPU broadcast TypeError." }, { "left": "issue:44913", "right": "issue:46121", "accept": false, - "reason": "Both touch rope/config handling, but one is a config field reverting on reload and the other is a TypeError in JSON-loaded list validation; not the same bug." + "reason": "Both rope-related, but one is config reload defaulting and the other is a JSON-list type error in validation; not the same bug." }, { "left": "issue:43526", "right": "issue:45479", "accept": false, - "reason": "Different layers and symptoms: BeitImageProcessorFast label reduction vs sequence-classification loss degeneracy across models." + "reason": "Different model families and failure modes: BeitImageProcessorFast label reduction vs degenerate classification loss." }, { "left": "issue:43697", "right": "issue:45216", "accept": false, - "reason": "Different models and failure modes: RTDetrV2 output drift vs Qwen3.5 checkpoint save_pretrained key corruption." + "reason": "Different concrete bugs: RTDetrV2 output mismatch vs Qwen3.5 save_pretrained checkpoint corruption." }, { - "left": "issue:43643", - "right": "issue:45375", + "left": "issue:43720", + "right": "issue:44898", "accept": false, - "reason": "Both involve missing config fields, but one is generic trust_remote_code AutoConfig loading and the other is a specific Qwen3_5MoeVisionConfig field dropped by strict serialization." + "reason": "Different subsystems and symptoms: BitNet packed-weight loading vs Perceiver positional interpolation failure." }, { "left": "issue:44466", "right": "issue:45005", - "accept": false, - "reason": "Related theme around tied weights, but the first is device-dependent lm_head serialization while the second is a broader translation-model tied-weights regression; not clearly the same concrete bug." + "accept": true, + "reason": "Both describe the same tied-weights serialization/save_pretrained regression in v5; 44466 looks like a narrower instance of the broader 45005 issue." }, { - "left": "issue:43720", - "right": "issue:44898", + "left": "issue:43643", + "right": "issue:45375", "accept": false, - "reason": "Unrelated problems: BitNet packed-weight loading vs Perceiver non-default resolution interpolation." + "reason": "Different config-field loss bugs: missing fields from trust_remote_code AutoConfig vs missing deepstack_visual_indexes in Qwen3_5MoeVisionConfig." }, { "left": "issue:45200", "right": "issue:46463", "accept": false, - "reason": "Different issues entirely: Gemma 4 text-only mm_token_type_ids defaulting vs a T5 decoder-input error-message bug." + "reason": "Different bugs: Gemma 4 mm_token_type_ids defaulting issue vs T5 decoder input validation error messaging." }, { "left": "issue:44493", "right": "issue:46050", "accept": false, - "reason": "Position-id unexpected keys and sinq quantized model reload/save corruption are separate serialization problems." + "reason": "Different serialization problems: unexpected position-id keys vs sinq quantization save/load breakage." }, { "left": "issue:41762", "right": "issue:44936", "accept": false, - "reason": "Different workflows and root causes: DeepSpeed ZeRO-3 loading failure vs trainer.evaluate() after train()." + "reason": "Different failures: Gemma3 ZeRO-3 loading index error vs trainer.evaluate after trainer.train." }, { "left": "issue:43452", "right": "issue:44843", "accept": false, - "reason": "Both are AutoTokenizer.from_pretrained failures, but one is gguf_file support and the other is an offline-mode regression in _patch_mistral_regex." + "reason": "Different causes: gguf_file handling in AutoTokenizer/AutoModelForCausalLM vs offline-model_info call in _patch_mistral_regex." }, { "left": "issue:43761", "right": "issue:46829", "accept": false, - "reason": "Both involve hidden_states outputs, but one is CLIPVisionModel returning None and the other is VideoPrism misplacing last_hidden_state in an output object." + "reason": "Different output bugs in different models: CLIPVisionModel hidden_states regression vs VideoPrism hidden_state field mix-up." }, { "left": "issue:43688", "right": "issue:43749", "accept": false, - "reason": "Different subsystems: MoE auxiliary-loss normalization vs FSDP CPU RAM efficient loading." + "reason": "Different training/loading issues: auxiliary-loss normalization vs FSDP CPU RAM efficient loading breakage." }, { "left": "issue:44361", "right": "issue:44442", "accept": false, - "reason": "Both are tokenizer-loading issues, but they affect different tokenizer implementations and different failure modes." + "reason": "Both tokenizer-related, but different tokenizers and failure types; not the same underlying bug." }, { "left": "issue:38175", "right": "issue:43295", "accept": false, - "reason": "One is unexpected zero probabilities in a SigLIP2 model; the other is a regression in processor.tokenizer access and image passing to tokenizer." + "reason": "Different regressions: siglip2 zero probabilities vs processor.tokenizer/image-to-tokenizer regression." }, { "left": "issue:43653", "right": "issue:44361", "accept": false, - "reason": "Different tokenizer bugs: BigBird special-token registration vs MLuke task-time AttributeError." + "reason": "Different tokenizer bugs: BigBird special-token registration vs MLukeTokenizer attribute error on tasks." }, { "left": "issue:43819", "right": "issue:44811", "accept": false, - "reason": "Different components and symptoms: DAC latent decoding mismatch vs Whisper batch_decode skip_special_tokens handling." + "reason": "Different code paths: DAC latent conversion mismatch vs Whisper batch_decode skip_special_tokens handling." }, { "left": "issue:43994", "right": "issue:44898", "accept": false, - "reason": "Both are vision-model regressions, but one is SigLIP2 inference quality and the other is Perceiver resizing/interpolation failure." + "reason": "Different model inference issues: siglip2 AutoModel/pipeline nonsense outputs vs Perceiver interpolation failure." }, { "left": "issue:45372", "right": "issue:46170", "accept": false, - "reason": "Different import/dependency problems: mistral_common ReasoningEffort import vs torch.distributed dependency issues on Windows/ROCm." + "reason": "Unrelated import/dependency problems: mistral_common ReasoningEffort import vs Windows ROCm torch.distributed dependency." }, { "left": "issue:44857", "right": "issue:47000", "accept": false, - "reason": "Both are RT-DETR loss crashes, but the reported failure modes differ: AMP/CUDA crash vs infeasible cost matrix." + "reason": "Different RT-DETR losses with different failure modes: AMP/CUDA crash vs infeasible cost matrix." }, { "left": "issue:46832", "right": "issue:46932", "accept": false, - "reason": "Different model families and fixes: RT-DETR backbone-feature-level mismatch vs Gemma4 quantized embedder dtype casting." + "reason": "Different models and bugs: RT-DETR backbone feature-level mismatch vs Gemma4 quantized embedder dtype casting." }, { "left": "issue:43525", "right": "issue:46701", "accept": false, - "reason": "Unrelated model/component failures: Llama4Config pad_token_id access vs DiffusionGemma image-text-to-text pipeline." + "reason": "Unrelated issues: Llama4Config missing pad_token_id vs DiffusionGemma pipeline failure." }, { "left": "issue:43531", "right": "issue:43742", "accept": false, - "reason": "Different bugs: Qwen3-MoE sliding_window behavior vs MobileLLM-125M loading key error." + "reason": "Different model-specific errors: Qwen3-MoE sliding window behavior vs MobileLLM key error on load." }, { "left": "issue:43295", "right": "issue:45356", "accept": false, - "reason": "Both are regressions, but one is processor/tokenizer API breakage and the other is Kimi-K2.5 codec and warning behavior; not the same underlying issue." + "reason": "Different regressions in different layers: processor/tokenizer image handling vs Kimi-K2.5 tokenizer codec/regex warning issues." } ] }, - "evaluator_result": null, - "evaluator_used": false, + "evaluator_result": { + "accept": true, + "feedback": "The assessment is grounded in the packet and stays conservative overall. The only accepted pair, issue:44466 \u2194 issue:45005, is plausibly the same tied-weights/save_pretrained regression at different specificity levels. The remaining soft-edge rejections are supported by clear title-level differences 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 }, { "label": "issue", "component_index": 1, - "component_count": 25, + "component_count": 26, "review_unit_index": 25, "review_unit_count": 61, "status": "reviewed", @@ -7812,13 +7813,13 @@ "issue:43749", "issue:43756", "issue:43761", + "issue:43782", "issue:43927", "issue:44361", "issue:44464", "issue:44466", "issue:44479", "issue:44484", - "issue:44554", "issue:44568", "issue:44625", "issue:44704", @@ -7844,15 +7845,15 @@ "issue:47013" ], "soft_pairs": [ - "issue:45542|issue:46178", "issue:29159|issue:33357", + "issue:45542|issue:46178", "issue:43653|issue:44568", "issue:43646|issue:43749", "issue:43688|issue:45704", "issue:43452|issue:43577", + "issue:43688|issue:45701", "issue:44843|issue:44991", "issue:43761|issue:47013", - "issue:43688|issue:45701", "issue:42898|issue:46491", "issue:43756|issue:46402", "issue:44464|issue:44743", @@ -7864,218 +7865,218 @@ "issue:33666|issue:36010", "issue:43550|issue:44361", "issue:44466|issue:46050", - "issue:43688|issue:45440", "issue:32090|issue:33666", + "issue:43688|issue:45440", "issue:44977|issue:46693", - "issue:44734|issue:46612", "issue:44464|issue:44977", + "issue:44734|issue:46612", "issue:41720|issue:44464", "issue:43504|issue:43550", "issue:43504|issue:43720", - "issue:43577|issue:44554" + "issue:43782|issue:46693" ], - "prepared_review_unit_hash": "be9ab29b26230f6d902528125ea0546a1d3b3946323a85091c1735135ab07af1", + "prepared_review_unit_hash": "7ca687a5ddab8b50da23aed8d9dd494bbb1848f1df6be0c5adcd22811dd8abfd", "component_budget": { "node_count": 505, "item_count": 505, "soft_pair_count": 1854, - "serialized_chars": 962354, - "estimated_input_tokens": 240589, - "estimated_eval_tokens": 481434 + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 29, - "serialized_chars": 36541, - "estimated_input_tokens": 9136, - "estimated_eval_tokens": 18528 + "serialized_chars": 36521, + "estimated_input_tokens": 9131, + "estimated_eval_tokens": 18518 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is a heterogeneous mix of unrelated Transformers bug reports: tokenizer regressions, multimodal model-loading issues, distributed-training failures, generation/cache bugs, and a few config/serialization problems. None of the soft pairs look like the same underlying defect or change.", - "confidence": 0.91, - "canonical_issue_reason": "issue:45701 is the broadest tokenizer-regression report in the set and the best representative for the tokenizer-related subset, though the overall cluster is not a true duplicate cluster.", + "summary": "This cluster is only loosely related by broad Transformer/tokenization/generation regression themes. The soft pairs mostly point to different code paths or different models/backends, so none should be merged as duplicates. The best representative issue is the slow-vs-fast tokenizer inconsistency report.", + "confidence": 0.64, + "canonical_issue_reason": "issue:29159 is the clearest, most central tokenizer-regression report here: it describes a concrete inconsistency between slow and fast tokenizer behavior, which is the broadest shared theme among the tokenizer-related items.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45701 is the most representative issue overall: broad user-facing tokenizer behavior change, multiple inbound references, and it anchors the only semi-coherent theme in the cluster.", + "best_issue_reason": "issue:29159 is the best single issue to anchor this cluster because it is concrete, widely applicable across tokenizer implementations, and closer to a canonical duplicate target than the more model-specific or version-specific reports.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:45542", - "right": "issue:46178", + "left": "issue:29159", + "right": "issue:33357", "accept": false, - "reason": "TensorBoard/TensorFlow backend setup vs a backend-mapping error for tensorflow_text; different failure paths." + "reason": "Tokenizer behavior inconsistency vs MacOS bus error loading a community CLIP model; unrelated failures." }, { - "left": "issue:29159", - "right": "issue:33357", + "left": "issue:45542", + "right": "issue:46178", "accept": false, - "reason": "Tokenizer slow/fast inconsistency vs a MacOS bus error in CLIP model loading; unrelated bugs." + "reason": "Both mention backend/import problems, but one is TensorBoard-without-TensorFlow and the other is tensorflow_text backend mapping; different bugs." }, { "left": "issue:43653", "right": "issue:44568", "accept": false, - "reason": "Both are tokenizer bugs, but one is BigBird mask-token decode behavior and the other is mdeberta special-token insertion; different tokenizers and symptoms." + "reason": "Both are tokenizer special-token regressions, but BigBird mask-token registration and mDeBERTa BOS/EOS insertion are different concrete issues." }, { "left": "issue:43646", "right": "issue:43749", "accept": false, - "reason": "Custom model initialization breakage vs FSDP CPU RAM efficient loading; not the same code path." + "reason": "Custom model initialization regression vs FSDP CPU RAM efficient loading breakage; same broad area, not the same bug." }, { "left": "issue:43688", "right": "issue:45704", "accept": false, - "reason": "Auxiliary loss normalization vs a T5 RMSNorm memory leak; different subsystems." + "reason": "Auxiliary-loss normalization bug vs T5/apex RMSNorm memory leak; unrelated." }, { "left": "issue:43452", "right": "issue:43577", "accept": false, - "reason": "gguf_file loading failure vs BLIP2 dtype propagation; different model-loading problems." + "reason": "gguf_file loading failure vs Blip2 dtype propagation issue; different model-loading paths and symptoms." + }, + { + "left": "issue:43688", + "right": "issue:45701", + "accept": false, + "reason": "Auxiliary-loss normalization vs generic tokenization changes across versions; not the same underlying problem." }, { "left": "issue:44843", "right": "issue:44991", "accept": false, - "reason": "Offline hub call in tokenizer patch vs tokenizer load failure for est-roberta; different root causes." + "reason": "Offline model_info call in AutoTokenizer patch vs tokenizer loading failure for a specific repo; different root causes." }, { "left": "issue:43761", "right": "issue:47013", "accept": false, - "reason": "CLIPVision hidden_states regression vs xLSTM train-mode unpacking crash; unrelated models and mechanics." - }, - { - "left": "issue:43688", - "right": "issue:45701", - "accept": false, - "reason": "Auxiliary loss normalization vs a broad tokenization-regression report; unrelated defects." + "reason": "CLIP hidden_states regression vs xLSTM return-last-states unpacking crash; both are output-shape issues but on different models/code paths." }, { "left": "issue:42898", "right": "issue:46491", "accept": false, - "reason": "v5 cleanup-space behavior change vs CodeLlama leading-space round-trip regression; both tokenizer-related but not the same bug." + "reason": "cleanup-tokenization-space behavior change vs CodeLlama leading-space strip regression; both tokenization-related but not the same bug." }, { "left": "issue:43756", "right": "issue:46402", "accept": false, - "reason": "Smollm3 RoPE-layer mismatch vs Sam3 text-encoder weights not loading; unrelated model issues." + "reason": "Smollm3 RoPE-layer mismatch vs Sam3 text encoder weights not loading; unrelated model-specific loading/config issues." }, { "left": "issue:44464", "right": "issue:44743", "accept": false, - "reason": "Chunked generation with compiled forward vs Qwen3.5 recurrent-state reset when seq_len>1; different generation paths." + "reason": "Chunked generation with compiled forward vs recurrent state reset when seq_len>1; both generation-related, but different failure modes and code paths." }, { "left": "issue:43653", "right": "issue:43927", "accept": false, - "reason": "BigBird special-token registration vs DiaConfig save/load token-ID loss; tokenizer behavior vs config persistence." + "reason": "BigBird special-token registration bug vs DiaConfig losing custom token IDs after save/load; both config/token metadata issues, not the same defect." }, { "left": "issue:44479", "right": "issue:44625", "accept": false, - "reason": "Video-input regression in Qwen VL models vs num_labels propagation in Qwen3.5; different config/input bugs." + "reason": "Video-input regression in Qwen VL models vs num_labels propagation in Qwen3.5; different model components." }, { "left": "issue:45491", "right": "issue:46002", "accept": false, - "reason": "Gemma3 mixed-length NaNs in embeddings vs NaNs in classification heads after checkout/init; both NaN reports but different layers and causes." + "reason": "Gemma3 mixed-length NaN embeddings vs NaNs in classification heads after checkout/init; both NaN reports but unrelated contexts." }, { "left": "issue:44484", "right": "issue:44704", "accept": false, - "reason": "A question about shard size vs AutoProcessor kwargs not reaching cached_file; unrelated." + "reason": "Question about save_pretrained shard size vs AutoProcessor kwargs forwarding bug; not duplicates." }, { "left": "issue:44821", "right": "issue:46821", "accept": false, - "reason": "AutoImageProcessor URL loading vs TimesFM window_size AttributeError; unrelated." + "reason": "AutoImageProcessor URL loading issue vs TimesFM window_size AttributeError; unrelated." }, { "left": "issue:33666", "right": "issue:36010", "accept": false, - "reason": "Qwen2-VL multi-GPU training vs GenerationMixin import error; unrelated." + "reason": "Qwen2-VL multi-GPU training issue vs GenerationMixin import error; unrelated." }, { "left": "issue:43550", "right": "issue:44361", "accept": false, - "reason": "Bamba torch.compile/SDPA failure vs MLukeTokenizer task AttributeError; unrelated." + "reason": "torch.compile/SDPA failure in Bamba vs MLukeTokenizer AttributeError; different subsystems." }, { "left": "issue:44466", "right": "issue:46050", "accept": false, - "reason": "lm_head.weight serialization inconsistency vs sinq quantized save/load breakage; different serialization bugs." + "reason": "Inconsistent lm_head serialization by device vs quantized-model save/load breakage; same general persistence area but different bugs." }, { - "left": "issue:43688", - "right": "issue:45440", + "left": "issue:32090", + "right": "issue:33666", "accept": false, - "reason": "Auxiliary-loss normalization vs DeepseekV3MoE divergence from remote implementation; unrelated model behavior." + "reason": "Trainer broadcast TypeError vs Qwen2-VL multi-GPU training; both training-related, not the same issue." }, { - "left": "issue:32090", - "right": "issue:33666", + "left": "issue:43688", + "right": "issue:45440", "accept": false, - "reason": "Trainer broadcast NoneType error vs Qwen2-VL multi-GPU training; both distributed-related but not the same bug." + "reason": "Auxiliary-loss normalization vs DeepSeekV3MoE divergence from remote implementation; different model logic." }, { "left": "issue:44977", "right": "issue:46693", "accept": false, - "reason": "Qwen3.5 flash-attention generation correctness bug vs FA2 long-context performance regression; correctness vs performance and different symptoms." + "reason": "Qwen3.5 flash-attention generation correctness regression vs FA2 long-context performance regression; related area, different problems." }, { - "left": "issue:44734", - "right": "issue:46612", + "left": "issue:44464", + "right": "issue:44977", "accept": false, - "reason": "Serve KV-cache continuation indexing crash vs beam-search cache reorder being skipped for several models; both cache-related but distinct failures." + "reason": "Compiled-forward chunked generation inconsistency vs flash-attention generation failure; not the same code-path issue." }, { - "left": "issue:44464", - "right": "issue:44977", + "left": "issue:44734", + "right": "issue:46612", "accept": false, - "reason": "Compiled-forward chunked generation inconsistency vs Qwen3.5 flash-attention generation bug; different regressions." + "reason": "KV-cache continuation indexing crash in serve vs beam-search cache reorder skipped for certain models; both cache-related but distinct bugs." }, { "left": "issue:41720", "right": "issue:44464", "accept": false, - "reason": "Auto device mapping cudaErrorAssert vs chunked generation inconsistency; unrelated code paths." + "reason": "Auto device mapping CUDA assert vs chunked generation/compiled forward inconsistency; unrelated." }, { "left": "issue:43504", "right": "issue:43550", "accept": false, - "reason": "Beit legacy-field load bug vs Bamba compile/SDPA bug; unrelated." + "reason": "Legacy field prevents Beit semantic-segmentation load vs Bamba torch.compile SDPA failure; unrelated." }, { "left": "issue:43504", "right": "issue:43720", "accept": false, - "reason": "Beit legacy-field load bug vs BitNet packed-weight loading bug; unrelated." + "reason": "Legacy-field load failure vs BitNet packed-weight unpacking failure; different model-loading defects." }, { - "left": "issue:43577", - "right": "issue:44554", + "left": "issue:43782", + "right": "issue:46693", "accept": false, - "reason": "BLIP2 dtype propagation issue vs an MPS attention correctness issue; different models and failure modes." + "reason": "Qwen3VL weight_only loading error vs FA2 long-context performance regression; different failure modes." } ] }, @@ -8089,7 +8090,7 @@ { "label": "issue", "component_index": 1, - "component_count": 25, + "component_count": 26, "review_unit_index": 26, "review_unit_count": 61, "status": "reviewed", @@ -8113,13 +8114,13 @@ "issue:43504", "issue:43525", "issue:43572", + "issue:43577", "issue:43606", "issue:43611", "issue:43643", "issue:43646", "issue:43688", "issue:43720", - "issue:43782", "issue:43825", "issue:43844", "issue:43927", @@ -8144,25 +8145,24 @@ "issue:46209", "issue:46365", "issue:46489", - "issue:46693", "issue:46710", "issue:46846" ], "soft_pairs": [ - "issue:43782|issue:46693", + "issue:43577|issue:44554", "issue:43611|issue:46846", "issue:43278|issue:43688", "issue:33666|issue:36331", "issue:43262|issue:43295", "issue:43720|issue:43927", - "issue:43643|issue:44561", "issue:43606|issue:43927", "issue:43504|issue:44610", + "issue:43643|issue:44561", "issue:39692|issue:43493", "issue:44514|issue:45290", "issue:44964|issue:45405", - "issue:43493|issue:43825", "issue:43720|issue:44361", + "issue:43493|issue:43825", "issue:44554|issue:46365", "issue:43525|issue:43572", "issue:44977|issue:46209", @@ -8175,236 +8175,233 @@ "issue:39692|issue:43295", "issue:43493|issue:43844", "issue:43278|issue:43844", - "issue:44568|issue:46489", "issue:43646|issue:44466", + "issue:44568|issue:46489", "issue:46209|issue:46710", "issue:33357|issue:33666" ], - "prepared_review_unit_hash": "1840da1831d362ebfab24cb28cef39cc99264627f7c8faeaafd3d3e61e8d3bb4", + "prepared_review_unit_hash": "cd5030d7f39e558e999f63331cc1f985bd45f0b417141ef420c4ee99d55f47b4", "component_budget": { "node_count": 505, "item_count": 505, "soft_pair_count": 1854, - "serialized_chars": 962354, - "estimated_input_tokens": 240589, - "estimated_eval_tokens": 481434 + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 }, "budget": { - "node_count": 48, - "item_count": 48, + "node_count": 47, + "item_count": 47, "soft_pair_count": 30, - "serialized_chars": 36860, - "estimated_input_tokens": 9215, - "estimated_eval_tokens": 18686 + "serialized_chars": 36349, + "estimated_input_tokens": 9088, + "estimated_eval_tokens": 18432 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "Most pairs are superficial similarity matches across unrelated Transformers regressions. Only one clear duplicate pair appears: the apply_chat_template crash with assistant tool-call messages and no content.", - "confidence": 0.91, - "canonical_issue_reason": "issue:44514 is the broader, earlier report of the apply_chat_template crash; it captures the core bug that also appears in 45290, which is just a narrower repro variant.", + "summary": "The set is heterogeneous and the proposed soft links are mostly title-level similarity, not the same concrete bug or change. I found no strong duplicate cluster and would not merge any of the soft pairs.", + "confidence": 0.89, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "issue:44514 is the best issue to anchor this duplicate pair because it states the underlying failure mode most generally and is the clearest target for consolidation.", + "best_issue_reason": null, "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:43782", - "right": "issue:46693", + "left": "issue:43577", + "right": "issue:44554", "accept": false, - "reason": "Different bugs: Qwen3VL from_pretrained weight_only error vs FA2 long-context performance regression." + "reason": "Different problems: one is dtype staying fp32 in BLIP2 loading; the other is an MPS attention correctness issue with mismatched head dims." }, { "left": "issue:43611", "right": "issue:46846", "accept": false, - "reason": "Both concern model loading/TP plans, but one is base_model_prefix loading and the other is missing linear_attn TP entries; distinct code paths." + "reason": "Both mention model loading, but one is a base_model_prefix regression and the other is a tensor-parallel planning/OOM issue in Qwen3.5." }, { "left": "issue:43278", "right": "issue:43688", "accept": false, - "reason": "Unrelated training/eval dtype mismatch vs auxiliary-loss normalization bug." + "reason": "Both involve numerical behavior, but one is eval/training dtype drift and the other is auxiliary-loss normalization in mixture-of-experts models." }, { "left": "issue:33666", "right": "issue:36331", "accept": false, - "reason": "Different issues: Qwen2-VL multi-GPU training vs Trainer custom compute_loss signature breakage." + "reason": "Completely different areas: Qwen2-VL multi-GPU training vs a CustomTrainer compute_loss signature regression." }, { "left": "issue:43262", "right": "issue:43295", "accept": false, - "reason": "Audio chat-template sampling-rate default vs multimodal processor.tokenizer regression; not the same defect." + "reason": "Audio processor sampling-rate defaulting and processor.tokenizer/image-handling regression are separate processor bugs." }, { "left": "issue:43720", "right": "issue:43927", "accept": false, - "reason": "Packed-weight loading bug in BitNet vs DiaConfig token-id persistence issue." - }, - { - "left": "issue:43643", - "right": "issue:44561", - "accept": false, - "reason": "Both involve trust_remote_code/v5 changes, but one is missing fields in AutoConfig and the other is removal of is_torch_fx_available; not the same bug." + "reason": "Different failure modes: unpacked packed weights during accelerate loading vs DiaConfig losing custom token IDs after save/load." }, { "left": "issue:43606", "right": "issue:43927", "accept": false, - "reason": "CPU offload device mismatch vs custom token IDs lost on save/load; unrelated." + "reason": "Unrelated code paths: CPU offload device mismatch in Bark vs config serialization/token-ID loss in Dia." }, { "left": "issue:43504", "right": "issue:44610", "accept": false, - "reason": "Legacy preset-loading failure in BEiT vs processor/model input-size mismatch in OmDet-Turbo." + "reason": "Different models and bugs: Beit legacy-field loading vs OmDet-Turbo processor/model input-size mismatch." + }, + { + "left": "issue:43643", + "right": "issue:44561", + "accept": false, + "reason": "One is missing fields from AutoConfig with trust_remote_code; the other is a missing utility function breaking remote-code models." }, { "left": "issue:39692", "right": "issue:43493", "accept": false, - "reason": "SigLIP2 doc-example problems vs SigLIP2 implementation discrepancy with JAX; different underlying issues." + "reason": "Both concern SigLIP2, but one is a documentation/example issue and the other is an implementation mismatch with the original JAX model." }, { "left": "issue:44514", "right": "issue:45290", - "accept": true, - "reason": "Same crash: apply_chat_template fails on assistant messages with tool calls and no content; tokenize=True is just a narrower reproducer." + "accept": false, + "reason": "Both are chat-template related, but one crashes on batched padding=False input and the other on assistant tool-call messages with no content." }, { "left": "issue:44964", "right": "issue:45405", "accept": false, - "reason": "Model loading failure vs a version pin / release availability issue for peft." + "reason": "Model loading failure vs dependency-version bump; the latter is about an unreleased PEFT minimum version, not the Phi-4 multimodal bug." }, { - "left": "issue:43493", - "right": "issue:43825", + "left": "issue:43720", + "right": "issue:44361", "accept": false, - "reason": "SigLIP2 numerical discrepancy vs pipeline translation-task error message; unrelated." + "reason": "Different tokenizer/model areas: BitNet accelerate loading of packed weights vs MLukeTokenizer AttributeError on tasks." }, { - "left": "issue:43720", - "right": "issue:44361", + "left": "issue:43493", + "right": "issue:43825", "accept": false, - "reason": "Packed-weight loading bug vs MLukeTokenizer AttributeError; different subsystems." + "reason": "SigLIP2 implementation discrepancy and pipeline error-message wording are unrelated issues." }, { "left": "issue:44554", "right": "issue:46365", "accept": false, - "reason": "MPS attention correctness issue vs Trainer import conflict with datasets; unrelated." + "reason": "Attention correctness on MPS is unrelated to an import-time Trainer/datasets dependency problem." }, { "left": "issue:43525", "right": "issue:43572", "accept": false, - "reason": "Both are config token-id attribute problems, but they affect different models and fields (pad_token_id vs pad_token_idx) with different failure modes." + "reason": "Both mention missing token-related fields, but Llama4 pad_token_id and StableLm pad_token_idx are different configs and regressions." }, { "left": "issue:44977", "right": "issue:46209", "accept": false, - "reason": "Qwen3.5 flash-attention generation bug vs FP8 B200 inference failure; not the same root cause." + "reason": "Qwen3.5 flash-attention generation bug is unrelated to FP8 inference failure on B200 hardware." }, { "left": "issue:44625", "right": "issue:45584", "accept": false, - "reason": "num_labels propagation issue vs Whisper empty-transcription failure after special-token alignment." + "reason": "Different models and symptoms: Qwen3.5 num_labels propagation vs Whisper empty-transcription failure after token alignment." }, { "left": "issue:43452", "right": "issue:44451", "accept": false, - "reason": "General gguf_file loading breakage vs a specific model load failure; too little evidence they share one code-path." + "reason": "gguf_file loading regression is not the same as inability to load a specific pretrained model." }, { "left": "issue:45823", "right": "issue:46050", "accept": false, - "reason": "Gemma4 device mismatch with auto device_map vs quantized model save/load corruption; different bugs." + "reason": "Device mismatch during forward with device_map=auto is different from quantized-model save/load corruption." }, { "left": "issue:28218", "right": "issue:33357", "accept": false, - "reason": "Tokenizer added-space bug vs MacOS bus error in a CLIP model; unrelated." + "reason": "Tokenizer extra-space behavior and a Mac bus error on CLIP are unrelated; similarity is only that both affect tokenization/model usage." }, { "left": "issue:43994", "right": "issue:46192", "accept": false, - "reason": "SigLIP2 wrong outputs vs return_tensors being ignored with text_kwargs; not the same issue." + "reason": "SigLIP2 wrong outputs with AutoModel/pipeline and return_tensors being ignored when text_kwargs is passed are separate API bugs." }, { "left": "issue:42796", "right": "issue:44977", "accept": false, - "reason": "Tokenizer BPE-marker handling bug vs Qwen3.5 flash-attention generation issue; unrelated." + "reason": "Ministral BPE-marker handling and Qwen3.5 flash-attention generation are different tokenizer vs inference-path issues." }, { "left": "issue:39692", "right": "issue:43295", "accept": false, - "reason": "Doc-example errors in SigLIP2 vs processor.tokenizer/images regression in custom multimodal code; different bugs." + "reason": "Documentation example errors for SigLIP2 and a regression in processor.tokenizer/image handling are not the same bug." }, { "left": "issue:43493", "right": "issue:43844", "accept": false, - "reason": "SigLIP2 implementation discrepancy vs gradient explosion with HfDeepSpeedConfig + ZeRO-3; unrelated." + "reason": "SigLIP2 implementation mismatch and gradient blow-up under DeepSpeed ZeRO-3 are unrelated training dynamics bugs." }, { "left": "issue:43278", "right": "issue:43844", "accept": false, - "reason": "Embedding dtype mismatch vs gradient blow-up under ZeRO-3; different failure modes." + "reason": "A dtype mismatch in evaluation is not the same as abnormal gradient growth during ZeRO-3 training." }, { - "left": "issue:44568", - "right": "issue:46489", + "left": "issue:43646", + "right": "issue:44466", "accept": false, - "reason": "Different tokenizer regressions affecting different models; not one concrete bug." + "reason": "Custom model initialization breakage in v5 and tied lm_head serialization depending on device are different regressions." }, { - "left": "issue:43646", - "right": "issue:44466", + "left": "issue:44568", + "right": "issue:46489", "accept": false, - "reason": "Custom model initialization breakage vs lm_head.weight serialization inconsistency; different code paths." + "reason": "Both concern tokenizer behavior, but one is missing BOS/EOS addition and the other is DeepSeek-Coder BPE output corruption after a prior fix." }, { "left": "issue:46209", "right": "issue:46710", "accept": false, - "reason": "FP8 inference on B200 vs DeepSeek accuracy regression after upgrade; not the same bug." + "reason": "Hardware-specific FP8 inference failure is not the same as an accuracy regression on DeepSeek-R1 after upgrade." }, { "left": "issue:33357", "right": "issue:33666", "accept": false, - "reason": "Bus error on CLIP model vs Qwen2-VL multi-GPU training request; unrelated." + "reason": "A macOS bus error in CLIP and Qwen2-VL multi-GPU training are unrelated issues." } ] }, - "evaluator_result": { - "accept": true, - "feedback": "The summary is well grounded in the packet: the only accepted duplicate pair is indeed the apply_chat_template crash between 44514 and 45290, and the other soft-edge pairs are rejected conservatively with issue-specific reasons. No overstatement of evidence stands out." - }, - "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 }, { "label": "issue", "component_index": 1, - "component_count": 25, + "component_count": 26, "review_unit_index": 27, "review_unit_count": 61, "status": "reviewed", @@ -8443,7 +8440,6 @@ "issue:44849", "issue:44964", "issue:45072", - "issue:45200", "issue:45229", "issue:45245", "issue:45357", @@ -8466,23 +8462,23 @@ "soft_pairs": [ "issue:44488|issue:44843", "issue:44849|issue:46209", - "issue:44568|issue:45812", "issue:43278|issue:43701", + "issue:44568|issue:45812", "issue:44792|issue:45405", "issue:43927|issue:44361", "issue:43575|issue:45405", "issue:43824|issue:44589", "issue:44849|issue:45357", "issue:46821|issue:46832", - "issue:43257|issue:43299", "issue:43504|issue:43606", + "issue:43257|issue:43299", "issue:45245|issue:46710", "issue:43334|issue:45759", "issue:44493|issue:46285", "issue:43334|issue:43572", - "issue:46399|issue:46846", "issue:44589|issue:45520", "issue:43572|issue:44291", + "issue:46399|issue:46846", "issue:43606|issue:44112", "issue:45072|issue:46832", "issue:43504|issue:43653", @@ -8491,211 +8487,210 @@ "issue:43873|issue:46050", "issue:43844|issue:44964", "issue:43994|issue:44493", - "issue:45229|issue:45823", - "issue:45200|issue:45229" + "issue:45229|issue:45823" ], - "prepared_review_unit_hash": "7ebbe6ddc886cd0ae6c288fb1a63b0d60f3d51997c7ceb32ce1bde34a7dd2b61", + "prepared_review_unit_hash": "50ca0adb3d94b05cbfbb1a77d60381aab80fcd7a0167d5280eb3b4c7bcf3c571", "component_budget": { "node_count": 505, "item_count": 505, "soft_pair_count": 1854, - "serialized_chars": 962354, - "estimated_input_tokens": 240589, - "estimated_eval_tokens": 481434 + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 }, "budget": { - "node_count": 48, - "item_count": 48, - "soft_pair_count": 29, - "serialized_chars": 36504, - "estimated_input_tokens": 9126, - "estimated_eval_tokens": 18508 + "node_count": 47, + "item_count": 47, + "soft_pair_count": 28, + "serialized_chars": 35616, + "estimated_input_tokens": 8904, + "estimated_eval_tokens": 18064 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly a loose set of unrelated model-loading, tokenizer, config, and device/offload regressions. None of the soft-pair candidates look like true duplicates; they mostly share broad Transformer-version or Qwen/Gemma adjacency rather than the same concrete bug.", - "confidence": 0.76, - "canonical_issue_reason": "issue:43824 is the clearest representative standalone bug: a high-discussion, user-visible loading/import regression with multiple references, and it fits the general dev-regression theme better than the more model-specific entries.", + "summary": "The set is highly heterogeneous: most items are distinct model-loading, tokenizer, device/offload, quantization, or CI regressions across different model families. None of the proposed soft edges look like the same underlying bug/change, so they should all be rejected.", + "confidence": 0.95, + "canonical_issue_reason": "No safe canonical issue: these issues span unrelated regressions (Qwen/VL loading, tokenizer special-token handling, offload/device mismatches, quantization persistence, config errors, CI failures), so a single issue would be misleading as a duplicate target.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43824 has the strongest signal for a canonical issue in this cluster: broad impact, clear failure mode, and stronger discussion/reference activity than most others.", + "best_issue_reason": "Issue 44843 is the broadest single user-facing regression in the set, but it only covers the AutoTokenizer/_patch_mistral_regex offline-path bug and does not subsume the rest.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44488", "right": "issue:44843", "accept": false, - "reason": "Different failures: model loading for cjvt/sleng-bert vs offline-mode break from unconditional model_info() lookup." + "reason": "Both are model-loading related, but one is a specific cjvt/sleng-bert load failure and the other is an HF_HUB_OFFLINE/model_info regression; different code paths." }, { "left": "issue:44849", "right": "issue:46209", "accept": false, - "reason": "Same family name, but different bugs: hidden-states output issue vs FP8 inference failure on B200." + "reason": "Qwen3.5 hidden-state output bug vs FP8 inference on B200 are unrelated features and failure modes." }, { - "left": "issue:44568", - "right": "issue:45812", + "left": "issue:43278", + "right": "issue:43701", "accept": false, - "reason": "Both are tokenizer regressions, but they affect different models and different code paths." + "reason": "BF16-to-FP32 eval dtype change is unrelated to a resume_from_checkpoint key mismatch." }, { - "left": "issue:43278", - "right": "issue:43701", + "left": "issue:44568", + "right": "issue:45812", "accept": false, - "reason": "Embedding dtype drift in eval is unrelated to resume_from_checkpoint key mismatch." + "reason": "Both concern tokenizer behavior, but one is mDeBERTa special-token insertion and the other is a Granite model token-id regression; not the same bug." }, { "left": "issue:44792", "right": "issue:45405", "accept": false, - "reason": "Janus image-generation test failure is unrelated to the unreleased PEFT version bump." + "reason": "A janus image-generation test failure is unrelated to a PEFT version bump." }, { "left": "issue:43927", "right": "issue:44361", "accept": false, - "reason": "Both involve config/tokenizer behavior, but one is custom token IDs on save/load and the other is an AttributeError in MLukeTokenizer tasks." + "reason": "DiaConfig custom token IDs after save/load and MLukeTokenizer task AttributeError involve different models and different failure paths." }, { "left": "issue:43575", "right": "issue:45405", "accept": false, - "reason": "Qwen2 TP OOM is unrelated to the PEFT release-version issue." + "reason": "Qwen2 TP OOM has nothing to do with the MIN_PEFT_VERSION release issue." }, { "left": "issue:43824", "right": "issue:44589", "accept": false, - "reason": "ImportError for a missing model class is a different problem from Float8 storage lookup failure." + "reason": "Missing Qwen2_5_VL import and Float8 storage lookup failure are unrelated import/runtime errors." }, { "left": "issue:44849", "right": "issue:45357", "accept": false, - "reason": "Same model family, but forward-output bug and save_pretrained key corruption are different code paths." + "reason": "Qwen3.5 hidden-states behavior and Qwen3.5 visual-encoder save_pretrained keys are different bugs in different subsystems." }, { "left": "issue:46821", "right": "issue:46832", "accept": false, - "reason": "Different models and different runtime errors: TimesFM window_size AttributeError vs RT-DETR feature-level TypeError." + "reason": "TimesFM window_size AttributeError and RT-DETR feature-level mismatch are unrelated model-specific issues." }, { - "left": "issue:43257", - "right": "issue:43299", + "left": "issue:43504", + "right": "issue:43606", "accept": false, - "reason": "Both mention Qwen MOE loading, but one is weight conversion with accelerate+deepspeed and the other is a Qwen3VL dev-version load regression." + "reason": "BEiT legacy-field loading and bark-small CPU offload device mismatch do not share the same underlying code path." }, { - "left": "issue:43504", - "right": "issue:43606", + "left": "issue:43257", + "right": "issue:43299", "accept": false, - "reason": "BEiT legacy-field loading and bark-small CPU offload device mismatch are unrelated bugs." + "reason": "Both mention Qwen3 MoE loading, but one is accelerate+deepspeed weight conversion and the other is Qwen3VL loading in dev; too different to merge." }, { "left": "issue:45245", "right": "issue:46710", "accept": false, - "reason": "Category-count overflow and accuracy regression after upgrade are different symptoms and likely different causes." + "reason": "2^24 category limit and DeepSeek accuracy regression are unrelated." }, { "left": "issue:43334", "right": "issue:45759", "accept": false, - "reason": "Both are Qwen multimodal config issues, but one is a missing pad_token_id attribute and the other is from_config not unwrapping text_config." + "reason": "Both touch Qwen multimodal configs, but one is a missing pad_token_id attribute and the other is from_config not unwrapping text_config; different fixes." }, { "left": "issue:44493", "right": "issue:46285", "accept": false, - "reason": "Unexpected key position-id failures are not the same as the CLIPTextModel structural inconsistency." + "reason": "Unexpected position-id keys across many models vs CLIP text-model structure inconsistency are different regressions." }, { "left": "issue:43334", "right": "issue:43572", "accept": false, - "reason": "Both are missing config fields after version updates, but they affect different models and different attributes." - }, - { - "left": "issue:46399", - "right": "issue:46846", - "accept": false, - "reason": "Qwen3.5 kernel registration error is unrelated to the TP-plan linear_attn OOM/shape issue." + "reason": "Qwen3-VL pad_token_id and StableLm pad_token_idx are different missing-field errors in different configs." }, { "left": "issue:44589", "right": "issue:45520", "accept": false, - "reason": "Different backend availability failures: Float8 storage lookup vs flash_attn import key error on Python 3.13." + "reason": "Float8 storage lookup failure and flash_attn KeyError on Python 3.13 are unrelated dependency/runtime issues." }, { "left": "issue:43572", "right": "issue:44291", "accept": false, - "reason": "StableLmConfig pad_token_idx loss is unrelated to the unexpected _is_hf_initialized argument in init_empty_weights." + "reason": "StableLm missing pad_token_idx and init_empty_weights unexpected _is_hf_initialized argument are separate config/API regressions." + }, + { + "left": "issue:46399", + "right": "issue:46846", + "accept": false, + "reason": "Qwen3.5 missing decorator for apply_rotary_pos_emb registration and missing linear_attn TP-plan entries are different Qwen3.5 bugs." }, { "left": "issue:43606", "right": "issue:44112", "accept": false, - "reason": "The CI test failure is not the same underlying bug as the real CPU offload device mismatch." + "reason": "CPU offload device mismatch and a stale GraniteSpeech CI override test are not the same bug." }, { "left": "issue:45072", "right": "issue:46832", "accept": false, - "reason": "dtype mismatches in bfloat16 inference are unrelated to RT-DETR backbone feature-level mismatch." + "reason": "bfloat16 dtype mismatches in SwitchTransformers/TimmWrapperModel are unrelated to RT-DETR feature-level shape errors." }, { "left": "issue:43504", "right": "issue:43653", "accept": false, - "reason": "BEiT legacy-field loading and BigBirdTokenizer special-token registration are different issues." + "reason": "BEiT legacy-field loading and BigBirdTokenizer mask-token registration are different issues despite both involving model/tokenizer setup." }, { "left": "issue:44464", "right": "issue:46424", "accept": false, - "reason": "Both touch generation/cache behavior, but chunked-generation inconsistency is not the same bug as max_cache_length only applying to quantized cache." + "reason": "Chunked generation inconsistency with compiled forward is not the same as cache max length only applying to quantized cache." }, { "left": "issue:43937", "right": "issue:46701", "accept": false, - "reason": "GLM-5 GenerationConfig validation and DiffusionGemma pipeline failure are separate model-specific issues." + "reason": "Invalid GenerationConfig for GLM-5 and DiffusionGemma pipeline failure are unrelated generation problems." }, { "left": "issue:43873", "right": "issue:46050", "accept": false, - "reason": "Quantization/offloading behavior and save/load breakage after sinq quantization are different bugs." + "reason": "Quantization/offload not working and quantized model save/load breakage are related broadly, but they describe different concrete failures and paths." }, { "left": "issue:43844", "right": "issue:44964", "accept": false, - "reason": "ZeRO-3 gradient growth is unrelated to the Phi-4 multimodal loading failure." + "reason": "ZeRO-3 gradient growth and Phi-4 multimodal loading failure are unrelated." }, { - "left": "issue:43994", - "right": "issue:44493", + "left": "issue:45229", + "right": "issue:45823", "accept": false, - "reason": "SigLIP2 nonsensical outputs and unexpected position-id keys are different failures with different causes." + "reason": "Gemma4 multi-GPU OOM and Gemma4 PLE device mismatch both mention Gemma4, but one is memory pressure and the other is a forward device-placement bug." }, { - "left": "issue:45229", - "right": "issue:45823", + "left": "issue:45357", + "right": "issue:45812", "accept": false, - "reason": "Both are Gemma4 and multi-GPU related, but one is CUDA OOM and the other is a forward device-mismatch bug." + "reason": "Qwen3.5 visual-encoder key regression and Granite tokenizer ID regression are different model families and different serialization/tokenization problems." }, { - "left": "issue:45200", - "right": "issue:45229", + "left": "issue:46050", + "right": "issue:46873", "accept": false, - "reason": "Text-only mm_token_type_ids defaulting and multi-GPU CUDA OOM are not the same bug." + "reason": "No such pair in the input cluster." } ] }, @@ -8709,7 +8704,7 @@ { "label": "issue", "component_index": 1, - "component_count": 25, + "component_count": 26, "review_unit_index": 28, "review_unit_count": 61, "status": "reviewed", @@ -8735,11 +8730,9 @@ "issue:43450", "issue:43493", "issue:43540", - "issue:43550", "issue:43572", "issue:43696", "issue:43701", - "issue:43844", "issue:43873", "issue:43957", "issue:44186", @@ -8753,6 +8746,7 @@ "issue:45084", "issue:45198", "issue:45200", + "issue:45229", "issue:45406", "issue:45446", "issue:45520", @@ -8770,12 +8764,13 @@ ], "soft_pairs": [ "issue:44869|issue:46752", + "issue:45200|issue:45229", "issue:45588|issue:46050", - "issue:42490|issue:43696", "issue:29159|issue:32090", + "issue:42490|issue:43696", "issue:29942|issue:30333", - "issue:44977|issue:46701", "issue:28218|issue:34567", + "issue:44977|issue:46701", "issue:45084|issue:45520", "issue:43065|issue:46369", "issue:15354|issue:38175", @@ -8795,205 +8790,204 @@ "issue:45823|issue:46192", "issue:43572|issue:46821", "issue:43295|issue:44610", - "issue:45200|issue:45823", - "issue:43550|issue:43844" + "issue:45200|issue:45823" ], - "prepared_review_unit_hash": "e73a609c007369e590e57d679ab0cc27652e80d2f16b303626b63d207b0365c0", + "prepared_review_unit_hash": "e374a2c507739cae09605455730dcbd67f59963d4d3aa02b8c79d0a7b28a4e4f", "component_budget": { "node_count": 505, "item_count": 505, "soft_pair_count": 1854, - "serialized_chars": 962354, - "estimated_input_tokens": 240589, - "estimated_eval_tokens": 481434 + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 }, "budget": { - "node_count": 48, - "item_count": 48, + "node_count": 47, + "item_count": 47, "soft_pair_count": 28, - "serialized_chars": 36143, - "estimated_input_tokens": 9036, - "estimated_eval_tokens": 18328 + "serialized_chars": 35531, + "estimated_input_tokens": 8883, + "estimated_eval_tokens": 18022 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This set is heterogeneous: the soft pairs span unrelated tokenizer, serving, attention, model-loading, and training issues. None of the candidate pairs looks like the same underlying bug/change, so there is no clear canonical issue for the whole set.", - "confidence": 0.91, - "canonical_issue_reason": "No single canonical issue fits: the candidates are scattered across unrelated bugs, so this cluster does not collapse to one duplicate topic.", + "summary": "This cluster is heterogeneous rather than duplicate-heavy: it mixes unrelated tokenizer, serving, quantization, device-mapping, and model-specific bugs. I rejected every soft pair as a true duplicate.", + "confidence": 0.97, + "canonical_issue_reason": "No real canonical duplicate set here; the items span unrelated bugs. If a representative must be chosen, issue:43295 is the cleanest regression report, but it does not subsume the others.", "canonical_pr_reason": null, - "best_issue_reason": "None of the issues is a strong global representative because the set is not a coherent duplicate cluster.", + "best_issue_reason": "issue:43295 is the strongest standalone representative because it has a concrete regression description and user-impacting repro, but the cluster is too heterogeneous to treat it as a true duplicate anchor.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44869", "right": "issue:46752", "accept": false, - "reason": "Different crash sites and behaviors: word timestamp decoding vs empty chat-template input handling." + "reason": "Different failure modes and code paths: Whisper word-timestamp decoding vs apply_chat_template on an empty list." }, { - "left": "issue:45588", - "right": "issue:46050", + "left": "issue:45200", + "right": "issue:45229", "accept": false, - "reason": "Flash-attention AttributeError on sink-less models is unrelated to quantization save/load breakage." + "reason": "Gemma 4 multimodal token-type defaults vs a multi-GPU OOM report; not the same bug." }, { - "left": "issue:42490", - "right": "issue:43696", + "left": "issue:45588", + "right": "issue:46050", "accept": false, - "reason": "Local-path trailing-slash import warning is unrelated to a CUDA out-of-memory failure." + "reason": "flash_attention.py s_aux=None crash and sinq save/load breakage are unrelated." }, { "left": "issue:29159", "right": "issue:32090", "accept": false, - "reason": "Tokenizer slow/fast inconsistency is unrelated to Trainer broadcasting NoneType values." + "reason": "Tokenizer slow/fast inconsistency vs Trainer _gpu_broadcast_one NoneType error." }, { - "left": "issue:29942", - "right": "issue:30333", + "left": "issue:42490", + "right": "issue:43696", "accept": false, - "reason": "Failing Flash Attention 2 tests and MLflow reporting on failed jobs are different bugs." + "reason": "Local-path import warning and GPT-oss CUDA OOM are unrelated." }, { - "left": "issue:44977", - "right": "issue:46701", + "left": "issue:29942", + "right": "issue:30333", "accept": false, - "reason": "Qwen3.5 flash-attention generation bug and DiffusionGemma pipeline failure are different code paths." + "reason": "Flash Attention 2 test failures vs MLFlow reporting after failed training jobs." }, { "left": "issue:28218", "right": "issue:34567", "accept": false, - "reason": "Added-token spacing behavior is unrelated to TrainerState token-count tracking." + "reason": "Added-token spacing bug vs TrainerState num_input_tokens_seen not updating." + }, + { + "left": "issue:44977", + "right": "issue:46701", + "accept": false, + "reason": "Both involve generation, but Qwen3.5 flash-attention behavior and DiffusionGemma pipeline failure are different code paths." }, { "left": "issue:45084", "right": "issue:45520", "accept": false, - "reason": "Template-node compilation error is unrelated to the Python 3.13 flash_attn import KeyError." + "reason": "Compilation error for non-template nodes vs Python 3.13 flash_attn import KeyError." }, { "left": "issue:43065", "right": "issue:46369", "accept": false, - "reason": "Dummy Conv2d in Sam3PixelDecoder is not the same as a memory overflow in sam3-litetext." + "reason": "Same model family, but dummy Conv2d design and litetext memory overflow are not the same underlying bug." }, { "left": "issue:15354", "right": "issue:38175", "accept": false, - "reason": "torch.jit.script GeneratorExp support and SigLIP2 zero-probabilities are unrelated model issues." + "reason": "torch.jit.script GeneratorExp support and SigLIP2 zero-probability output are unrelated." }, { "left": "issue:44554", "right": "issue:45446", "accept": false, - "reason": "MPS attention correctness and a PyTorch version-check bug are different problems." + "reason": "MPS attention correctness issue and flex_attention import version check are different problems." }, { "left": "issue:41720", "right": "issue:46710", "accept": false, - "reason": "A800 auto device-map cuda assert is unrelated to an accuracy regression after version upgrade." + "reason": "Qwen3 device-mapping CUDA assert and DeepSeek accuracy regression are unrelated." }, { "left": "issue:43493", "right": "issue:43701", "accept": false, - "reason": "SigLIP2 implementation discrepancy and checkpoint key mismatch are distinct failures." + "reason": "SigLIP2 HF-vs-JAX discrepancy and resume_from_checkpoint key mismatch do not describe the same bug." }, { "left": "issue:44423", "right": "issue:44734", "accept": false, - "reason": "Continuous-batching multimodal crash and KV-cache continuation indexing bug are different serve paths." + "reason": "Both are serve crashes, but multimodal continuous-batching and KV-cache continuation indexing are distinct issues." }, { "left": "issue:45406", "right": "issue:46121", "accept": false, - "reason": "Gemma4Processor missing _tokenizer is unrelated to JSON-list rope parameter validation." + "reason": "Gemma4Processor missing _tokenizer and rope-params list type handling are unrelated." }, { "left": "issue:44977", "right": "issue:46710", "accept": false, - "reason": "Flash-attention generation failure for Qwen3.5 is unrelated to the DeepSeek accuracy regression." + "reason": "Qwen3.5 flash-attention generation issue is not the same as a DeepSeek accuracy regression." }, { "left": "issue:37428", "right": "issue:43957", "accept": false, - "reason": "Flash-attention utility import failure and meta-device model-loading breakage are separate issues." + "reason": "Flash-attention import name breakage and meta-device model loading failures are separate bugs." }, { "left": "issue:43295", "right": "issue:43450", "accept": false, - "reason": "Processor/tokenizer regression is unrelated to video processor batch-shape handling." + "reason": "Processor.tokenizer/images regression vs batched video-shape bug; both are processor-related but not the same defect." }, { "left": "issue:45072", "right": "issue:45198", "accept": false, - "reason": "bfloat16 dtype mismatches are not the same as Wav2Vec2 save/tokenization failures." + "reason": "bfloat16 dtype mismatch CI issue and Wav2Vec2 save/tokenization failure are unrelated." }, { "left": "issue:45588", "right": "issue:45759", "accept": false, - "reason": "Flash-attention sinkless-model crash and composite-config text_config unwrapping are different bugs." + "reason": "flash_attention sinkless-model crash and multimodal text_config unwrapping are different code paths." }, { "left": "issue:43540", "right": "issue:45200", "accept": false, - "reason": "Video input processing ValueError is unrelated to Gemma 4 mm_token_type_ids defaults." + "reason": "Video-input processing error in Qwen3OmniMoe and Gemma 4 token-type defaults are unrelated." }, { "left": "issue:43295", "right": "issue:43873", "accept": false, - "reason": "Processor/tokenizer regression is unrelated to offloading behavior with quantization." + "reason": "Processor regression and quantized offloading failure are different issues." }, { "left": "issue:44186", "right": "issue:46192", "accept": false, - "reason": "LayoutLMv2Tokenizer padding/NER crash and return_tensors being ignored are separate tokenizer issues." + "reason": "LayoutLMv2Tokenizer crash on NER/padding and return_tensors handling are distinct tokenizer behaviors." }, { "left": "issue:45823", "right": "issue:46192", "accept": false, - "reason": "Gemma4 PLE device mismatch is unrelated to return_tensors handling." + "reason": "Gemma4 device-mapping mismatch and return_tensors handling are unrelated." }, { "left": "issue:43572", "right": "issue:46821", "accept": false, - "reason": "StableLmConfig pad_token_idx regression and TimesFM window_size AttributeError are different config bugs." + "reason": "StableLmConfig pad_token_idx missing and TimesFM window_size AttributeError are different model/config bugs." }, { "left": "issue:43295", "right": "issue:44610", "accept": false, - "reason": "Processor/tokenizer regression is not the same as OmDet-Turbo processor size mismatch." + "reason": "Processor API regression and OmDet-Turbo input-size mismatch are separate bugs." }, { "left": "issue:45200", "right": "issue:45823", "accept": false, - "reason": "Missing default mm_token_type_ids and PLE device mismatch affect different Gemma 4 paths." - }, - { - "left": "issue:43550", - "right": "issue:43844", - "accept": false, - "reason": "Bamba torch.compile SDPA failure and ZeRO-3 gradient inflation are unrelated training/runtime bugs." + "reason": "Gemma4 token-type defaulting and Gemma4 PLE device mismatch are different failure modes." } ] }, @@ -9007,7 +9001,7 @@ { "label": "issue", "component_index": 1, - "component_count": 25, + "component_count": 26, "review_unit_index": 29, "review_unit_count": 61, "status": "reviewed", @@ -9034,15 +9028,16 @@ "issue:43452", "issue:43493", "issue:43525", + "issue:43550", "issue:43572", "issue:43575", "issue:43653", "issue:43824", + "issue:43844", "issue:43856", "issue:43950", "issue:43975", "issue:44112", - "issue:44442", "issue:44466", "issue:44521", "issue:44554", @@ -9050,7 +9045,6 @@ "issue:44849", "issue:44987", "issue:45127", - "issue:45198", "issue:45200", "issue:45290", "issue:45356", @@ -9066,6 +9060,7 @@ "issue:46932" ], "soft_pairs": [ + "issue:43550|issue:43844", "issue:29870|issue:29942", "issue:43450|issue:43575", "issue:45520|issue:46178", @@ -9085,204 +9080,197 @@ "issue:42915|issue:43824", "issue:43824|issue:43975", "issue:43653|issue:44112", - "issue:43950|issue:46285", "issue:43452|issue:46512", + "issue:43950|issue:46285", "issue:43262|issue:43450", "issue:43334|issue:43525", "issue:45200|issue:46899", - "issue:45200|issue:46932", - "issue:44442|issue:45198" + "issue:45200|issue:46932" ], - "prepared_review_unit_hash": "586425a6babb803e6da27e511048ae77dd3ee353be1b75b1ce5294558fb4ea45", + "prepared_review_unit_hash": "0cb1483f57cfb619d72fc8aba8810be93426ffe973c4406d24d48402233febc4", "component_budget": { "node_count": 505, "item_count": 505, "soft_pair_count": 1854, - "serialized_chars": 962354, - "estimated_input_tokens": 240589, - "estimated_eval_tokens": 481434 + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 }, "budget": { "node_count": 47, "item_count": 47, "soft_pair_count": 26, - "serialized_chars": 35358, - "estimated_input_tokens": 8840, - "estimated_eval_tokens": 17936 + "serialized_chars": 35372, + "estimated_input_tokens": 8843, + "estimated_eval_tokens": 17942 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "Only one soft pair looks like a true duplicate: the Gemma4 4-bit multimodal embedder dtype-cast regression. The rest are unrelated failures that happen to share broad subsystems (tokenizers, chat templates, imports, model loading) but do not describe the same concrete bug.", + "summary": "All soft-edge candidates look like false positives: they span unrelated tokenizer, config, import, generation, multimodal, and training bugs, not the same underlying change or defect.", "confidence": 0.91, - "canonical_issue_reason": "Use issue 46899 as the canonical issue: it is the clearest original report of the Gemma4 4-bit quantization failure in multimodal embedders, while 46932 is the same bug framed as a fix/duplication reference.", + "canonical_issue_reason": "issue:44521 is the most active/open report in the set with strong inbound discussion and a concrete, reproducible bug description, so it is the best single anchor if one issue must represent this mixed cluster.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 46899 is the best representative because it describes the concrete failing behavior directly and matches 46932\u2019s same dtype-casting bug.", + "best_issue_reason": "issue:44521 has the clearest ongoing signal and the broadest direct follow-up potential among these unrelated issues, making it the most suitable representative issue.", "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:43550", + "right": "issue:43844", + "accept": false, + "reason": "Different bugs: SDPA/torch.compile failure in Bamba-9B-v2 vs ZeRO-3 gradient instability in DeepSpeed." + }, { "left": "issue:29870", "right": "issue:29942", "accept": false, - "reason": "Different failures: Bart generate() decoder_attention_mask bug vs Flash Attention 2 test regressions." + "reason": "Bart generate() UnboundLocalError on decoder_attention_mask is unrelated to Flash Attention 2 test failures." }, { "left": "issue:43450", "right": "issue:43575", "accept": false, - "reason": "Unrelated problems: batched video processor output shape vs Qwen2-57B tensor-parallel OOM." + "reason": "Video processor batched-shape bug vs Qwen2-57B tensor-parallel OOM; different components and symptoms." }, { "left": "issue:45520", "right": "issue:46178", "accept": false, - "reason": "Both are import/backend issues, but they target different modules and errors; not the same bug." + "reason": "Python 3.13 flash_attn KeyError is not the same as a BACKENDS_MAPPING tensorflow_text ValueError." }, { "left": "issue:43572", "right": "issue:44987", "accept": false, - "reason": "Different regressions: missing pad_token_idx in StableLmConfig vs loading a specific model family." + "reason": "Missing pad_token_idx in StableLmConfig is distinct from a model-loading failure on transformers>=5.1.0." }, { "left": "issue:37428", "right": "issue:43824", "accept": false, - "reason": "Both are ImportErrors, but for different symbols and code paths." + "reason": "Both are ImportErrors, but for different missing symbols/classes in different modules." }, { "left": "issue:42490", "right": "issue:43493", "accept": false, - "reason": "Path-with-slash import warning is unrelated to the SigLIP2 implementation discrepancy." + "reason": "Trailing-slash local-path import warning is unrelated to the SigLIP2 HF-vs-JAX discrepancy." }, { "left": "issue:39692", "right": "issue:46192", "accept": false, - "reason": "Doc/example errors for SigLIP2 are not the same as return_tensors being ignored with text_kwargs." + "reason": "Docs example/model-processor mismatch is not the same bug as return_tensors being ignored with text_kwargs." }, { "left": "issue:44554", "right": "issue:45356", "accept": false, - "reason": "Different issues: MPS attention correctness vs Kimi tokenizer regression." + "reason": "MPS attention correctness issue and Kimi-K2.5 tokenizer codec regression affect different code paths." }, { "left": "issue:43232", "right": "issue:44849", "accept": false, - "reason": "Separate generation bugs affecting different behaviors and model families." + "reason": "Generation kwarg sync_gpus handling is unrelated to the Qwen3.5 output_hidden_states bug." }, { "left": "issue:43295", "right": "issue:43493", "accept": false, - "reason": "Custom processor/tokenizer regression is not the same as the SigLIP2 JAX-vs-HF discrepancy." + "reason": "processor.tokenizer regression for custom model code is unrelated to the SigLIP2 implementation discrepancy." }, { "left": "issue:44521", "right": "issue:45290", "accept": false, - "reason": "Both involve apply_chat_template, but one is all-zero assistant masks and the other is a crash on tool-call messages; different bugs." + "reason": "Both involve apply_chat_template, but one is all-zero assistant masks for multimodal input and the other is a tool-call crash; different failure modes." }, { "left": "issue:44466", "right": "issue:45127", "accept": false, - "reason": "Both touch tied weights/lm_head, but one is serialization inconsistency and the other is LoRA merge collapse." + "reason": "Serialization of lm_head.weight by device is not the same as LoRA merge collapse with extended vocabulary." }, { "left": "issue:45381", "right": "issue:46209", "accept": false, - "reason": "Different problems: Qwen2.5-VL video position IDs vs FP8 inference on B200." + "reason": "Qwen2.5-VL video position ids bug is unrelated to FP8 inference failure on B200." }, { "left": "issue:28218", "right": "issue:35141", "accept": false, - "reason": "Tokenizer spacing after added tokens is unrelated to post_init reinitializing resized embeddings." + "reason": "Tokenizer extra-space insertion and embedding reinitialization in post_init are separate issues." }, { "left": "issue:44737", "right": "issue:46894", "accept": false, - "reason": "CPU arange device bug in XLNet is unrelated to CohereASR training-loss shifting." + "reason": "XLNet CPU placement bug is unrelated to the CohereASR training-loss double-shift problem." }, { "left": "issue:43278", "right": "issue:43856", "accept": false, - "reason": "Embedding dtype mismatch in evaluation is unrelated to MoE training memory inefficiency." + "reason": "BF16-to-FP32 dtype drift in evaluate is not the same as MoE training memory inefficiency." }, { "left": "issue:42915", "right": "issue:43824", "accept": false, - "reason": "Qwen3Moe FP8 config failure and Qwen2.5-VL import error are different code paths." + "reason": "Qwen3Moe FineGrainedFP8Config failure and missing Qwen2_5_VLForConditionalGeneration import are different defects." }, { "left": "issue:43824", "right": "issue:43975", "accept": false, - "reason": "Importing a missing class is not the same as incorrect detokenization behavior." + "reason": "A missing exported class ImportError is unrelated to incorrect detokenization in v5." }, { "left": "issue:43653", "right": "issue:44112", "accept": false, - "reason": "BigBirdTokenizer special-token decode bug is unrelated to a GraniteSpeech CI test failure." + "reason": "BigBirdTokenizer special-token registration bug is unrelated to a stale GraniteSpeech CI test." }, { - "left": "issue:43950", - "right": "issue:46285", + "left": "issue:43452", + "right": "issue:46512", "accept": false, - "reason": "Non-persistent buffer corruption on from_pretrained is a different regression from CLIP text-model structure changes." + "reason": "gguf_file loading failures and from_config(dtype=...) leaving config.dtype stale are different lifecycle problems." }, { - "left": "issue:43452", - "right": "issue:46512", + "left": "issue:43950", + "right": "issue:46285", "accept": false, - "reason": "gguf_file loading failures are unrelated to stale config dtype after from_config." + "reason": "Non-persistent buffer corruption in from_pretrained is not the same as CLIP text-model structural inconsistency." }, { "left": "issue:43262", "right": "issue:43450", "accept": false, - "reason": "Audio chat-template sampling-rate default and batched video shape are different processor bugs." + "reason": "Audio chat_template sampling-rate default bug is unrelated to batched video processor shape output." }, { "left": "issue:43334", "right": "issue:43525", "accept": false, - "reason": "Both mention missing pad_token_id, but they affect different model configs and load paths; not clearly the same bug." + "reason": "Both are pad_token_id AttributeErrors, but on different model/config classes; insufficient evidence of one root cause." }, { "left": "issue:45200", "right": "issue:46899", "accept": false, - "reason": "Gemma4 text-only mm_token_type_ids defaulting is a different issue from quantized audio/vision embedder dtype casting." + "reason": "Gemma4 token-type-id defaults and 4-bit quantized embedder dtype casting are separate multimodal bugs." }, { "left": "issue:45200", "right": "issue:46932", "accept": false, - "reason": "Different Gemma4 regressions: missing mm_token_type_ids defaults vs quantized embedder dtype casting." - }, - { - "left": "issue:44442", - "right": "issue:45198", - "accept": false, - "reason": "Tokenizer loading failure for FastSpeech2Conformer is unrelated to Wav2Vec2 save_pretrained/tokenization failure." - }, - { - "left": "issue:46899", - "right": "issue:46932", - "accept": true, - "reason": "Same Gemma4 multimodal quantization bug: both describe unsafe dtype casting in audio/vision embedders causing broken processing." + "reason": "Same family, but token-type-id defaults and quantized vision/audio embedder casting do not look like the same underlying fix." } ] }, @@ -9296,7 +9284,7 @@ { "label": "issue", "component_index": 1, - "component_count": 25, + "component_count": 26, "review_unit_index": 30, "review_unit_count": 61, "status": "reviewed", @@ -9327,15 +9315,16 @@ "issue:43901", "issue:44112", "issue:44206", + "issue:44442", "issue:44466", "issue:44479", "issue:44561", "issue:44610", "issue:44792", - "issue:44855", "issue:44857", "issue:44898", "issue:45083", + "issue:45198", "issue:45200", "issue:45325", "issue:45412", @@ -9355,6 +9344,7 @@ "issue:46858" ], "soft_pairs": [ + "issue:44442|issue:45198", "issue:44792|issue:45563", "issue:44898|issue:45412", "issue:44479|issue:46192", @@ -9367,219 +9357,218 @@ "issue:44610|issue:44857", "issue:43819|issue:45083", "issue:43388|issue:43526", - "issue:45759|issue:46285", "issue:46512|issue:46821", + "issue:45759|issue:46285", "issue:44857|issue:45593", "issue:44206|issue:45584", "issue:43606|issue:45083", - "issue:46424|issue:46858", "issue:38175|issue:45412", + "issue:46424|issue:46858", "issue:43335|issue:43653", "issue:37428|issue:43425", "issue:28282|issue:33357", "issue:29942|issue:32090", "issue:43550|issue:44112", "issue:43611|issue:44466", - "issue:45520|issue:46170", "issue:43335|issue:43504", - "issue:44855|issue:45083" + "issue:45520|issue:46170" ], - "prepared_review_unit_hash": "8050397099be16ef9dbe67fb73176f3eb6e590ebd7e191ff668caa629d3a93a8", + "prepared_review_unit_hash": "8c6e550b0e7f2b07f48848c547f9ebe8bccbf49c9fa071e2b1c9b364b1c4de14", "component_budget": { "node_count": 505, "item_count": 505, "soft_pair_count": 1854, - "serialized_chars": 962354, - "estimated_input_tokens": 240589, - "estimated_eval_tokens": 481434 + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 }, "budget": { - "node_count": 47, - "item_count": 47, + "node_count": 48, + "item_count": 48, "soft_pair_count": 28, - "serialized_chars": 35913, - "estimated_input_tokens": 8979, - "estimated_eval_tokens": 18214 + "serialized_chars": 36413, + "estimated_input_tokens": 9104, + "estimated_eval_tokens": 18464 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The cluster is heterogeneous: the soft pairs share only broad subsystem similarity (models, processors, cache, docs, import_utils) but not the same concrete bug. None of the candidate pairs should be merged as duplicates.", - "confidence": 0.91, - "canonical_issue_reason": "No single issue is a strong canonical for this cluster; the items span unrelated regressions and API/doc mismatches, and none of the soft pairs are the same underlying bug.", + "summary": "The pairwise similarities are mostly superficial; these issues describe different models, different subsystems, or different failure modes, so none of the soft edges should be collapsed.", + "confidence": 0.95, + "canonical_issue_reason": "issue:44479 is the clearest self-contained regression report in the set, with explicit affected models and a concrete user-facing failure.", "canonical_pr_reason": null, - "best_issue_reason": "No issue is particularly representative of the whole set; the cluster is too mixed to justify a global best duplicate target.", + "best_issue_reason": "issue:44479 is the best representative issue because it is specific, reproducible, and describes a broad regression affecting multiple Qwen VL variants.", "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:44442", + "right": "issue:45198", + "accept": false, + "reason": "Different tokenizer families and failure modes: AutoTokenizer loading FastSpeech2ConformerTokenizer vs Wav2Vec2 save_pretrained/tokenization behavior." + }, { "left": "issue:44792", "right": "issue:45563", "accept": false, - "reason": "Different bugs: Janus image generation test failure vs a stale paging warning for num_return_sequences." + "reason": "Both mention generation-related tests, but one is a janus image generation failure and the other is a stale warning about num_return_sequences." }, { "left": "issue:44898", "right": "issue:45412", "accept": false, - "reason": "Different code paths: Perceiver resolution/interpolation failure vs RT-DETR memory release behavior." + "reason": "Unrelated bugs: Perceiver image classification resolution handling vs RT-DETR memory not being released." }, { "left": "issue:44479", "right": "issue:46192", "accept": false, - "reason": "Unrelated regressions: video-input handling in Qwen VL models vs return_tensors being ignored in text_kwargs." + "reason": "Both touch multimodal inputs, but one is a video-input regression in Qwen VL models while the other is text_kwargs/return_tensors precedence." }, { "left": "issue:45200", "right": "issue:45325", "accept": false, - "reason": "Both involve multimodal inputs, but the concrete bugs differ: Gemma mm_token_type_ids defaults vs Qwen2.5-VL rope position scaling." + "reason": "Different models and code paths: Gemma 4 text-only mm_token_type_ids defaulting vs Qwen2.5-VL rope index scaling for still images." }, { "left": "issue:43819", "right": "issue:44112", "accept": false, - "reason": "DAC latent reconstruction mismatch and GraniteSpeech stale device-override CI failure are unrelated." + "reason": "Different components entirely: DAC latent conversion mismatch vs a GraniteSpeech CI device-override test failure." }, { "left": "issue:41720", "right": "issue:46209", "accept": false, - "reason": "Both are GPU-related, but one is a Qwen3 auto device-mapping assert and the other is FP8 inference failure on B200." + "reason": "Both are CUDA/runtime issues, but they affect different models, hardware, and root causes." }, { "left": "issue:43335", "right": "issue:43550", "accept": false, - "reason": "Different models and failures: SwitchTransformers sparse-layer config bug vs Bamba torch.compile/SDPA failure." + "reason": "SwitchTransformers sparse-layer config bug is unrelated to Bamba torch.compile/SDPA failure." }, { "left": "issue:44561", "right": "issue:46234", "accept": false, - "reason": "Both touch import_utils, but they concern different missing helpers and different affected models." + "reason": "Both involve import_utils helpers, but they are missing different symbols and stem from different compatibility breaks." }, { "left": "issue:43827", "right": "issue:43901", "accept": false, - "reason": "Both are docs issues around pipelines, but one is about removed pipeline() references and the other about return_all_scores behavior." + "reason": "Both are docs issues, but they concern different docs pages and different API behaviors." }, { "left": "issue:44610", "right": "issue:44857", "accept": false, - "reason": "Different failures in detection models: processor image-size mismatch vs float16 AMP crash in the loss." + "reason": "Different vision models and failures: incorrect processor image size vs mixed-precision loss crash." }, { "left": "issue:43819", "right": "issue:45083", "accept": false, - "reason": "DAC from_latents mismatch is unrelated to qwen3_omni_moe feature-length helper behavior." + "reason": "No shared bug beyond being in the same general audio/multimodal area; DAC latent mismatch is unrelated to qwen3_omni_moe feature-length arithmetic." }, { "left": "issue:43388", "right": "issue:43526", "accept": false, - "reason": "Both concern labels, but one is gather_for_metrics dropping tuple labels and the other is BeitImageProcessorFast reduce_labels output." + "reason": "Both mention labels, but one is a distributed metrics batching bug and the other is an image processor reduce_labels bug." }, { - "left": "issue:45759", - "right": "issue:46285", + "left": "issue:46512", + "right": "issue:46821", "accept": false, - "reason": "Different model-structure regressions: Qwen multimodal text_config unwrap vs CLIP text model class structure change." + "reason": "Completely different issues: stale config.dtype after from_config vs TimesFM window_size AttributeError." }, { - "left": "issue:46512", - "right": "issue:46821", + "left": "issue:45759", + "right": "issue:46285", "accept": false, - "reason": "Stale config.dtype after from_config(dtype=...) is unrelated to TimesFM window_size AttributeError." + "reason": "Both involve model/config structure, but the concrete problems and affected architectures differ." }, { "left": "issue:44857", "right": "issue:45593", "accept": false, - "reason": "Different object-detection bugs: AMP crash in LwDetrImageLoss vs missing auxiliary losses when denoising is off." + "reason": "Different detection-model training bugs: AMP/CUDA crash versus auxiliary losses being skipped when denoising is disabled." }, { "left": "issue:44206", "right": "issue:45584", "accept": false, - "reason": "Unrelated issues: LasrFeatureExtractor center-arg crash vs Whisper empty transcription after align_special_tokens." + "reason": "Different audio pipeline failures affecting different models and different trigger conditions." }, { "left": "issue:43606", "right": "issue:45083", "accept": false, - "reason": "CPU offload device mismatch in bark-small is unrelated to qwen3_omni_moe feature-length calculation." + "reason": "CPU offload device mismatch for bark-small is unrelated to qwen3_omni_moe feature-length helper behavior." }, { - "left": "issue:46424", - "right": "issue:46858", + "left": "issue:38175", + "right": "issue:45412", "accept": false, - "reason": "Both are cache-related, but one is about max_cache_length semantics and the other about get_seq_length returning a device tensor." + "reason": "Different symptoms and subsystems: zero probabilities in SigLIP2 vs RT-DETR memory leak." }, { - "left": "issue:38175", - "right": "issue:45412", + "left": "issue:46424", + "right": "issue:46858", "accept": false, - "reason": "Different models and symptoms: SIGLIP zero probabilities vs RT-DETR memory leak." + "reason": "Both are cache-related, but one is about max_cache_length semantics and the other is a device-tensor return breaking static-cache slice arithmetic." }, { "left": "issue:43335", "right": "issue:43653", "accept": false, - "reason": "SwitchTransformers sparse-layer construction and BigBirdTokenizer special-token registration are unrelated." + "reason": "Distinct bugs in different subsystems: SwitchTransformers sparse-layer construction vs BigBird tokenizer special-token registration." }, { "left": "issue:37428", "right": "issue:43425", "accept": false, - "reason": "Both mention torch compatibility, but one is a missing import in flash_attention_utils and the other is a broad Torch 2.10 incompatibility report." + "reason": "Missing import symbol in flash_attention_utils is not the same bug as Torch 2.10 incompatibility." }, { "left": "issue:28282", "right": "issue:33357", "accept": false, - "reason": "ImportError due to missing PyTorch and a MacOS bus error with a community CLIP model are different problems." + "reason": "One is a missing PyTorch ImportError, the other is a MacOS bus error in a specific CLIP model." }, { "left": "issue:29942", "right": "issue:32090", "accept": false, - "reason": "Flash Attention 2 test failures and Trainer _gpu_broadcast_one NoneType errors are unrelated." + "reason": "Unrelated failures: Flash Attention 2 test regressions vs a Trainer broadcast TypeError." }, { "left": "issue:43550", "right": "issue:44112", "accept": false, - "reason": "torch.compile/SDPA failure in Bamba is not the same as the GraniteSpeech stale device override test." + "reason": "Different causes and code paths: Bamba torch.compile/SDPA failure vs GraniteSpeech stale device-override CI test." }, { "left": "issue:43611", "right": "issue:44466", "accept": false, - "reason": "Both are v5 regressions, but one is model loading with base_model_prefix and the other is lm_head.weight serialization inconsistency." - }, - { - "left": "issue:45520", - "right": "issue:46170", - "accept": false, - "reason": "Different import/runtime issues: missing flash_attn key on Python 3.13 vs Windows ROCm torch.distributed dependency problems." + "reason": "Both relate to model loading/serialization, but one is about base_model_prefix compatibility and the other about lm_head.weight serialization differing by device." }, { "left": "issue:43335", "right": "issue:43504", "accept": false, - "reason": "Different model bugs: SwitchTransformers sparse-layer creation vs Beit semantic-segmentation legacy-field loading failure." + "reason": "Different models and bugs: sparse-layer creation in SwitchTransformers vs legacy-field loading failure in BeitForSemanticSegmentation." }, { - "left": "issue:44855", - "right": "issue:45083", + "left": "issue:45520", + "right": "issue:46170", "accept": false, - "reason": "DebertaV2 Python 3.13 import parsing error is unrelated to qwen3_omni_moe feature-length helper behavior." + "reason": "Separate environment-specific import problems: flash_attn key lookup on Python 3.13 vs Windows ROCm/torch.distributed dependency issues." } ] }, @@ -9593,7 +9582,7 @@ { "label": "issue", "component_index": 1, - "component_count": 25, + "component_count": 26, "review_unit_index": 31, "review_unit_count": 61, "status": "reviewed", @@ -9608,7 +9597,6 @@ "issue:29127", "issue:31501", "issue:32090", - "issue:33357", "issue:34689", "issue:35141", "issue:37428", @@ -9633,6 +9621,7 @@ "issue:44617", "issue:44743", "issue:44779", + "issue:44855", "issue:44857", "issue:45072", "issue:45083", @@ -9653,244 +9642,244 @@ "issue:46697" ], "soft_pairs": [ + "issue:44855|issue:45083", "issue:44448|issue:44743", - "issue:43278|issue:43825", "issue:37428|issue:44351", - "issue:43335|issue:45083", "issue:44514|issue:44521", + "issue:43335|issue:45083", "issue:45290|issue:46326", - "issue:43335|issue:44112", "issue:43335|issue:45593", + "issue:43335|issue:44112", "issue:29127|issue:32090", + "issue:43278|issue:43825", "issue:45588|issue:45859", "issue:42490|issue:43854", "issue:34689|issue:35141", "issue:43720|issue:44112", "issue:43819|issue:44493", - "issue:46431|issue:46566", "issue:39692|issue:43526", + "issue:46431|issue:46566", "issue:43606|issue:44857", "issue:45310|issue:45478", "issue:44448|issue:44779", "issue:28218|issue:32090", "issue:39692|issue:44493", - "issue:26413|issue:31501", + "issue:44617|issue:46697", "issue:45459|issue:46234", + "issue:26413|issue:31501", "issue:45127|issue:45245", - "issue:44617|issue:46697", "issue:44206|issue:45823", "issue:43701|issue:45083", - "issue:32090|issue:33357", "issue:44857|issue:45072" ], - "prepared_review_unit_hash": "9c723c71280ec24d5c777c97e1cf65a97617988ee5c37f5ec5fb5b075b6b5d15", + "prepared_review_unit_hash": "55ee729ed80bf1d4f4f4da54c7464eb1329f5e7d541505f1081585e731908f10", "component_budget": { "node_count": 505, "item_count": 505, "soft_pair_count": 1854, - "serialized_chars": 962354, - "estimated_input_tokens": 240589, - "estimated_eval_tokens": 481434 + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 29, - "serialized_chars": 36542, - "estimated_input_tokens": 9136, - "estimated_eval_tokens": 18528 + "serialized_chars": 36608, + "estimated_input_tokens": 9152, + "estimated_eval_tokens": 18560 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly a set of unrelated issue reports with only one clear duplicate pair: the two identical Qwen3.5 MoE `from_pretrained` regression reports. The rest differ by model, component, or failure mode enough to keep separate.", - "confidence": 0.97, - "canonical_issue_reason": "Issue 45310 is the better canonical report: it is the earlier and more established report of the same `transformers>=5.4.0` Qwen3.5 MoE `from_pretrained` failure, with the duplicate later issue 45478 matching it closely.", + "summary": "This cluster is mostly a loose thematic grouping of unrelated bug reports. The only clear duplicate pair is the Qwen3.5 MoE `from_pretrained` regression (45310/45478); the rest are different code paths, models, or symptom classes.", + "confidence": 0.93, + "canonical_issue_reason": "Issue 45310 is the best canonical issue: it is the earlier report of the exact same Qwen3.5 MoE `from_pretrained` failure later repeated in 45478.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 45310 is the strongest representative of the only true duplicate pair in this set; it has the same concrete failure, earlier timestamp, and enough context to serve as the primary tracker.", + "best_issue_reason": "45310 is the strongest representative of the cluster because it has the clearest true-duplicate match and the most stable, specific problem statement.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:44448", - "right": "issue:44743", + "left": "issue:44855", + "right": "issue:45083", "accept": false, - "reason": "Different bugs: Pegasus output drift vs Qwen recurrent-state reset; no shared concrete code-path." + "reason": "Both are bugs, but they affect unrelated modules and failures: DeBERTa import parsing vs a qwen3_omni_moe helper behavior issue." }, { - "left": "issue:43278", - "right": "issue:43825", + "left": "issue:44448", + "right": "issue:44743", "accept": false, - "reason": "Unrelated symptoms: embedding dtype change in eval vs a pipeline error message for unsupported translation tasks." + "reason": "Different models and different problems: Pegasus output drift across versions vs Qwen3.5 recurrent-state reset with cache." }, { "left": "issue:37428", "right": "issue:44351", "accept": false, - "reason": "Both are import-related, but they concern different missing symbols in different modules." + "reason": "Both are import errors, but for different missing symbols in different subsystems; not the same underlying bug." }, { - "left": "issue:43335", - "right": "issue:45083", + "left": "issue:44514", + "right": "issue:44521", "accept": false, - "reason": "Different model families and code paths: SwitchTransformers sparse-layer construction vs qwen3_omni_moe feature length helper." + "reason": "Both involve `apply_chat_template`, but one is a batched-padding crash and the other is incorrect assistant masks for multimodal inputs." }, { - "left": "issue:44514", - "right": "issue:44521", + "left": "issue:43335", + "right": "issue:45083", "accept": false, - "reason": "Both touch `apply_chat_template`, but one crashes on batched padding=False input while the other produces zero assistant masks; not the same bug." + "reason": "Unrelated issues: SwitchTransformers config bug vs qwen3_omni_moe feature-extraction behavior." }, { "left": "issue:45290", "right": "issue:46326", "accept": false, - "reason": "Both involve chat templating, but one is a tool-call/content crash and the other is a `mistral_common` final-message handling bug." + "reason": "Both touch chat templating/tokenization, but the failing paths differ: tool-call handling vs `continue_final_message` on mistral_common tokenizers." }, { "left": "issue:43335", - "right": "issue:44112", + "right": "issue:45593", "accept": false, - "reason": "Config construction bug vs stale CI test failure; different artifacts and failure modes." + "reason": "Different underlying bugs in different models: SwitchTransformers config creation vs D-FINE auxiliary-loss behavior." }, { "left": "issue:43335", - "right": "issue:45593", + "right": "issue:44112", "accept": false, - "reason": "Sparse-layer initialization and missing auxiliary losses are unrelated model behaviors." + "reason": "No shared code path; one is a model config bug and the other is a CI failure about stale device overrides." }, { "left": "issue:29127", "right": "issue:32090", "accept": false, - "reason": "LayoutLMv3 error-message clarity vs Trainer `_gpu_broadcast_one` NoneType crash are unrelated." + "reason": "Completely different failures: LayoutLMv3 error-message clarity vs Trainer `_gpu_broadcast_one` type handling." + }, + { + "left": "issue:43278", + "right": "issue:43825", + "accept": false, + "reason": "Different symptoms and code paths: dtype drift in evaluation vs a pipeline task-support error message." }, { "left": "issue:45588", "right": "issue:45859", "accept": false, - "reason": "Different offload/device issues in different modules; one is flash-attention auxiliary state, the other is Qwen rotary embedding CPU offload." + "reason": "Both involve offload/attention-related code, but one is `s_aux=None` handling in flash attention and the other is a CPU-offload incompatibility in rotary embeddings." }, { "left": "issue:42490", "right": "issue:43854", "accept": false, - "reason": "Local-path import warning vs a specific GLM-4.7-Flash loading failure in tests; not the same underlying bug." + "reason": "A local-path import warning is unrelated to a unit-test model-loading failure for GLM-4.7-Flash." }, { "left": "issue:34689", "right": "issue:35141", "accept": false, - "reason": "Model-loading regression vs embedding reinitialization after resizing; distinct code paths." + "reason": "Different regressions: Llama 3.2 Vision model loading vs output embedding reinitialization after resizing token embeddings." }, { "left": "issue:43720", "right": "issue:44112", "accept": false, - "reason": "Packed-weight handling during accelerate loading vs a stale GraniteSpeech device-override test failure." + "reason": "Unrelated: packed-weight loading failure in BitNet vs a stale CI device-override test." }, { "left": "issue:43819", "right": "issue:44493", "accept": false, - "reason": "DAC latent reconstruction mismatch and unexpected position-id keys are different model issues." + "reason": "Different bug classes: DAC latent/forward mismatch vs unexpected position-id keys in many models." }, { - "left": "issue:46431", - "right": "issue:46566", + "left": "issue:39692", + "right": "issue:43526", "accept": false, - "reason": "Missing chat template error vs DiffusionGemma staying on meta device; unrelated." + "reason": "Docs example errors for SigLIP2 are not the same as the BEiT Fast processor label-reduction bug." }, { - "left": "issue:39692", - "right": "issue:43526", + "left": "issue:46431", + "right": "issue:46566", "accept": false, - "reason": "Documentation example errors vs BEiT fast processor label reduction bug; different scopes." + "reason": "Different problems: missing chat template on a processor vs DiffusionGemma remaining on meta with `device_map=\"auto\"`." }, { "left": "issue:43606", "right": "issue:44857", "accept": false, - "reason": "CPU offload device mismatch vs CUDA float16 loss crash; not the same failure." + "reason": "Both are runtime failures, but one is device mismatch with CPU offload and the other is an AMP/CUDA crash in a loss function." }, { "left": "issue:45310", "right": "issue:45478", "accept": true, - "reason": "Same Qwen3.5 MoE `from_pretrained` regression on `transformers>=5.4.0`, with matching title and closely aligned report details." + "reason": "Exact same title and same Qwen3.5 MoE `from_pretrained` regression; these are clear duplicates of the same bug." }, { "left": "issue:44448", "right": "issue:44779", "accept": false, - "reason": "Pegasus output differences and Deepseek tokenizer regression are separate model behaviors." + "reason": "Different models and failure modes: Pegasus output mismatch vs Deepseek tokenizer regression." }, { "left": "issue:28218", "right": "issue:32090", "accept": false, - "reason": "Tokenizer added-space bug vs Trainer broadcast NoneType crash are unrelated." + "reason": "Tokenizer spacing after added tokens is unrelated to a Trainer broadcast type error." }, { "left": "issue:39692", "right": "issue:44493", "accept": false, - "reason": "SigLIP2 docs/example failures and position-id key warnings are different problems." + "reason": "Documentation/example issues are unrelated to unexpected position-id keys in model loading." }, { - "left": "issue:26413", - "right": "issue:31501", + "left": "issue:44617", + "right": "issue:46697", "accept": false, - "reason": "Checkpoint resumption failure and FineWeb training startup failure are not the same bug." + "reason": "CUDA OOM in a video model is not the same bug as over-padding images in Swin2SR." }, { "left": "issue:45459", "right": "issue:46234", "accept": false, - "reason": "Tokenizer protobuf error masking vs missing import utility symbol; different issues." + "reason": "Different import-related issues: tokenizer error masking vs a missing utility symbol." }, { - "left": "issue:45127", - "right": "issue:45245", + "left": "issue:26413", + "right": "issue:31501", "accept": false, - "reason": "LoRA merge collapse with extended vocab vs category-count limit runtime error are unrelated." + "reason": "Both touch training flow, but one is checkpoint resume epoch-iterator failure and the other is a training job that never starts." }, { - "left": "issue:44617", - "right": "issue:46697", + "left": "issue:45127", + "right": "issue:45245", "accept": false, - "reason": "CUDA OOM in Sam3Video vs Swin2SR over-padding; not a duplicate." + "reason": "Unrelated: LoRA merge collapse with extended vocab vs a category-count limit error." }, { "left": "issue:44206", "right": "issue:45823", "accept": false, - "reason": "Unsupported `center` arg in a feature extractor vs Gemma4 PLE device mismatch; different code paths." - }, - { - "left": "issue:43701", - "right": "issue:45083", - "accept": false, - "reason": "Checkpoint key mismatch and qwen3_omni_moe feature-length helper behavior are unrelated." + "reason": "Different model-specific failures: unsupported feature-extractor arg vs device-mapped forward mismatch." }, { - "left": "issue:32090", - "right": "issue:33357", + "left": "issue:44857", + "right": "issue:45072", "accept": false, - "reason": "NoneType GPU broadcast error vs MacOS bus error loading a CLIP model; different failures." + "reason": "Both mention dtype issues, but they are different models and distinct failure paths." }, { - "left": "issue:44857", - "right": "issue:45072", + "left": "issue:43701", + "right": "issue:45083", "accept": false, - "reason": "Both mention dtype, but one crashes in loss computation under AMP while the other is a bfloat16 inference mismatch; not the same underlying bug." + "reason": "Checkpoint key mismatch is unrelated to a qwen3_omni_moe helper behavior bug." } ] }, "evaluator_result": { "accept": true, - "feedback": "The summary is grounded in the packet: the only clearly duplicate pair is the identical Qwen3.5 MoE `from_pretrained` reports (45310/45478), and the remaining soft-edge rejections are appropriately conservative given the differing titles and failure modes. No overstatement stands out." + "feedback": "The summary is well grounded in the packet and stays conservative on the soft-edge pairs. The only accepted duplicate, issue:45310 / issue:45478, is clearly supported by the identical title and matching Qwen3.5 MoE from_pretrained failure. The rest of the pair judgments are appropriately cautious and do not overstate duplicate evidence." }, "evaluator_used": true, "retried": false, @@ -9901,7 +9890,7 @@ { "label": "issue", "component_index": 1, - "component_count": 25, + "component_count": 26, "review_unit_index": 32, "review_unit_count": 61, "status": "reviewed", @@ -9919,6 +9908,7 @@ "issue:29942", "issue:32090", "issue:33290", + "issue:33357", "issue:34689", "issue:35141", "issue:37428", @@ -9941,7 +9931,6 @@ "issue:44493", "issue:44561", "issue:44568", - "issue:44779", "issue:44855", "issue:44977", "issue:45083", @@ -9961,24 +9950,25 @@ "issue:47000" ], "soft_pairs": [ + "issue:32090|issue:33357", "issue:42490|issue:43575", "issue:45593|issue:47000", "issue:43550|issue:45083", - "issue:38175|issue:44493", "issue:27957|issue:34689", "issue:29942|issue:35141", + "issue:38175|issue:44493", "issue:29870|issue:32090", "issue:43278|issue:43329", - "issue:44360|issue:44561", "issue:45290|issue:45381", "issue:45920|issue:46192", "issue:44448|issue:44977", "issue:10000|issue:15354", + "issue:44360|issue:44561", "issue:44977|issue:45229", "issue:44855|issue:46399", "issue:29870|issue:33290", - "issue:43927|issue:45698", "issue:45290|issue:46655", + "issue:43927|issue:45698", "issue:40444|issue:43526", "issue:43278|issue:43381", "issue:37428|issue:42175", @@ -9987,219 +9977,218 @@ "issue:43873|issue:45229", "issue:42490|issue:43873", "issue:25251|issue:32090", - "issue:44561|issue:44779", "issue:43976|issue:45440", "issue:44448|issue:44466", "issue:44568|issue:45491" ], - "prepared_review_unit_hash": "90600116dffdf1e63ac0d2f7dd9c9f0f13fc0ab84aa4fa457e32815d2a9ea0a2", + "prepared_review_unit_hash": "f6d2ae50e3f343379a60104f959718e73a1ea529e1002fa09715f44d75961210", "component_budget": { "node_count": 505, "item_count": 505, "soft_pair_count": 1854, - "serialized_chars": 962354, - "estimated_input_tokens": 240589, - "estimated_eval_tokens": 481434 + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 30, - "serialized_chars": 36984, - "estimated_input_tokens": 9246, - "estimated_eval_tokens": 18748 + "serialized_chars": 36980, + "estimated_input_tokens": 9245, + "estimated_eval_tokens": 18746 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This set is a collection of mostly unrelated Transformers issues spanning tokenizers, generation, attention backends, training, and model loading. None of the proposed soft pairs look like the same underlying bug or change, so I would not merge any of them as duplicates. The broadest representative issue is the v5 tokenizer regression affecting multiple models.", - "confidence": 0.93, - "canonical_issue_reason": "Issue 45920 is the most representative because it describes a broad v5 tokenizer regression with multiple affected models and broader ecosystem impact than the narrower single-model reports.", + "summary": "This set does not form a single duplicate cluster; it\u2019s a collection of unrelated transformer bugs/regressions across different models and subsystems. None of the proposed soft edges look like the same underlying issue, so all are rejected.", + "confidence": 0.24, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "Issue 45920 is the strongest overall representative: broad scope, multiple affected models, and clear regression framing make it the best summary of the cluster's dominant theme.", + "best_issue_reason": "issue:45920 is the strongest representative because it describes a broad, concrete v5 regression affecting multiple model families with clear user impact.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:42490", - "right": "issue:43575", + "left": "issue:32090", + "right": "issue:33357", + "accept": false, + "reason": "Different failures and stacks: Trainer GPU broadcast NoneType vs MacOS bus error in CLIP." + }, + { + "left": "issue:42490", + "right": "issue:43575", "accept": false, - "reason": "Different problems: local model path/import warning vs tensor-parallel OOM while loading a large Qwen model." + "reason": "Import warning for local paths is unrelated to Qwen2.5 tensor-parallel OOM on model load." }, { "left": "issue:45593", "right": "issue:47000", "accept": false, - "reason": "Different detection/training bugs: missing auxiliary losses in D-FINE vs RTDetrLoss infeasible cost matrix." + "reason": "Different detection bugs: missing auxiliary losses vs infeasible RTDetr cost matrix." }, { "left": "issue:43550", "right": "issue:45083", "accept": false, - "reason": "Unrelated areas: torch.compile/SDPA failure in Bamba vs a feature-length helper bug in qwen3_omni_moe." - }, - { - "left": "issue:38175", - "right": "issue:44493", - "accept": false, - "reason": "Different symptoms and code paths: zero probabilities in SigLIP2 vs unexpected position-id keys across many models." + "reason": "torch.compile+SDPA crash in Bamba is not the same bug as qwen3_omni_moe feature-extraction helper behavior." }, { "left": "issue:27957", "right": "issue:34689", "accept": false, - "reason": "Flash Attention 2 support issue for XLM-RoBERTa vs a model-loading breakage for Llama 3.2 Vision." + "reason": "Flash Attention 2 support for XLMRoberta is unrelated to Llama 3.2 Vision model loading breakage." }, { "left": "issue:29942", "right": "issue:35141", "accept": false, - "reason": "Flash Attention test failures are not the same as tied-weights reinitialization during embedding resize." + "reason": "Failing Flash Attention 2 tests do not match token-embedding resize/reinit behavior." + }, + { + "left": "issue:38175", + "right": "issue:44493", + "accept": false, + "reason": "Siglip2 zero-probability output issue and unexpected position-id keys are different symptoms and code paths." }, { "left": "issue:29870", "right": "issue:32090", "accept": false, - "reason": "Different runtime failures: BART generate() decoder mask bug vs Trainer broadcast of NoneType." + "reason": "BART generate decoder mask assignment bug is unrelated to Trainer _gpu_broadcast_one NoneType error." }, { "left": "issue:43278", "right": "issue:43329", "accept": false, - "reason": "Embedding dtype drift in eval is unrelated to the multimodal token-count video-branch undefined variables." - }, - { - "left": "issue:44360", - "right": "issue:44561", - "accept": false, - "reason": "DSA indexer activation bug is unrelated to the v5 removal of is_torch_fx_available breaking trust_remote_code models." + "reason": "Embedding dtype drift in eval is a different problem from undefined video-token helper variables." }, { "left": "issue:45290", "right": "issue:45381", "accept": false, - "reason": "Chat template tool-call crash vs Qwen2.5-VL video vision_position_ids bug are different code paths." + "reason": "Chat-template tool-call crash and Qwen2.5-VL video position-id bug affect different functions." }, { "left": "issue:45920", "right": "issue:46192", "accept": false, - "reason": "Both tokenizer-related, but one is wrong token IDs across models while the other is return_tensors being ignored in a specific API shape." + "reason": "Both are tokenizer-related, but one is wrong token IDs and the other is return_tensors being ignored." }, { "left": "issue:44448", "right": "issue:44977", "accept": false, - "reason": "Different model families and failures: Pegasus v4/v5 output divergence vs Qwen3.5 flash-attention generation bug." + "reason": "Pegasus v4/v5 output divergence is not the same concrete bug as Qwen3.5 flash-attention generation failure." }, { "left": "issue:10000", "right": "issue:15354", "accept": false, - "reason": "German DistilBERT issue vs torch.jit GeneratorExp export failure are unrelated." + "reason": "German DistilBert issue and torch.jit.script GeneratorExp export failure are unrelated." + }, + { + "left": "issue:44360", + "right": "issue:44561", + "accept": false, + "reason": "DSA indexer missing ReLU is unrelated to removal of is_torch_fx_available breaking trust_remote_code." }, { "left": "issue:44977", "right": "issue:45229", "accept": false, - "reason": "Flash-attention generation bug in Qwen3.5 vs multi-GPU CUDA OOM for Gemma4 are not the same defect." + "reason": "Qwen3.5 flash-attention generation bug and Gemma4 multi-GPU OOM are different failure modes." }, { "left": "issue:44855", "right": "issue:46399", "accept": false, - "reason": "Python 3.13 import/IndentationError from comment placement is unrelated to a missing-decorator registration crash in Qwen3.5." + "reason": "Python 3.13 IndentationError from torch.jit.script parsing is unrelated to a missing decorator in apply_rotary_pos_emb registration." }, { "left": "issue:29870", "right": "issue:33290", "accept": false, - "reason": "Decoder attention mask assignment bug vs Adafactor/DeepSpeed OOM are different failure modes." + "reason": "BART generation bug and DeepSpeed Adafactor OOM are different subsystems and symptoms." }, { - "left": "issue:43927", - "right": "issue:45698", + "left": "issue:45290", + "right": "issue:46655", "accept": false, - "reason": "DiaConfig custom token IDs lost on save/load is a different persistence bug than loading the wrong custom module after save_pretrained." + "reason": "Assistant-tool-call chat template crash is not the same as SmolVLM video padding miscalculation." }, { - "left": "issue:45290", - "right": "issue:46655", + "left": "issue:43927", + "right": "issue:45698", "accept": false, - "reason": "Chat template tool-call crash is unrelated to incorrect padding in SmolVLM's video processor." + "reason": "One is custom token IDs lost in config save/load; the other is the wrong custom module loaded after save_pretrained." }, { "left": "issue:40444", "right": "issue:43526", "accept": false, - "reason": "Qwen2.5-VL iterable dataset/multi-image finetuning failure is not the same as BeitImageProcessorFast reduce_labels returning one label." + "reason": "Qwen2.5-VL iterable multi-image training failure and BeitImageProcessorFast reduce_labels bug are unrelated." }, { "left": "issue:43278", "right": "issue:43381", "accept": false, - "reason": "Eval dtype change is unrelated to gradient checkpointing being disallowed in eval mode." + "reason": "Eval-time BF16->FP32 embedding dtype change is not the same bug as gradient checkpointing being disallowed in eval mode." }, { "left": "issue:37428", "right": "issue:42175", "accept": false, - "reason": "Flash support import error is unrelated to missing TensorFlow in the torch-only pip extra." + "reason": "Missing flash-support import and tensorflow not included in torch extra are different packaging/import issues." }, { "left": "issue:32090", "right": "issue:33290", "accept": false, - "reason": "Trainer broadcast NoneType error and DeepSpeed Adafactor OOM are different bugs." + "reason": "Trainer broadcast NoneType error is unrelated to OOM with Adafactor in DeepSpeed." }, { "left": "issue:46421", "right": "issue:46858", "accept": false, - "reason": "Spurious recompilation from cache lazy init is a different issue than get_seq_length() returning a device tensor and breaking arithmetic." + "reason": "Both involve cache/compile, but one is spurious recompilation from lazy init and the other is device-tensor seq-length arithmetic failure." }, { "left": "issue:43873", "right": "issue:45229", "accept": false, - "reason": "Quantization offloading behavior is unrelated to Gemma4 multi-GPU CUDA OOM." + "reason": "Quantization offloading not working and Gemma4 multi-GPU CUDA OOM are not the same underlying bug." }, { "left": "issue:42490", "right": "issue:43873", "accept": false, - "reason": "Local-path import warning is unrelated to quantization offloading malfunction." + "reason": "A trailing-slash local-model import warning is unrelated to quantization offloading behavior." }, { "left": "issue:25251", "right": "issue:32090", "accept": false, - "reason": "Pipeline top_k nesting bug is unrelated to Trainer broadcast of NoneType." - }, - { - "left": "issue:44561", - "right": "issue:44779", - "accept": false, - "reason": "Missing is_torch_fx_available breaks trust_remote_code imports; DeepSeek tokenizer regression is a different tokenizer path." + "reason": "Pipeline top_k nesting change has no overlap with Trainer GPU broadcast NoneType errors." }, { "left": "issue:43976", "right": "issue:45440", "accept": false, - "reason": "Python-version compatibility issue is unrelated to native DeepseekV3MoE diverging from remote implementation." + "reason": "Python-version compatibility issue is unrelated to DeepseekV3MoE divergence from remote implementation." }, { "left": "issue:44448", "right": "issue:44466", "accept": false, - "reason": "Both are v5-related but concern different behaviors: Pegasus output divergence vs lm_head.weight serialization tied-weights inconsistency." + "reason": "Different regression types: model output mismatch vs lm_head serialization/tied-weights inconsistency." }, { "left": "issue:44568", "right": "issue:45491", "accept": false, - "reason": "mDeBERTa BOS/EOS tokenization regression is unrelated to Gemma3 NaN embeddings from mixed-length batching." + "reason": "Tokenizer BOS/EOS omission and Gemma3 NaN embeddings are different code paths and failure modes." } ] }, @@ -10213,7 +10202,7 @@ { "label": "issue", "component_index": 1, - "component_count": 25, + "component_count": 26, "review_unit_index": 33, "review_unit_count": 61, "status": "reviewed", @@ -10230,10 +10219,8 @@ "issue:33357", "issue:34567", "issue:34689", - "issue:35141", "issue:39692", "issue:41720", - "issue:42490", "issue:42915", "issue:42994", "issue:43065", @@ -10252,7 +10239,9 @@ "issue:44361", "issue:44442", "issue:44479", + "issue:44561", "issue:44743", + "issue:44779", "issue:44871", "issue:44977", "issue:45198", @@ -10274,6 +10263,7 @@ ], "soft_pairs": [ "issue:29127|issue:34567", + "issue:44561|issue:44779", "issue:44871|issue:44977", "issue:33357|issue:39692", "issue:45588|issue:45823", @@ -10294,238 +10284,230 @@ "issue:43756|issue:45440", "issue:43653|issue:44442", "issue:43412|issue:45588", - "issue:45459|issue:45712", "issue:43065|issue:43994", - "issue:45491|issue:46693", "issue:44265|issue:45198", + "issue:45491|issue:46693", + "issue:45459|issue:45712", "issue:29127|issue:33357", "issue:43065|issue:45412", - "issue:45372|issue:45406", "issue:43575|issue:43701", - "issue:45245|issue:45563", - "issue:42490|issue:43381", - "issue:29127|issue:35141" + "issue:45372|issue:45406", + "issue:45245|issue:45563" ], - "prepared_review_unit_hash": "5c2ac897c41f9c30534974525fe6c10f0e598e8107e2989086261cf621606251", + "prepared_review_unit_hash": "5166ccbe164be9486c4b5c38c37dc3a34821569f103ebedcb448c9ff8ed1e7e2", "component_budget": { "node_count": 505, "item_count": 505, "soft_pair_count": 1854, - "serialized_chars": 962354, - "estimated_input_tokens": 240589, - "estimated_eval_tokens": 481434 + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 32, - "serialized_chars": 37439, - "estimated_input_tokens": 9360, - "estimated_eval_tokens": 18976 + "soft_pair_count": 31, + "serialized_chars": 37086, + "estimated_input_tokens": 9272, + "estimated_eval_tokens": 18800 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "These issues are not a duplicate cluster: they span unrelated bugs across tokenizers, vision, generation, training state, export, quantization, and infrastructure. All soft edges are best rejected.", - "confidence": 0.95, - "canonical_issue_reason": "No obvious canonical duplicate: the set is a grab-bag of unrelated issues, so no single issue can safely represent the cluster.", + "summary": "These items are a broad mix of unrelated bug reports across tokenizers, generation, export, processor loading, and model-specific regressions; none of the soft pairs look like true duplicates of the same underlying bug/change.", + "confidence": 0.92, + "canonical_issue_reason": "issue:46693 is the strongest representative issue: it is current, clearly framed as a concrete regression, and has active discussion/inbound references, making it a good cluster anchor despite the cluster being heterogeneous.", "canonical_pr_reason": null, - "best_issue_reason": "issue:46693 is the most active and concrete regression report, so it's the least-bad standalone anchor, but it still doesn't unify the rest of the items.", + "best_issue_reason": "issue:46693 is the best overall issue to surface because it is recent, broadly impactful, and sufficiently concrete to track and triage.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:29127", "right": "issue:34567", "accept": false, - "reason": "LayoutLMv3 error messaging vs TrainerState token counting are unrelated bugs in different subsystems." + "reason": "Different subsystems and failure modes: LayoutLMv3 error messaging vs TrainerState token counting." + }, + { + "left": "issue:44561", + "right": "issue:44779", + "accept": false, + "reason": "One is a v5 compatibility break for trust_remote_code models; the other is a DeepSeek tokenizer correctness regression." }, { "left": "issue:44871", "right": "issue:44977", "accept": false, - "reason": "Gemma3 eos_token_id mismatch and Qwen3.5 flash-attention generation failure are different model-specific problems." + "reason": "Gemma eos_token_id mismatch and Qwen flash-attention generation bug are unrelated." }, { "left": "issue:33357", "right": "issue:39692", "accept": false, - "reason": "MacOS CLIP bus error and SigLIP2 doc/example issues are not the same underlying failure." + "reason": "MacOS bus error on CLIP is not the same as SigLIP2 documentation/example issues." }, { "left": "issue:45588", "right": "issue:45823", "accept": false, - "reason": "Flash-attention s_aux=None crash and Gemma4 PLE device mismatch are distinct code-path bugs." + "reason": "Flash-attention sinkless-model crash and Gemma4 device-map mismatch are different code paths." }, { "left": "issue:43381", "right": "issue:43701", "accept": false, - "reason": "Eval-mode gradient checkpointing limitation and resume_from_checkpoint key mismatch are unrelated training issues." + "reason": "Eval-mode gradient checkpointing limitation is unrelated to resume_from_checkpoint key mismatch." }, { "left": "issue:43278", "right": "issue:43575", "accept": false, - "reason": "Embedding dtype drift in evaluate and Qwen2 TP OOM are different problems." + "reason": "Training/eval dtype drift and tensor-parallel OOM are separate issues." }, { "left": "issue:45198", "right": "issue:45803", "accept": false, - "reason": "Wav2Vec2 save/tokenization failure and Fuyu bare-except error handling are unrelated." + "reason": "Wav2Vec2 save/tokenization failure is unrelated to Fuyu's bare-except error handling bug." }, { "left": "issue:41720", "right": "issue:45229", "accept": false, - "reason": "Both involve large-model inference, but Qwen3 auto device-map cuda assert and Gemma4 multi-GPU OOM are different failures." + "reason": "CUDA assert during auto device mapping differs from multi-GPU OOM on Gemma4." }, { "left": "issue:44977", "right": "issue:45563", "accept": false, - "reason": "Qwen3.5 flash-attention generation bug and stale warning for num_return_sequences are unrelated." + "reason": "Generation failure with flash-attention and a stale warning in paged generate() are not the same bug." }, { "left": "issue:44186", "right": "issue:45198", "accept": false, - "reason": "LayoutLMv2Tokenizer NER/padding crash and Wav2Vec2 save/tokenization failure are different tokenizer bugs." + "reason": "LayoutLMv2 tokenizer crashes on NER padding/truncation; Wav2Vec2 has save/tokenization failure." }, { "left": "issue:44361", "right": "issue:45198", "accept": false, - "reason": "MLukeTokenizer AttributeError and Wav2Vec2 save/tokenization failure do not share the same root cause." + "reason": "MLuke tokenizer AttributeError is unrelated to Wav2Vec2 save/tokenization issues." }, { "left": "issue:43611", "right": "issue:46121", "accept": false, - "reason": "Base_model_prefix loading regression and rope-parameter JSON list TypeError are unrelated." + "reason": "base_model_prefix loading break and rope-parameter JSON list TypeError are different changes." }, { "left": "issue:29870", "right": "issue:34689", "accept": false, - "reason": "BART decoder_attention_mask bug and Llama 3.2 Vision loading break are different issues." + "reason": "BART generate() UnboundLocalError and Llama 3.2 Vision loading break are unrelated." }, { "left": "issue:43653", "right": "issue:44743", "accept": false, - "reason": "BigBirdTokenizer special-token registration and Qwen3.5 recurrent-state reset are unrelated." + "reason": "BigBird special-token registration and Qwen recurrent-state reset are different tokenizer/model bugs." }, { "left": "issue:44206", "right": "issue:44479", "accept": false, - "reason": "Feature extractor center-arg crash and video input regression affect different preprocessing paths." + "reason": "Unsupported center arg in image processing is unrelated to the video-input regression across Qwen VL models." }, { "left": "issue:29942", "right": "issue:30064", "accept": false, - "reason": "Flash Attention 2 test failures and void segmentation map processing are unrelated." + "reason": "Failing Flash Attention 2 tests and void segmentation-map processing are not the same underlying bug." }, { "left": "issue:42915", "right": "issue:42994", "accept": false, - "reason": "Qwen3Moe FineGrainedFP8Config failure and quantized model saving failure are different quantization bugs." + "reason": "FineGrainedFP8Config failure and quantized model saving failure are different quantization issues." }, { "left": "issue:45372", "right": "issue:45588", "accept": false, - "reason": "Gemma4 processor import error and flash-attention s_aux=None crash are unrelated." + "reason": "ReasoningEffort import break in Gemma4 processor loading is separate from flash_attention s_aux=None crash." }, { "left": "issue:43756", "right": "issue:45440", "accept": false, - "reason": "Smollm3 RoPE-layer mismatch and DeepseekV3MoE divergence are different model-implementation parity issues." + "reason": "Smollm3 RoPE-layer mismatch and DeepseekV3MoE divergence are model-specific but distinct." }, { "left": "issue:43653", "right": "issue:44442", "accept": false, - "reason": "BigBirdTokenizer decode bug and AutoTokenizer loading FastSpeech2ConformerTokenizer are unrelated tokenizer problems." + "reason": "BigBird tokenizer decode bug and AutoTokenizer loading FastSpeech2ConformerTokenizer are unrelated." }, { "left": "issue:43412", "right": "issue:45588", "accept": false, - "reason": "Executorch export failure and flash-attention crash are different code paths." + "reason": "Executorch export/training-guard failure is unrelated to flash_attention integration crash." }, { - "left": "issue:45459", - "right": "issue:45712", + "left": "issue:43065", + "right": "issue:43994", "accept": false, - "reason": "Protobuf error masking in tokenizer conversion and leftover dummy PT classes are unrelated maintenance issues." + "reason": "Sam3 dummy Conv2d cleanup and SigLIP2 nonsensical outputs are different issues." }, { - "left": "issue:43065", - "right": "issue:43994", + "left": "issue:44265", + "right": "issue:45198", "accept": false, - "reason": "Sam3PixelDecoder dummy Conv2d and SigLIP2 nonsensical outputs are unrelated vision-model bugs." + "reason": "torch.export with torch_compilable_check and Wav2Vec2 save/tokenization failure are unrelated." }, { "left": "issue:45491", "right": "issue:46693", "accept": false, - "reason": "Gemma3 NaN embeddings and FA2 long-context performance regression are different correctness vs performance issues." + "reason": "Both mention FA2, but one is a NaN correctness bug and the other is a performance regression; not the same concrete fix." }, { - "left": "issue:44265", - "right": "issue:45198", + "left": "issue:45459", + "right": "issue:45712", "accept": false, - "reason": "torch.export failure with torch_compilable_check and Wav2Vec2 save/tokenization failure are unrelated." + "reason": "Tokenizer error masking via bare except and leftover dummy classes are unrelated repository issues." }, { "left": "issue:29127", "right": "issue:33357", "accept": false, - "reason": "LayoutLMv3 error-message clarity and CLIP MacOS bus error are unrelated." + "reason": "LayoutLMv3 error messaging and CLIP MacOS bus error do not share the same bug." }, { "left": "issue:43065", "right": "issue:45412", "accept": false, - "reason": "Dummy Sam3PixelDecoder class cleanup and RT-DETR memory release are unrelated." - }, - { - "left": "issue:45372", - "right": "issue:45406", - "accept": false, - "reason": "Gemma4 processor import failure and Gemma4Processor missing _tokenizer in serve are different concrete bugs." + "reason": "Dummy class cleanup in SAM3 is unrelated to RT-DETR memory-release behavior." }, { "left": "issue:43575", "right": "issue:43701", "accept": false, - "reason": "Qwen2 TP OOM and resume_from_checkpoint key mismatch do not share a root cause." - }, - { - "left": "issue:45245", - "right": "issue:45563", - "accept": false, - "reason": "Category-count runtime error and stale generate warning are unrelated." + "reason": "Tensor-parallel OOM on Qwen2 is unrelated to checkpoint key mismatch." }, { - "left": "issue:42490", - "right": "issue:43381", + "left": "issue:45372", + "right": "issue:45406", "accept": false, - "reason": "Trailing-slash import warning and eval-mode gradient checkpointing are unrelated." + "reason": "Gemma4 processor import failure and serve-time missing _tokenizer attribute are related by model family but not clearly the same concrete bug." }, { - "left": "issue:29127", - "right": "issue:35141", + "left": "issue:45245", + "right": "issue:45563", "accept": false, - "reason": "LayoutLMv3 clarity issue and post_init reinitializing resized embeddings are unrelated." + "reason": "Category-count runtime error and stale num_return_sequences warning are unrelated." } ] }, @@ -10539,7 +10521,7 @@ { "label": "issue", "component_index": 1, - "component_count": 25, + "component_count": 26, "review_unit_index": 34, "review_unit_count": 61, "status": "reviewed", @@ -10554,6 +10536,7 @@ "issue:33290", "issue:34689", "issue:35141", + "issue:42490", "issue:42915", "issue:43066", "issue:43334", @@ -10598,17 +10581,19 @@ "issue:46821" ], "soft_pairs": [ + "issue:42490|issue:43381", + "issue:29127|issue:35141", "issue:45440|issue:45593", - "issue:45372|issue:45459", "issue:43994|issue:45412", + "issue:45372|issue:45459", "issue:45406|issue:46701", "issue:43421|issue:45491", - "issue:43066|issue:44462", "issue:45412|issue:46369", "issue:43575|issue:44387", + "issue:43066|issue:44462", "issue:44442|issue:45072", - "issue:44464|issue:45584", "issue:43526|issue:45834", + "issue:44464|issue:45584", "issue:43824|issue:44279", "issue:42915|issue:43931", "issue:43526|issue:46693", @@ -10622,207 +10607,255 @@ "issue:43334|issue:43531", "issue:46121|issue:46399", "issue:45588|issue:46821", - "issue:45250|issue:46640", "issue:43381|issue:43873", + "issue:45250|issue:46640", "issue:44514|issue:45381", "issue:43646|issue:44077" ], - "prepared_review_unit_hash": "41fc2c701a6861dbd35d87471566a8dfe7af5b495051afd06b0e194d2d392ea3", + "prepared_review_unit_hash": "b9921f961547dd6e27140cb06ac2625a94f855b59f1380a917e1d257edf6331d", "component_budget": { "node_count": 505, "item_count": 505, "soft_pair_count": 1854, - "serialized_chars": 962354, - "estimated_input_tokens": 240589, - "estimated_eval_tokens": 481434 + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 }, "budget": { - "node_count": 47, - "item_count": 47, - "soft_pair_count": 28, - "serialized_chars": 35341, - "estimated_input_tokens": 8836, - "estimated_eval_tokens": 17928 + "node_count": 48, + "item_count": 48, + "soft_pair_count": 30, + "serialized_chars": 36514, + "estimated_input_tokens": 9129, + "estimated_eval_tokens": 18514 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "No soft-edge pair looks like the same underlying bug/change; the overlaps are mostly broad subsystem similarity (Flash Attention, Qwen, tokenizer, memory/OOM) rather than the same concrete failure mode.", - "confidence": 0.93, - "canonical_issue_reason": "issue:43873 is the best standalone anchor here: it is open, detailed, and has the strongest discussion/inbound-reference signal, making it the most actionable representative issue in this set.", + "summary": "These soft matches are mostly title-level similarity only; the paired issues describe different models, subsystems, or failure modes, so none should be deduplicated.", + "confidence": 0.94, + "canonical_issue_reason": "No clear canonical issue: the set is heterogeneous and the apparent similarities are mostly broad phrasing, not the same underlying bug.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43873 is the strongest issue to keep as the global representative because it has the most community signal and the clearest actionable bug report.", + "best_issue_reason": "issue:46640 is the closest umbrella for the only recurring theme (Flash Attention 2), but it is still too generic to serve as a true duplicate hub.", "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:42490", + "right": "issue:43381", + "accept": false, + "reason": "Local-path import warning vs gradient-checkpointing eval-mode behavior; unrelated bugs." + }, + { + "left": "issue:29127", + "right": "issue:35141", + "accept": false, + "reason": "LayoutLMv3 error-message clarity vs token-embedding resize/reinit; different code paths." + }, { "left": "issue:45440", "right": "issue:45593", "accept": false, - "reason": "Different bugs: DeepSeekV3MoE implementation divergence vs D-FINE auxiliary-loss behavior when denoising is off." + "reason": "DeepseekV3MoE output divergence vs D-FINE auxiliary-loss logic; unrelated model bugs." }, { - "left": "issue:45372", - "right": "issue:45459", + "left": "issue:43994", + "right": "issue:45412", "accept": false, - "reason": "Both involve tokenizer/processor loading, but one is a mistral_common import breakage and the other hides protobuf-related tokenizer errors; not the same failure." + "reason": "SigLIP2 bad outputs vs RT-DETR memory release; different symptoms and subsystems." }, { - "left": "issue:43994", - "right": "issue:46369", + "left": "issue:45372", + "right": "issue:45459", "accept": false, - "reason": "Different models and symptoms: SigLIP2 nonsensical outputs vs SAM3-lite memory overflow." + "reason": "Gemma processor import failure vs tokenizer-error masking when protobuf is missing; not the same bug." }, { "left": "issue:45406", "right": "issue:46701", "accept": false, - "reason": "Both touch Gemma-related paths, but one is a serve-time missing _tokenizer attribute and the other is an image-text-to-text pipeline output issue." + "reason": "Gemma4Processor missing _tokenizer vs DiffusionGemma generation output/pipeline failure; different issues." }, { "left": "issue:43421", "right": "issue:45491", "accept": false, - "reason": "Tokenizer post-processor syncing and Gemma3 NaN embeddings are unrelated code paths and failure modes." - }, - { - "left": "issue:43066", - "right": "issue:44462", - "accept": false, - "reason": "Tokenizer decoder-type regression and ignoring tokenizer.json are different tokenizer-loading problems." + "reason": "Tokenizer post-processor update feature vs Gemma3 NaN embeddings with sliding-window attention; unrelated." }, { "left": "issue:45412", "right": "issue:46369", "accept": false, - "reason": "Both are memory-related, but one is a deletion/GC leak in RT-DETR and the other is an overflow in sam3-litetext." + "reason": "RT-DETR deletion memory leak vs SAM3-LiteText memory overflow; both memory-related but distinct bugs." }, { "left": "issue:43575", "right": "issue:44387", "accept": false, - "reason": "Both mention OOM, but one is TP loading of a specific Qwen model and the other is an int4 quantization reserved-memory regression." + "reason": "Both are OOM reports, but one is TP load OOM and the other is int4 quantization reserved-memory regression." }, { - "left": "issue:44442", - "right": "issue:45072", + "left": "issue:43066", + "right": "issue:44462", "accept": false, - "reason": "FastSpeech2Conformer tokenizer loading and dtype mismatches in bfloat16 inference are different issues." + "reason": "Wrong tokenizer decoder type in v5 vs AutoTokenizer ignoring tokenizer.json; different loading failures." }, { - "left": "issue:44464", - "right": "issue:45584", + "left": "issue:44442", + "right": "issue:45072", "accept": false, - "reason": "Chunked generation inconsistency with compiled forward and Whisper empty-transcription failure after token alignment are distinct bugs." + "reason": "Tokenizer loading failure vs dtype mismatch in bfloat16 inference; unrelated." }, { "left": "issue:43526", "right": "issue:45834", "accept": false, - "reason": "Beit label-reduction behavior and Kosmos2.5 OCR index errors are unrelated." + "reason": "BEiT label reduction bug vs Kosmos2.5 long-OCR index error; different models and failure modes." + }, + { + "left": "issue:44464", + "right": "issue:45584", + "accept": false, + "reason": "Chunked generation inconsistency with compiled forward vs Whisper empty-transcription failure after token alignment; not the same bug." }, { "left": "issue:43824", "right": "issue:44279", "accept": false, - "reason": "An ImportError for Qwen2_5_VLForConditionalGeneration is not the same as a generic dependency issue." + "reason": "Missing Qwen2.5-VL import symbol vs vague dependency issue; too broad and not a shared concrete defect." }, { "left": "issue:42915", "right": "issue:43931", "accept": false, - "reason": "FineGrainedFP8Config failure in Qwen3MoE and weight-shape mismatch in Qwen3-VL are different model/config problems." + "reason": "Qwen3MoE FP8 config failure vs Qwen3-VL weight-shape mismatch; different models and root causes." }, { - "left": "issue:43526", - "right": "issue:46693", + "left": "issue:43334", + "right": "issue:43531", + "accept": false, + "reason": "Missing pad_token_id on Qwen3-VL vs sliding_window issue on Qwen3-MoE; unrelated." + }, + { + "left": "issue:46121", + "right": "issue:46399", + "accept": false, + "reason": "RoPE validation TypeError vs missing decorator for apply_rotary_pos_emb registration; unrelated code paths." + }, + { + "left": "issue:45588", + "right": "issue:46821", + "accept": false, + "reason": "flash_attention.py s_aux=None crash vs TimesFM window_size AttributeError; same broad area only, not same bug." + }, + { + "left": "issue:43381", + "right": "issue:43873", + "accept": false, + "reason": "Gradient checkpointing in eval mode vs offloading under quantization; unrelated." + }, + { + "left": "issue:45250", + "right": "issue:46640", + "accept": false, + "reason": "Both mention Flash Attention 2, but one is a generic feature/title and the other a specific issue report; not the same concrete change/bug." + }, + { + "left": "issue:44514", + "right": "issue:45381", + "accept": false, + "reason": "Batched apply_chat_template padding crash vs Qwen2.5-VL video position-id bug; different processing stages." + }, + { + "left": "issue:43646", + "right": "issue:44077", "accept": false, - "reason": "Label reduction in BeitImageProcessorFast and FA2 long-context performance regression are unrelated." + "reason": "Custom model initialization breakage vs patchtsmixer post_init policy; related only at a very high level." }, { "left": "issue:27957", "right": "issue:33290", "accept": false, - "reason": "XLMRoberta Flash Attention 2 support and Adafactor/DeepSpeed OOM are different bugs." + "reason": "XLMRoberta Flash Attention 2 report vs Adafactor/deepspeed OOM; unrelated." }, { "left": "issue:27957", "right": "issue:29127", "accept": false, - "reason": "Flash Attention support for XLMRoberta and a LayoutLMv3 error-message clarity issue do not match." + "reason": "Flash Attention 2 on XLMRoberta vs LayoutLMv3 error clarity; no underlying overlap." }, { "left": "issue:44857", "right": "issue:45588", "accept": false, - "reason": "float16 AMP crash in LwDetrImageLoss and s_aux=None AttributeError in flash_attention.py are different code paths." + "reason": "LwDetrImageLoss AMP crash vs flash-attention sinkless-model AttributeError; different failures." }, { "left": "issue:43994", "right": "issue:46369", "accept": false, - "reason": "SigLIP2 wrong outputs and SAM3-lite memory overflow are not the same bug." + "reason": "SigLIP2 nonsensical outputs vs SAM3-LiteText memory overflow; unrelated." }, { "left": "issue:43526", "right": "issue:44186", "accept": false, - "reason": "BeitImageProcessorFast reduce_labels bug and LayoutLMv2Tokenizer crash on NER/padding are different tokenizer/processor issues." + "reason": "BEiTImageProcessorFast reduce_labels bug vs LayoutLMv2Tokenizer NER/padding crash; different components." }, { "left": "issue:33290", "right": "issue:34689", "accept": false, - "reason": "DeepSpeed Adafactor OOM and Llama 3.2 90B Vision load breakage are unrelated failures." + "reason": "Adafactor/deepspeed OOM vs Llama 3.2 Vision model-loading breakage; unrelated." }, { "left": "issue:33290", "right": "issue:35141", "accept": false, - "reason": "OOM with Adafactor/DeepSpeed and embedding reinitialization after resize are different bugs." + "reason": "Optimizer OOM vs embedding reinitialization bug; different code paths." }, { "left": "issue:43334", "right": "issue:43531", "accept": false, - "reason": "Missing pad_token_id on Qwen3-VL text config and a sliding_window issue in Qwen3-MoE are distinct." + "reason": "Qwen3-VL missing pad_token_id vs Qwen3-MoE sliding-window issue; unrelated." }, { "left": "issue:46121", "right": "issue:46399", "accept": false, - "reason": "JSON-loaded list handling in rope validation and missing decorator registration for apply_rotary_pos_emb are unrelated." + "reason": "RoPE validation TypeError vs Qwen3.5 kernel registration decorator issue; unrelated." }, { "left": "issue:45588", "right": "issue:46821", "accept": false, - "reason": "flash_attention.py s_aux=None crash and TimesFM window_size AttributeError are different subsystems and errors." + "reason": "flash_attention.py AttributeError vs TimesFM window_size AttributeError; same generic error type only, not same bug." }, { - "left": "issue:45250", - "right": "issue:46640", + "left": "issue:43381", + "right": "issue:43873", "accept": false, - "reason": "The titles are both generic FA2 reports, but there is no concrete evidence they describe the same bug or change." + "reason": "Eval-mode gradient checkpointing restriction vs quantization offloading behavior; unrelated." }, { - "left": "issue:43381", - "right": "issue:43873", + "left": "issue:45250", + "right": "issue:46640", "accept": false, - "reason": "Gradient checkpointing in eval mode and quantization offloading problems are different behaviors." + "reason": "Generic Flash Attention 2 feature/title overlap is insufficient to merge as duplicates." }, { "left": "issue:44514", "right": "issue:45381", "accept": false, - "reason": "Batched apply_chat_template padding crash and wrong vision_position_ids in Qwen2.5-VL video input are different processor/model issues." + "reason": "Batched processor crash vs video vision-position-id bug; not the same underlying defect." }, { "left": "issue:43646", "right": "issue:44077", "accept": false, - "reason": "Generic Transformers 5.0 custom initialization breakage and patchtsmixer post_init policy are not the same concrete bug." + "reason": "Both mention initialization/post_init, but the reported problems are different and not mergeable into one fix." } ] }, @@ -10836,7 +10869,7 @@ { "label": "issue", "component_index": 1, - "component_count": 25, + "component_count": 26, "review_unit_index": 35, "review_unit_count": 61, "status": "reviewed", @@ -10902,14 +10935,14 @@ "issue:43262|issue:44206", "issue:44442|issue:44898", "issue:42490|issue:44743", - "issue:45003|issue:45440", "issue:43720|issue:44442", + "issue:45003|issue:45440", "issue:42175|issue:46234", "issue:27957|issue:29870", "issue:44077|issue:44479", "issue:43493|issue:43827", - "issue:43316|issue:43525", "issue:44062|issue:45907", + "issue:43316|issue:43525", "issue:43873|issue:44871", "issue:45325|issue:46697", "issue:43643|issue:44913", @@ -10917,203 +10950,203 @@ "issue:44855|issue:45137", "issue:43316|issue:45759", "issue:44987|issue:46178", - "issue:44190|issue:44496", "issue:44442|issue:45803", + "issue:44190|issue:44496", "issue:43761|issue:44079", "issue:43931|issue:45216", "issue:46829|issue:46932", "issue:43232|issue:43388" ], - "prepared_review_unit_hash": "3c228f96f399f9793b2b1bd84fff5a5ffdb7aee84e25a79ae5d7e5c6dacb2dd2", + "prepared_review_unit_hash": "40de04a84bfb01f5b7b27bafff921f590ed7a8272542e87444b7a2ac8543bc9d", "component_budget": { "node_count": 505, "item_count": 505, "soft_pair_count": 1854, - "serialized_chars": 962354, - "estimated_input_tokens": 240589, - "estimated_eval_tokens": 481434 + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 27, - "serialized_chars": 36162, - "estimated_input_tokens": 9041, - "estimated_eval_tokens": 18338 + "serialized_chars": 36167, + "estimated_input_tokens": 9042, + "estimated_eval_tokens": 18340 }, "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 generation, model/config loading, offloading, audio/vision processors, and docs. None of the soft pairs look like the same underlying bug or change, so I would not merge any of them as duplicates.", - "confidence": 0.94, - "canonical_issue_reason": "If a canonical issue must be chosen, issue:43749 is the best anchor because it has the most discussion/inbound references and reflects a broad, high-impact loading regression. That said, the cluster overall does not form a single cohesive duplicate set.", + "summary": "This cluster is heterogeneous: the issues cover many unrelated models, loading paths, processors, and generation bugs. None of the soft pairs look like the same underlying bug/change, so all should be kept separate.", + "confidence": 0.97, + "canonical_issue_reason": "No single canonical issue stands out because the items span unrelated subsystems and failure modes.", "canonical_pr_reason": null, - "best_issue_reason": "Issue:43749 is the broadest and most connected report in the set, making it the most suitable representative issue for the cluster, even though the surrounding items are mostly unrelated.", + "best_issue_reason": "No issue here is a strong global representative; the cluster does not form one duplicate family.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43761", "right": "issue:45859", "accept": false, - "reason": "Different models and failure modes: CLIPVision hidden_states regression vs Qwen3_5Moe CPU-offload incompatibility." + "reason": "Both involve Qwen/rotary behavior, but one is a CLIPVision hidden-states regression and the other is CPU offload incompatibility in Qwen3_5Moe; different bugs and code paths." }, { "left": "issue:44977", "right": "issue:45245", "accept": false, - "reason": "One is a flash-attention generation bug; the other is a category-count runtime limit. No shared code path." + "reason": "Qwen3.5 flash-attention generation failure is unrelated to a category-count runtime error in a separate vision/classification path." }, { "left": "issue:44743", "right": "issue:44811", "accept": false, - "reason": "Recurrent-state reset in Qwen modular code vs Whisper batch_decode skipping params are unrelated." + "reason": "One is recurrent-state reset in modular_qwen3_5 cache handling; the other is Whisper processor batch_decode skip_special_tokens handling. Different components." }, { "left": "issue:43262", "right": "issue:44206", "accept": false, - "reason": "Both are audio-related, but one is chat template sampling-rate handling and the other is a feature-extractor arg crash." + "reason": "Audio processor sampling-rate default bug and LasrFeatureExtractor unsupported center argument are distinct processor issues." }, { "left": "issue:44442", "right": "issue:44898", "accept": false, - "reason": "Tokenizer loading failure vs Perceiver image-classification resizing/interpolation failure are distinct bugs." + "reason": "FastSpeech2ConformerTokenizer loading failure is not the same as Perceiver interpolation/padding failure; different tokenizer vs image-model code." }, { "left": "issue:42490", "right": "issue:44743", "accept": false, - "reason": "Local-path trailing-slash import warning is unrelated to Qwen recurrent-state caching." + "reason": "Local-path import warning and Qwen recurrent cache reset are unrelated." }, { - "left": "issue:45003", - "right": "issue:45440", + "left": "issue:43720", + "right": "issue:44442", "accept": false, - "reason": "Unsafe sys.modules access and DeepSeekV3 implementation divergence are different code paths and symptoms." + "reason": "BitNet packed-weight unpacking during accelerate loading is a model-loading bug, not the tokenizer-loading failure described here." }, { - "left": "issue:43720", - "right": "issue:44442", + "left": "issue:45003", + "right": "issue:45440", "accept": false, - "reason": "BitNet packed-weight loading bug vs FastSpeech2 tokenizer loading failure are unrelated." + "reason": "Unsafe sys.modules access in modeling_utils is a generic import issue, while DeepseekV3MoE divergence is a model-implementation mismatch." }, { "left": "issue:42175", "right": "issue:46234", "accept": false, - "reason": "Missing TensorFlow in extras install and a missing import_utils helper are separate packaging/import issues." + "reason": "TensorFlow missing from a pip extra and a missing import_utils availability symbol are separate packaging/import problems." }, { "left": "issue:27957", "right": "issue:29870", "accept": false, - "reason": "Flash Attention 2 support for XLMRoberta is not the same bug as Bart generate() using decoder_attention_mask before assignment." + "reason": "Flash Attention 2 support for XLMRoberta and a BART generate decoder_attention_mask bug are different models and failures." }, { "left": "issue:44077", "right": "issue:44479", "accept": false, - "reason": "PatchTSMixer post_init policy and Qwen video-input regression are different model-specific issues." + "reason": "patchtsmixer post_init restriction and Qwen video-input regression are unrelated model-specific changes." }, { "left": "issue:43493", "right": "issue:43827", "accept": false, - "reason": "SigLIP2 implementation drift is unrelated to documentation still referencing removed pipeline() APIs." + "reason": "SigLIP2 HF/JAX discrepancy and docs still referencing pipeline() are unrelated; one is model parity, the other documentation cleanup." }, { - "left": "issue:43316", - "right": "issue:43525", + "left": "issue:44062", + "right": "issue:45907", "accept": false, - "reason": "Gemma3TextConfig API mismatch and Llama4Config missing pad_token_id are separate configuration bugs." + "reason": "AddedToken duplicate special-arg error and list_repo_files revision forwarding are unrelated API bugs." }, { - "left": "issue:44062", - "right": "issue:45907", + "left": "issue:43316", + "right": "issue:43525", "accept": false, - "reason": "AddedToken keyword conflict and list_repo_files revision forwarding touch completely different utilities." + "reason": "Gemma3TextConfig API discrepancy and missing pad_token_id on Llama4Config are different config-attribute problems." }, { "left": "issue:43873", "right": "issue:44871", "accept": false, - "reason": "Quantization/offloading behavior and Gemma eos_token_id config inconsistency are not the same defect." + "reason": "Quantization/offloading behavior and Gemma-3 eos_token_id inconsistency are not the same bug." }, { "left": "issue:45325", "right": "issue:46697", "accept": false, - "reason": "Qwen2.5-VL rope-index temporal scaling and Swin2SR padding logic are unrelated image/video preprocessing issues." + "reason": "Qwen2.5-VL rope indexing for still images and Swin2SR over-padding images are different image/video preprocessing bugs." }, { "left": "issue:43643", "right": "issue:44913", "accept": false, - "reason": "trust_remote_code config field loss and GPTNeoX rotary_pct reload regression are different config-loading problems." + "reason": "trust_remote_code missing fields in AutoConfig and GPTNeoX rotary_pct not persisting on reload are distinct config-loading issues." }, { "left": "issue:43749", "right": "issue:43931", "accept": false, - "reason": "FSDP CPU RAM efficient loading failure and Qwen3-VL shape mismatch are distinct loading failures with different root causes." + "reason": "FSDP CPU RAM efficient loading and Qwen3-VL weight-shape mismatch are separate model-loading paths." }, { "left": "issue:44855", "right": "issue:45137", "accept": false, - "reason": "TorchScript parsing on Python 3.13 and DeepSpeed ZeRO3 deque underflow are unrelated runtime errors." + "reason": "Python 3.13 TorchScript parsing failure in DebertaV2 and DeepSpeed ZeRO3 deque pop error are unrelated runtime issues." }, { "left": "issue:43316", "right": "issue:45759", "accept": false, - "reason": "Gemma3TextConfig API discrepancy is not the same as AutoModelForCausalLM failing to unwrap text_config for Qwen multimodal configs." + "reason": "Both touch composite configs, but one is a Gemma3TextConfig API discrepancy and the other is AutoModelForCausalLM not unwrapping text_config for Qwen multimodal configs; different concrete behavior." }, { "left": "issue:44987", "right": "issue:46178", "accept": false, - "reason": "Model loading failure for physical-intelligence/fast and missing tensorflow_text backend mapping are separate backend/config issues." + "reason": "Physical-intelligence/fast loading failure and tensorflow_text backend registration error are different backend/import issues." }, { - "left": "issue:44190", - "right": "issue:44496", + "left": "issue:44442", + "right": "issue:45803", "accept": false, - "reason": "Local dataset loading in a training script is unrelated to an unrecognized model/config type error." + "reason": "Tokenizer loading failure and Fuyu batch feature exception handling are not the same bug." }, { - "left": "issue:44442", - "right": "issue:45803", + "left": "issue:44190", + "right": "issue:44496", "accept": false, - "reason": "Tokenizer loading failure and FuyuBatchFeature's bare except swallowing errors are different code paths." + "reason": "Local dataset loading in a training script and unrecognized model/config.json loading are unrelated load-time errors." }, { "left": "issue:43761", "right": "issue:44079", "accept": false, - "reason": "Hidden-states regression in CLIPVisionModel and ModelOutput key assignment bug do not describe the same behavior." + "reason": "CLIPVision hidden_states regression and ModelOutput key assignment when None are different output-object bugs." }, { "left": "issue:43931", "right": "issue:45216", "accept": false, - "reason": "Qwen3-VL weight-shape mismatch and Qwen3.5 save_pretrained regression are separate model/version issues." + "reason": "Weight-shape mismatch on model load and incorrect Qwen3.5 save_pretrained output are different regressions." }, { "left": "issue:46829", "right": "issue:46932", "accept": false, - "reason": "VideoPrism output field mix-up and Gemma4 quantized embedder dtype casting are different problems." + "reason": "VideoPrism hidden_states misassignment and Gemma4 quantized embedders dtype casting are unrelated model-output vs dtype-casting issues." }, { "left": "issue:43232", "right": "issue:43388", "accept": false, - "reason": "Generation kwarg synchronization bug and gather_for_metrics tuple-label truncation are unrelated training/generation issues." + "reason": "Generation kwargs update after sync_gpus and gather_for_metrics dropping tuple labels are distinct distributed-training bugs." } ] }, @@ -11127,7 +11160,7 @@ { "label": "issue", "component_index": 1, - "component_count": 25, + "component_count": 26, "review_unit_index": 36, "review_unit_count": 61, "status": "reviewed", @@ -11191,16 +11224,16 @@ "issue:43701|issue:46506", "issue:42617|issue:43575", "issue:43575|issue:43742", - "issue:42907|issue:45698", "issue:43550|issue:43701", "issue:27957|issue:32090", + "issue:42907|issue:45698", "issue:25251|issue:28282", "issue:43606|issue:43701", "issue:16998|issue:22355", "issue:43295|issue:43825", "issue:43295|issue:44871", - "issue:46234|issue:46291", "issue:44871|issue:45584", + "issue:46234|issue:46291", "issue:44464|issue:46421", "issue:44077|issue:44534", "issue:43316|issue:43716", @@ -11216,20 +11249,20 @@ "issue:44610|issue:46369", "issue:45803|issue:46170" ], - "prepared_review_unit_hash": "68645e186463a694afcaf95b62201b63df3a76d6ff9f533dc1f36e72506e14fc", + "prepared_review_unit_hash": "2d33e6266108702e537718d3ce2d8a1ae0cd74894cae45eb359a4b78494afb34", "component_budget": { "node_count": 505, "item_count": 505, "soft_pair_count": 1854, - "serialized_chars": 962354, - "estimated_input_tokens": 240589, - "estimated_eval_tokens": 481434 + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 28, - "serialized_chars": 35854, + "serialized_chars": 35855, "estimated_input_tokens": 8964, "estimated_eval_tokens": 18184 }, @@ -11238,180 +11271,180 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is broad and mostly heterogeneous: loading/config/import regressions, compile/runtime bugs, and a few multimodal model issues. I don\u2019t see any soft pair that clearly describes the same underlying bug or fix. Issue 45698 is the best anchor because it is the broadest custom-module loading regression and has the most discussion/inbound references among the set.", - "confidence": 0.19, - "canonical_issue_reason": "No clear duplicate stands out in this mixed cluster; issue 45698 is the most central/high-signal loading regression and the best representative anchor.", + "summary": "This cluster is a set of mostly unrelated issues that only look superficially similar by shared terms like loading, configs, cache, generation, or performance. None of the soft pairs appear to describe the same concrete bug/change, so all soft edges should be rejected.", + "confidence": 0.96, + "canonical_issue_reason": "issue:45698 is the strongest representative issue: it has the most discussion and inbound references, and it describes a concrete, reproducible save_pretrained/from_pretrained custom-module loading failure.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 45698 has the strongest combination of breadth and activity, and it sits closest to the recurring save/load/custom-module theme present in parts of the cluster.", + "best_issue_reason": "issue:45698 is also the best overall issue candidate because it is well-scoped, high-signal, and centrally about a user-facing loading regression.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43761", "right": "issue:44188", "accept": false, - "reason": "Both are regressions, but they hit different code paths: CLIPVision hidden states vs torch.compile attention branching." + "reason": "Both mention regressions around forward/attention behavior, but one is CLIPVisionModel hidden_states being None and the other is torch.compile branching causing diverging attention kernels; different code paths and failure modes." }, { "left": "issue:43701", "right": "issue:46506", "accept": false, - "reason": "Checkpoint key mismatch and CPU/CUDA output divergence are different failure modes with different triggers." + "reason": "resume_from_checkpoint key mismatch is a checkpoint/state dict naming issue, while PPDocLayoutV3 CPU vs CUDA differences are numerical/device divergence; not the same bug." }, { "left": "issue:42617", "right": "issue:43575", "accept": false, - "reason": "Different problems: a script execution failure vs tp load-time OOM on a specific model." + "reason": "3d_parallel.py failing is a script/runtime problem, while Qwen2 tp causing OOM is a model-loading memory footprint issue; they are not the same underlying defect." }, { "left": "issue:43575", "right": "issue:43742", "accept": false, - "reason": "OOM under tensor-parallel loading is not the same bug as a key error while loading MobileLLM." - }, - { - "left": "issue:42907", - "right": "issue:45698", - "accept": false, - "reason": "Both involve save/load, but one is dequantized model saving and the other is wrong custom module resolution after save_pretrained." + "reason": "Both involve loading models, but one is tensor-parallel OOM for Qwen2-57B-A14B-Instruct and the other is a key error for MobileLLM-125M; unrelated failures." }, { "left": "issue:43550", "right": "issue:43701", "accept": false, - "reason": "torch.compile SDPA behavior on Bamba is unrelated to resume_from_checkpoint key mismatches." + "reason": "torch.compile with SDPA on Bamba-9B-v2 is an execution/graphing issue, not the same as a checkpoint key mismatch on resume_from_checkpoint." }, { "left": "issue:27957", "right": "issue:32090", "accept": false, - "reason": "FlashAttention support for XLMRoberta and a Trainer broadcast TypeError are unrelated bugs." + "reason": "XLMRoberta Flash Attention 2 support and Trainer GPU broadcast NoneType errors are different subsystems and distinct failure modes." + }, + { + "left": "issue:42907", + "right": "issue:45698", + "accept": false, + "reason": "Saving dequantized Ministral/Devstral models is a quantization-save issue, while 45698 is wrong custom module resolution after save_pretrained; different bugs." }, { "left": "issue:25251", "right": "issue:28282", "accept": false, - "reason": "Pipeline top_k output nesting and missing PyTorch import errors are unrelated." + "reason": "top_k changing pipeline output shape is an API/return-format issue; AutoModel missing PyTorch is an environment import error, not a duplicate." }, { "left": "issue:43606", "right": "issue:43701", "accept": false, - "reason": "CPU offload device mismatch is a different runtime issue from checkpoint key mismatches." + "reason": "CPU offload device mismatch in bark-small is a device placement bug, not the same as checkpoint key mismatch when resuming training." }, { "left": "issue:16998", "right": "issue:22355", "accept": false, - "reason": "A model_max_length question and an onnx import error are unrelated." + "reason": "model_max_length behavior for DeBERTa-V3 and missing transformers.onnx are unrelated configuration vs import/package issues." }, { "left": "issue:43295", "right": "issue:43825", "accept": false, - "reason": "Both are v5-related, but one is processor/tokenizer regression and the other is an incorrect pipeline error message." + "reason": "processor.tokenizer/image-to-tokenizer regression in v4.57.5 and a misleading pipeline translation error message are different regressions in different code paths." }, { "left": "issue:43295", "right": "issue:44871", "accept": false, - "reason": "Processor/tokenizer handling and eos_token_id config mismatch are different model-interface bugs." + "reason": "Both involve multimodal model config inconsistencies, but one is processor.tokenizer/image passing regression and the other is eos_token_id mismatch; different concrete issues." }, { - "left": "issue:46234", - "right": "issue:46291", + "left": "issue:44871", + "right": "issue:45584", "accept": false, - "reason": "Different import failures: missing utility symbol vs kernels dependency/version handling." + "reason": "Gemma-3 eos_token_id mismatch and Whisper empty transcription after align_special_tokens are separate tokenizer/generation bugs." }, { - "left": "issue:44871", - "right": "issue:45584", + "left": "issue:46234", + "right": "issue:46291", "accept": false, - "reason": "Gemma eos_token_id configuration and Whisper empty-transcription handling are unrelated." + "reason": "Missing is_falcon_mamba_ssm_available export and kernels v0.15 requiring version/revision are both import-time issues, but they involve different packages and root causes." }, { "left": "issue:44464", "right": "issue:46421", "accept": false, - "reason": "Both mention chunked generation/cache behavior, but one is output inconsistency and the other is spurious recompilation from lazy cache init." + "reason": "Chunked generation inconsistency with compiled forward and cache lazy initialization causing recompilation are both compile/cache-related, but not the same behavior or fix." }, { "left": "issue:44077", "right": "issue:44534", "accept": false, - "reason": "Different issues: PatchTSMixer post_init validation vs non-persistent buffers being filled with junk." + "reason": "patchtsmixer post_init allowance and non-persistent buffers filled with junk are unrelated model-initialization/config regressions." }, { "left": "issue:43316", "right": "issue:43716", "accept": false, - "reason": "Gemma3TextConfig API inconsistency and Mistral-3 image preprocessor dtype mismatch are not the same bug." + "reason": "Gemma3TextConfig API discrepancy and Mistral-3 image preprocessor dtype mismatch are different model-family issues with different symptoms." }, { "left": "issue:43316", "right": "issue:46821", "accept": false, - "reason": "These are separate model/config problems with different affected components and symptoms." + "reason": "Gemma3TextConfig API mismatch and TimesFM window_size AttributeError are unrelated config/API bugs." }, { "left": "issue:44871", "right": "issue:46050", "accept": false, - "reason": "EOS token config mismatch is unrelated to quantized save/load breakage." + "reason": "Gemma-3 eos_token_id inconsistency and sinq quantized save/load breakage are different serialization/config problems." }, { "left": "issue:43646", "right": "issue:44913", "accept": false, - "reason": "Custom model initialization breakage and GPTNeoX rotary_pct reload regression are different config/loading issues." + "reason": "Transformers 5.0.0 breaking custom model initialization and GPTNeoX rotary_pct not persisting are distinct config-loading regressions." }, { "left": "issue:43452", "right": "issue:44496", "accept": false, - "reason": "Both are loading failures, but gguf handling and missing model_type in config have different root causes." + "reason": "gguf_file breaking AutoTokenizer/AutoModel loading and an unrecognized model/config.json error are different model-loading failures." }, { "left": "issue:45803", "right": "issue:46365", "accept": false, - "reason": "Bare except swallowing errors in Fuyu conversion is unrelated to the Trainer import problem with datasets." + "reason": "FuyuBatchFeature swallowing KeyboardInterrupt is a code-path exception-handling bug, while Trainer import failing until datasets is uninstalled is an environment/package conflict." }, { "left": "issue:44162", "right": "issue:44792", "accept": false, - "reason": "ESM2 breakage and a janus image-generation test failure are unrelated." + "reason": "ESM2 being broken and janus test_model_generate_images failure are unrelated model/test failures." }, { "left": "issue:43404", "right": "issue:43441", "accept": false, - "reason": "Same general model family, but weight tying and FlashAttention failure are distinct concrete bugs." + "reason": "Mistral3 lm_head weight tying affects image-text generation behavior; Ministral-3 FlashAttention failure is an attention-kernel compatibility issue." }, { "left": "issue:45632", "right": "issue:46424", "accept": false, - "reason": "Cache path collisions for trust_remote_code and max_cache_length being ignored are different cache-system bugs." + "reason": "trust_remote_code cache path collisions and max_cache_length only honored for Quantized Cache are both cache-related, but one is a path collision bug and the other is a config-option propagation bug." }, { "left": "issue:44242", "right": "issue:46894", "accept": false, - "reason": "MoE load-balancing loss omission and CohereASR training-loss double shift are unrelated training-loss issues." + "reason": "Load balancing loss not added when output_router_logits=False and CohereASR training-loss double-shift are both training-loss bugs, but they affect different mechanisms and fixes." }, { "left": "issue:44610", "right": "issue:46369", "accept": false, - "reason": "Processor input-size mismatch and model memory overflow are not the same defect." + "reason": "OmDet-Turbo processor producing 640px inputs vs model expecting 224px is an image-size preprocessing mismatch; sam3-litetext memory overflow is a resource usage issue." }, { "left": "issue:45803", "right": "issue:46170", "accept": false, - "reason": "A swallowed-exception bug and a Windows ROCm import dependency issue are unrelated." + "reason": "Bare except swallowing errors in FuyuBatchFeature is unrelated to Windows/ROCm import errors caused by torch.distributed dependency." } ] }, @@ -11425,7 +11458,7 @@ { "label": "issue", "component_index": 1, - "component_count": 25, + "component_count": 26, "review_unit_index": 37, "review_unit_count": 61, "status": "reviewed", @@ -11468,6 +11501,7 @@ "issue:45245", "issue:45406", "issue:45542", + "issue:45584", "issue:45588", "issue:45600", "issue:45632", @@ -11486,8 +11520,8 @@ "soft_pairs": [ "issue:43065|issue:44617", "issue:43540|issue:44514", - "issue:43650|issue:46202", "issue:45542|issue:46234", + "issue:43650|issue:46202", "issue:43504|issue:45698", "issue:43572|issue:46178", "issue:44792|issue:44977", @@ -11497,204 +11531,214 @@ "issue:43673|issue:44743", "issue:43352|issue:46640", "issue:43525|issue:43531", - "issue:41950|issue:43475", "issue:44534|issue:44913", + "issue:41950|issue:43475", "issue:44190|issue:44964", "issue:35141|issue:42490", "issue:44734|issue:45406", "issue:45072|issue:46932", "issue:45632|issue:46617", "issue:44869|issue:46832", - "issue:45245|issue:46858", "issue:44855|issue:45588", - "issue:46082|issue:46697" + "issue:46082|issue:46697", + "issue:45245|issue:46858", + "issue:44617|issue:45584" ], - "prepared_review_unit_hash": "9ca6689aa69e5e689c272e545337a7145fb20d11db62b8e13af145afbf2f7d73", + "prepared_review_unit_hash": "19c2297a7d59ead90c136766bc48909ea63b1f7986d9052b87d864885bfc63a1", "component_budget": { "node_count": 505, "item_count": 505, "soft_pair_count": 1854, - "serialized_chars": 962354, - "estimated_input_tokens": 240589, - "estimated_eval_tokens": 481434 + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 }, "budget": { - "node_count": 47, - "item_count": 47, - "soft_pair_count": 24, - "serialized_chars": 34395, - "estimated_input_tokens": 8599, - "estimated_eval_tokens": 17454 + "node_count": 48, + "item_count": 48, + "soft_pair_count": 25, + "serialized_chars": 35271, + "estimated_input_tokens": 8818, + "estimated_eval_tokens": 17892 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The cluster is highly heterogeneous; the soft pairs show only superficial subsystem overlap, not the same underlying bug/change. No soft edge should be merged. If one issue must be chosen as a representative, #45698 is the strongest core-loading bug with the most discussion and inbound references, but it is not a true duplicate hub for the rest of the set.", - "confidence": 0.94, - "canonical_issue_reason": "#45698 has the strongest signal and broadest impact (core from_pretrained/save_pretrained loading path, highest discussion/inbound refs), so it is the best single representative if a canonical issue is required.", + "summary": "The cluster is mostly heterogeneous, with one plausible duplicate thread around `trust_remote_code` cache-path handling. Most other pairs are only loosely related by subsystem and should stay separate.", + "confidence": 0.87, + "canonical_issue_reason": "`issue:45632` is the clearest root-cause description in the only plausible duplicate thread: cache-path resolution for `trust_remote_code` models. It best represents the shared bug surface with `issue:46617`.", "canonical_pr_reason": null, - "best_issue_reason": "#45698 is the most actionable and well-supported issue in the set; it is a concrete, high-impact bug with clear reproduction and broad user relevance.", + "best_issue_reason": "`issue:45632` is the strongest global issue anchor because it is the most specific, root-cause-oriented cache-path report and has the clearest duplicate counterpart in this set.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43065", "right": "issue:44617", "accept": false, - "reason": "Both are SAM3-related, but one is a dummy Conv2d in a decoder and the other is CUDA OOM in video inference; different failures and code paths." + "reason": "Different bugs: a dummy conv layer in Sam3PixelDecoder vs a CUDA OOM in Sam3Video." }, { "left": "issue:43540", "right": "issue:44514", "accept": false, - "reason": "Both mention video/Qwen models, but one is a processing ValueError and the other is a batched chat-template crash with padding=False; unrelated bugs." + "reason": "Both involve video/multimodal flows, but one is a Qwen3OmniMoe video-input validation error and the other is batched chat-template padding behavior." }, { - "left": "issue:43650", - "right": "issue:46202", + "left": "issue:45542", + "right": "issue:46234", "accept": false, - "reason": "No shared concrete bug: one is a placeholder/data issue, the other is a Gemma pipeline roundtrip tokenization problem." + "reason": "Unrelated import/backends problems: TensorFlow/tensorboard backend detection vs a missing `is_falcon_mamba_ssm_available` symbol." }, { - "left": "issue:45542", - "right": "issue:46234", + "left": "issue:43650", + "right": "issue:46202", "accept": false, - "reason": "Both touch backend/import utilities, but one is a TensorBoard-without-TF backend error and the other is a missing import_utils symbol; different defects." + "reason": "No shared underlying bug; one is a vague data request and the other is a Gemma token roundtrip issue." }, { "left": "issue:43504", "right": "issue:45698", "accept": false, - "reason": "One is a BEiT legacy-field model preset load failure; the other is wrong custom module selection after save/load. Similar area, not same bug." + "reason": "Both concern model loading, but the concrete failures differ: legacy field handling vs wrong custom module selection after save/load." }, { "left": "issue:43572", "right": "issue:46178", "accept": false, - "reason": "These are unrelated configuration/backend issues: StableLm pad_token_idx after v5.0 vs a tensorflow_text backend mapping error." + "reason": "Different problems: a missing config field after v5 update vs an unsupported backend entry (`tensorflow_text`)." }, { "left": "issue:44792", "right": "issue:44977", "accept": false, - "reason": "Both involve generation on model families, but one is a Janus image test failure and the other is Qwen3.5 flash-attention behavior; not the same path." + "reason": "Different models and failure modes: janus image-generation test failure vs Qwen3.5 flash-attention generation bug." }, { "left": "issue:44792", "right": "issue:45600", "accept": false, - "reason": "Janus image generation failure and Sam3LiteText auto_mapping CI breakage are different models and different root causes." + "reason": "CI/test failure versus auto-mapping references to removed configs; not the same underlying change or bug." }, { "left": "issue:45127", "right": "issue:45200", "accept": false, - "reason": "Both mention Gemma-family fine-tuning, but LoRA merge/collapse with extended vocab is unrelated to missing mm_token_type_ids defaults." + "reason": "Both touch model setup, but one is LoRA/vocab collapse and the other is a Gemma 4 multimodal token-id defaulting issue." }, { "left": "issue:43388", "right": "issue:43688", "accept": false, - "reason": "One is gather_for_metrics truncating tuple labels; the other is auxiliary-loss normalization in MoE models. Different training bugs." + "reason": "Different training bugs: metric gathering drops labels vs auxiliary-loss normalization." }, { "left": "issue:43673", "right": "issue:44743", "accept": false, - "reason": "Both concern cache/state handling in generation, but one is missing cache in chunked prefill and the other is recurrent states resetting for seq_len>1; not the same concrete bug." + "reason": "Both mention cache/state handling, but the concrete failures and code paths are different enough that they are not the same bug." }, { "left": "issue:43352", "right": "issue:46640", "accept": false, - "reason": "Both mention Flash Attention 2, but one is NemotronH unsupported-FA2 gating and the other is a generic FA2 issue with different context." + "reason": "Too broad to merge: one is a Nemotron-specific FA2 support error, the other is a generic flash-attention issue with no concrete shared code path." }, { "left": "issue:43525", "right": "issue:43531", "accept": false, - "reason": "Llama4Config pad_token_id missing and Qwen3-MoE sliding_window handling are unrelated configuration/model bugs." + "reason": "Different config/model bugs: missing `pad_token_id` in Llama4Config vs sliding-window behavior in Qwen3-MoE." }, { - "left": "issue:41950", - "right": "issue:43475", + "left": "issue:44534", + "right": "issue:44913", "accept": false, - "reason": "Video-classification processor selection and SAM3 vision encoder missing attribute are separate issues in different pipelines." + "reason": "Both are configuration/state related, but one is about non-persistent buffers being initialized with junk and the other about `rotary_pct` not persisting on reload." }, { - "left": "issue:44534", - "right": "issue:44913", + "left": "issue:41950", + "right": "issue:43475", "accept": false, - "reason": "One is non-persistent buffers filled with junk in v5, the other is GPTNeoXConfig rotary_pct not persisting on reload; similar persistence theme only." + "reason": "Related only by processor/vision area; one is wrong processor selection, the other is a missing output attribute in SAM 3 video." }, { "left": "issue:44190", "right": "issue:44964", "accept": false, - "reason": "Local dataset loading in a training script and model loading of microsoft/Phi-4-multimodal-instruct are different failures." + "reason": "Different loading failures: a local dataset issue in a training script vs a model-load failure for Phi-4 multimodal." }, { "left": "issue:35141", "right": "issue:42490", "accept": false, - "reason": "Token-embedding resize reinitialization and import warning from a trailing slash are entirely different code paths." + "reason": "Unrelated: embedding reinitialization after resize vs an import warning from a trailing-slash local model path." }, { "left": "issue:44734", "right": "issue:45406", "accept": false, - "reason": "KV-cache continuation tensor indexing and Gemma4Processor missing _tokenizer are unrelated runtime errors." + "reason": "Distinct serve/multimodal crashes: KV-cache continuation tensor indexing vs missing `_tokenizer` on Gemma4Processor." }, { "left": "issue:45072", "right": "issue:46932", "accept": false, - "reason": "Both involve dtype/quantization concerns, but one is CI inference mismatches and the other is Gemma4 embedder casting; not the same bug." + "reason": "Different dtype issues in different contexts; one is a CI inference mismatch, the other is quantized Gemma4 embedder casting." }, { "left": "issue:45632", "right": "issue:46617", - "accept": false, - "reason": "Both are trust_remote_code cache-loading issues, but one is leaf-directory cache collision and the other is a FileNotFoundError from cache loading; similar surface, different root cause." + "accept": true, + "reason": "Both point to the same `trust_remote_code` cache-path bug surface: a bad cache path/collision can plausibly produce the reported `FileNotFoundError`." }, { "left": "issue:44869", "right": "issue:46832", "accept": false, - "reason": "Whisper timestamp decode crash and RT-DETR feature-level mismatch are completely unrelated." - }, - { - "left": "issue:45245", - "right": "issue:46858", - "accept": false, - "reason": "Category-cardinality overflow and static-cache slice arithmetic on non-CUDA backends are different low-level failures." + "reason": "Different model families and failures: Whisper word-timestamp decoding vs RT-DETR feature-level handling." }, { "left": "issue:44855", "right": "issue:45588", "accept": false, - "reason": "Python 3.13 import parsing/torch.jit script issue is unrelated to flash_attention.py crashing on s_aux=None." + "reason": "Python 3.13 parse/import failure is unrelated to a flash-attention crash on `s_aux=None`." }, { "left": "issue:46082", "right": "issue:46697", "accept": false, - "reason": "LlamaConfig head_dim validation and Swin2SR over-padding are different model/configuration bugs with no shared code path." + "reason": "Unrelated model issues: LlamaConfig validation vs Swin2SR over-padding." + }, + { + "left": "issue:45245", + "right": "issue:46858", + "accept": false, + "reason": "Different low-level bugs: category-count overflow vs device-tensor slice arithmetic in static cache prefill." + }, + { + "left": "issue:44617", + "right": "issue:45584", + "accept": false, + "reason": "Different SAM3 OOM vs Whisper empty-transcription failure; only the general ML domain overlaps." } ] }, - "evaluator_result": null, - "evaluator_used": false, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet and stays conservative overall. The only accepted soft edge, issue:45632 vs issue:46617, is plausibly the same trust_remote_code cache-path failure surface. The other rejections are cautious and supported by the issue titles, with no obvious overstatement." + }, + "evaluator_used": true, "retried": false, - "accepted_nontrivial_soft_edge": false, + "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null }, { "label": "issue", "component_index": 1, - "component_count": 25, + "component_count": 26, "review_unit_index": 38, "review_unit_count": 61, "status": "reviewed", @@ -11755,9 +11799,8 @@ ], "soft_pairs": [ "issue:43827|issue:43994", - "issue:44617|issue:45584", - "issue:43673|issue:46421", "issue:29870|issue:35141", + "issue:43673|issue:46421", "issue:43931|issue:44060", "issue:43644|issue:44913", "issue:44242|issue:46897", @@ -11766,8 +11809,8 @@ "issue:44933|issue:45362", "issue:43335|issue:44855", "issue:44077|issue:44877", - "issue:43550|issue:44190", "issue:45381|issue:46369", + "issue:43550|issue:44190", "issue:41720|issue:45245", "issue:44479|issue:44734", "issue:43065|issue:44610", @@ -11779,220 +11822,217 @@ "issue:45127|issue:45357", "issue:45381|issue:46655", "issue:43295|issue:45276", - "issue:44779|issue:46489", "issue:44112|issue:45600", + "issue:44779|issue:46489", "issue:45290|issue:46752" ], - "prepared_review_unit_hash": "672de7239c903ebd8f0aec4d77159df44d664ae241fca48107252e69528b3ff1", + "prepared_review_unit_hash": "a977f79ee9dfb258c11fa6b34afd4802c1a8d9e1aeba3196b2215d1de12fc7e4", "component_budget": { "node_count": 505, "item_count": 505, "soft_pair_count": 1854, - "serialized_chars": 962354, - "estimated_input_tokens": 240589, - "estimated_eval_tokens": 481434 + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 28, - "serialized_chars": 36223, - "estimated_input_tokens": 9056, - "estimated_eval_tokens": 18368 + "soft_pair_count": 27, + "serialized_chars": 35923, + "estimated_input_tokens": 8981, + "estimated_eval_tokens": 18218 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is heterogeneous: it mixes unrelated bugs, regressions, docs issues, CI failures, and model-specific edge cases across many different code paths. None of the soft pairs looks like the same underlying bug or change, so there is no real duplicate cluster here.", - "confidence": 0.99, - "canonical_issue_reason": "No single canonical issue fits this cluster; the issues are mostly unrelated and span different models, processors, generation paths, and configuration behavior.", - "canonical_pr_reason": "No pull requests are present.", - "best_issue_reason": "No issue is a good representative duplicate target because the set is not a coherent duplicate group.", - "best_pr_reason": "No pull requests are present.", + "summary": "Most pairs are thematically related but not the same underlying bug. The only strong duplicate-level match is the DeepSeek tokenizer regression pair; the rest differ in concrete failure mode, model, or code path.", + "confidence": 0.87, + "canonical_issue_reason": "issue:44779 is the broadest/root report for the DeepSeek tokenizer regression in v5, and issue:46489 reads like a narrower follow-up that explicitly calls out a gap in the prior fix.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:44779 is the best anchor because it states the regression plainly, is broader in scope, and can subsume the later DeepSeek-Coder v1 follow-up.", + "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43827", "right": "issue:43994", "accept": false, - "reason": "Docs still referencing pipeline() after v5 removals is unrelated to SigLIP2 producing bad outputs via AutoModel/pipeline." + "reason": "Both mention pipelines, but one is a docs/API-removal problem and the other is a model/output correctness issue for SigLIP2; different bugs." }, { - "left": "issue:44617", - "right": "issue:45584", + "left": "issue:29870", + "right": "issue:35141", "accept": false, - "reason": "Sam3Video CUDA OOM and Whisper empty-transcription failure are different models and different failure modes." + "reason": "One is a BART generate() decoder mask bug, the other is a resize_token_embeddings/post_init weight reinit bug; not the same code path." }, { "left": "issue:43673", "right": "issue:46421", "accept": false, - "reason": "Both mention chunked prefill/cache, but one is a missing GenerationMixin cache in v5.0.0 and the other is spurious recompilation from lazy cache init." - }, - { - "left": "issue:29870", - "right": "issue:35141", - "accept": false, - "reason": "Bart generate() decoder_attention_mask initialization bug is unrelated to output embeddings being reinitialized after resize_token_embeddings." + "reason": "Both involve chunked prefill/cache behavior, but one is missing cache state and the other is spurious recompilation from lazy initialization; adjacent, not duplicate." }, { "left": "issue:43931", "right": "issue:44060", "accept": false, - "reason": "Qwen3-VL weight shape mismatch on load is not the same as the Qwen3-Next tied-weights warning involving dt_bias." + "reason": "Weight-shape mismatch on Qwen3-VL loading is a different failure from the Qwen3-Next tied-weights warning bug." }, { "left": "issue:43644", "right": "issue:44913", "accept": false, - "reason": "Non-persistent buffer junk filling in Transformers 5.0.0 is unrelated to GPTNeoX rotary_pct not persisting on reload." + "reason": "A generic non-persistent-buffer junk-fill regression is not the same as GPTNeoXConfig's rotary_pct reload defaulting." }, { "left": "issue:44242", "right": "issue:46897", "accept": false, - "reason": "Missing load-balancing loss and Florence2 double-shift training loss are different bugs in different models/code paths." + "reason": "These are both training-loss issues, but they affect different models and different loss computation paths." }, { "left": "issue:25251", "right": "issue:27957", "accept": false, - "reason": "Pipeline top_k nesting change and XLMRoberta Flash Attention 2 support are unrelated issues." + "reason": "Top_k nested-list output from pipeline is unrelated to XLMRoberta Flash Attention 2 support." }, { "left": "issue:45563", "right": "issue:45584", "accept": false, - "reason": "Paged generate warning about num_return_sequences is unrelated to Whisper failing on empty transcription after align_special_tokens." + "reason": "A stale warning in paged generate() is unrelated to Whisper empty-transcription failure after align_special_tokens." }, { "left": "issue:44933", "right": "issue:45362", "accept": false, - "reason": "A nonexistent import from image_utils has nothing to do with Qwen3.5 chat crashes." + "reason": "A nonexistent image_utils import is a different issue from Qwen3.5 chat crashes." }, { "left": "issue:43335", "right": "issue:44855", "accept": false, - "reason": "SwitchTransformers sparse-layer creation and DebertaV2Model Python 3.13 import parsing are unrelated." + "reason": "Sparse-layer config construction and Python 3.13 IndentationError on import are unrelated bugs." }, { "left": "issue:44077", "right": "issue:44877", "accept": false, - "reason": "patchtsmixer post_init policy and granite_speech strict config loading are different configuration issues." + "reason": "Optional post_init handling in patchtsmixer is not the same as strict-config loading for granite_speech." }, { - "left": "issue:43550", - "right": "issue:44190", + "left": "issue:45381", + "right": "issue:46369", "accept": false, - "reason": "Bamba torch.compile/SDPA failure is unrelated to the local dataset loading script bug." + "reason": "Qwen video vision_position_ids regression and Sam3-LiteText memory overflow are different code paths." }, { - "left": "issue:45381", - "right": "issue:46369", + "left": "issue:43550", + "right": "issue:44190", "accept": false, - "reason": "Qwen2.5-VL video vision_position_ids regression and Sam3-LiteText memory overflow are different model-specific problems." + "reason": "torch.compile/SDPA model failure and local dataset loading failure are unrelated." }, { "left": "issue:41720", "right": "issue:45245", "accept": false, - "reason": "Qwen3 auto device mapping cudaErrorAssert and the 2^24 categories limit are unrelated." + "reason": "Qwen3 device-mapping cudaErrorAssert has nothing in common with the 2^24 category limit error." }, { "left": "issue:44479", "right": "issue:44734", "accept": false, - "reason": "Qwen video-input regression and /v1/responses KV-cache continuation indexing are different code paths." + "reason": "Video-input regression across Qwen models is different from KV-cache continuation tensor indexing in transformers serve." }, { "left": "issue:43065", "right": "issue:44610", "accept": false, - "reason": "Dummy Conv2d in Sam3PixelDecoder is unrelated to OmDet-Turbo's processor/model input-size mismatch." + "reason": "A dummy Conv2d in Sam3PixelDecoder is unrelated to OmDet-Turbo's processor/model image-size mismatch." }, { "left": "issue:43335", "right": "issue:44190", "accept": false, - "reason": "This repeats the earlier unrelated pair: SwitchTransformers config bug vs image-classification dataset loading." + "reason": "Config sparsity construction and local dataset loading are unrelated failures." }, { "left": "issue:45588", "right": "issue:47013", "accept": false, - "reason": "Flash-attention sink-less model crash and xLSTM unpacking failure are unrelated implementation bugs." + "reason": "Flash-attention sinkless-model AttributeError and xLSTM unpacking crash are different implementation bugs." }, { "left": "issue:44617", "right": "issue:46369", "accept": false, - "reason": "Both are memory-related, but Sam3Video OOM and Sam3-LiteText overflow are different model paths and not the same bug." + "reason": "Sam3Video OOM and Sam3-LiteText memory overflow are both memory-related but not the same bug." }, { "left": "issue:39692", "right": "issue:44190", "accept": false, - "reason": "SigLIP2 docs/example errors are unrelated to the image-classification no-trainer local dataset loading issue." + "reason": "SigLIP2 docs/model mismatch is unrelated to local dataset loading in the image-classification script." }, { "left": "issue:44464", "right": "issue:45563", "accept": false, - "reason": "Chunked-generation inconsistency with compiled forward is not the same as a stale warning in paged generate()." + "reason": "Chunked generation inconsistency under compiled forward is not the same as a stale warning on num_return_sequences." }, { "left": "issue:45127", "right": "issue:45357", "accept": false, - "reason": "LoRA merge collapse with extended vocab is unrelated to Qwen3.5 save_pretrained writing incorrect visual encoder keys." + "reason": "LoRA merge collapse with extended vocab and Qwen3.5 visual-encoder save_pretrained regression are different model-state bugs." }, { "left": "issue:45381", "right": "issue:46655", "accept": false, - "reason": "Both involve video processing, but Qwen2.5-VL vision_position_ids and SmolVLM padding calculation are different bugs." + "reason": "Qwen video vision-position regression and SmolVLM video padding calculation are different processor issues." }, { "left": "issue:43295", "right": "issue:45276", "accept": false, - "reason": "processor.tokenizer regression and gemma4 resize_token_embeddings not updating embeddings are separate issues." - }, - { - "left": "issue:44779", - "right": "issue:46489", - "accept": false, - "reason": "Deepseek tokenizer regression and the DeepSeek-Coder v1 gap in a prior fix are related at a high level but not clearly the same concrete bug." + "reason": "Processor/tokenizer API regression and gemma4 resize_token_embeddings not updating outputs are different bugs." }, { "left": "issue:44112", "right": "issue:45600", "accept": false, - "reason": "GraniteSpeech stale device-override CI failure is unrelated to removed Sam3LiteText configs in auto_mappings." + "reason": "GraniteSpeech CI stale device override and removed Sam3LiteText mappings are separate CI breakages." + }, + { + "left": "issue:44779", + "right": "issue:46489", + "accept": true, + "reason": "Both report the same DeepSeek tokenizer regression in v5; 46489 explicitly describes a remaining gap in the prior fix." }, { "left": "issue:45290", "right": "issue:46752", "accept": false, - "reason": "apply_chat_template crashing on tool-call assistant messages is a different edge case from crashing on an empty list." + "reason": "Assistant tool-call empty-content chat-template crash and empty-list chat-template IndexError are related only at a high level; not the same specific failure." } ] }, - "evaluator_result": null, - "evaluator_used": false, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet, and the soft-edge calls stay conservative. The only accepted duplicate-level pair, issue:44779 and issue:46489, is well-supported by the titles and the note that 46489 is a gap in the prior fix. No other pair is overstated as the same underlying bug." + }, + "evaluator_used": true, "retried": false, - "accepted_nontrivial_soft_edge": false, + "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null }, { "label": "issue", "component_index": 1, - "component_count": 25, + "component_count": 26, "review_unit_index": 39, "review_unit_count": 61, "status": "reviewed", @@ -12017,6 +12057,7 @@ "issue:43526", "issue:43531", "issue:43540", + "issue:43582", "issue:43643", "issue:43673", "issue:43854", @@ -12045,10 +12086,10 @@ "issue:46082", "issue:46121", "issue:46234", - "issue:46424", "issue:46506", "issue:46697", - "issue:46858" + "issue:46858", + "issue:46932" ], "soft_pairs": [ "issue:43540|issue:44734", @@ -12059,8 +12100,8 @@ "issue:44462|issue:45907", "issue:25251|issue:29870", "issue:44938|issue:45542", - "issue:43673|issue:46858", "issue:43854|issue:44964", + "issue:43673|issue:46858", "issue:44246|issue:45003", "issue:43643|issue:44496", "issue:45632|issue:46234", @@ -12075,200 +12116,200 @@ "issue:43404|issue:45216", "issue:36246|issue:38617", "issue:43531|issue:45600", - "issue:41950|issue:44016", "issue:44295|issue:46506", - "issue:43673|issue:46424" + "issue:41950|issue:44016", + "issue:43582|issue:46932" ], - "prepared_review_unit_hash": "a32eb65351470af3dbe50ec3becbb22e84cd0cc3da2e5d889f4e3773f6e82b87", + "prepared_review_unit_hash": "0dd7132dc513c54371cc3b590beecc5610fc016d68419c5b7052a1b9df7c5a95", "component_budget": { "node_count": 505, "item_count": 505, "soft_pair_count": 1854, - "serialized_chars": 962354, - "estimated_input_tokens": 240589, - "estimated_eval_tokens": 481434 + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 }, "budget": { - "node_count": 47, - "item_count": 47, + "node_count": 48, + "item_count": 48, "soft_pair_count": 27, - "serialized_chars": 35507, - "estimated_input_tokens": 8877, - "estimated_eval_tokens": 18010 + "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": "These items are mostly unrelated Transformers issue reports rather than a tight duplicate cluster. The soft pairs share broad keywords/subsystems (generation, loaders, processors, cache, imports) but not the same underlying bug, so none should be merged.", - "confidence": 0.81, - "canonical_issue_reason": "There is no strong single canonical duplicate here; if one must be chosen, issue 44964 is the cleanest representative because it is a concrete, user-facing model-load regression with a specific target and good discussion.", + "summary": "No soft-edge pair looks like the same underlying bug/change. The candidates span unrelated areas (generation, model loading, tokenizers, caches, CI, pipelines, and docs), so there is no clear duplicate cluster here.", + "confidence": 0.97, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "Issue 44964 is the best representative issue: it is narrowly scoped, reproducible, and clearly about a real runtime failure, whereas most other issues are either more peripheral, broader, or about different code paths.", + "best_issue_reason": null, "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43540", "right": "issue:44734", "accept": false, - "reason": "Different bugs and code paths: Qwen3OmniMoe video-input processing vs Transformers serve KV-cache continuation indexing." + "reason": "Different failures: Qwen3OmniMoe video-input processing vs transformers serve KV-cache continuation indexing." }, { "left": "issue:43526", "right": "issue:46697", "accept": false, - "reason": "Both are image-processor bugs, but one is about label reduction and the other about padding logic; not the same defect." + "reason": "Different image-processor bugs on different models: label reduction vs over-padding." }, { "left": "issue:45276", "right": "issue:45584", "accept": false, - "reason": "Unrelated model/behavior: Gemma4 embedding resize/tie-weight handling vs Whisper empty-transcription generation failure." + "reason": "Unrelated model-specific embedding resize bug vs Whisper empty-transcription generation failure." }, { "left": "issue:44521", "right": "issue:45381", "accept": false, - "reason": "Both involve multimodal chat/generation, but one is assistant mask computation and the other is video position ids; distinct bugs." + "reason": "Different multimodal issues with different symptoms and code paths: assistant masks vs vision_position_ids." }, { "left": "issue:44737", "right": "issue:45803", "accept": false, - "reason": "CPU tensor allocation in XLNet positional encoding is unrelated to Fuyu's broad exception handling bug." + "reason": "XLNet CPU tensor allocation bug is unrelated to Fuyu's broad exception handling." }, { "left": "issue:44462", "right": "issue:45907", "accept": false, - "reason": "Tokenizer file selection/precedence and repo listing revision forwarding are different failure modes." + "reason": "Tokenizer file selection bug is unrelated to list_repo_files revision forwarding." }, { "left": "issue:25251", "right": "issue:29870", "accept": false, - "reason": "Pipeline top_k nesting and Bart generate() UnboundLocalError are separate generation issues." + "reason": "Pipeline output nesting bug is unrelated to Bart decoder_attention_mask initialization." }, { "left": "issue:44938", "right": "issue:45542", "accept": false, - "reason": "Python 3.14 import failure is not the same as the TensorFlow-backend packaging mismatch." + "reason": "Python 3.14 import failure is unrelated to missing TensorFlow backend dependency handling." }, { - "left": "issue:43673", - "right": "issue:46858", + "left": "issue:43854", + "right": "issue:44964", "accept": false, - "reason": "Both concern cache internals, but one is missing cache in chunked_prefill and the other is device-tensor seq_length arithmetic in static cache." + "reason": "Different model-loading regressions affecting different models and likely different failing code paths." }, { - "left": "issue:43854", - "right": "issue:44964", + "left": "issue:43673", + "right": "issue:46858", "accept": false, - "reason": "Both are model-loading complaints, but they target different models and different underlying loading failures." + "reason": "Both cache-related, but one is a missing cache API in v5 chunked prefill and the other is device-tensor arithmetic in static cache." }, { "left": "issue:44246", "right": "issue:45003", "accept": false, - "reason": "Import slowness and unsafe sys.modules access are related only at a very high level, not the same bug." + "reason": "Import latency complaint is not the same bug as unsafe sys.modules access." }, { "left": "issue:43643", "right": "issue:44496", "accept": false, - "reason": "AutoConfig trust_remote_code field-loss and unrecognized-model/model_type handling are different config-loading problems." + "reason": "Different config-loading problems: missing fields with trust_remote_code vs missing model_type in config.json." }, { "left": "issue:45632", "right": "issue:46234", "accept": false, - "reason": "Cache-path collision for trust_remote_code is unrelated to the missing import utility symbol." + "reason": "Cache-path collision for local models is unrelated to missing import_utils availability helper." }, { "left": "issue:44964", "right": "issue:45600", "accept": false, - "reason": "Different model-loading failures: Phi-4 multimodal load regression vs stale auto_mappings references breaking CI." + "reason": "Different failures: model loading regression vs stale auto_mappings reference breaking CI." }, { "left": "issue:42175", "right": "issue:44279", "accept": false, - "reason": "Packaging/extras mismatch is not the same as a generic dependency issue." + "reason": "General dependency/install mismatch is too broad and does not match the specific TensorFlow-backend packaging bug." }, { "left": "issue:27957", "right": "issue:35141", "accept": false, - "reason": "Flash Attention support for XLMRoberta is unrelated to embedding reinitialization after resize_token_embeddings." + "reason": "Flash Attention support issue is unrelated to embedding reinitialization after resize_token_embeddings." }, { "left": "issue:44112", "right": "issue:44964", "accept": false, - "reason": "A stale CI test failure in GraniteSpeech is not the same as a model-loading regression for Phi-4 multimodal." + "reason": "CI test flake in GraniteSpeech is unrelated to Phi-4 multimodal loading." }, { "left": "issue:42796", "right": "issue:44521", "accept": false, - "reason": "Tokenizer BPE marker handling and multimodal assistant mask generation are different subsystems and bugs." + "reason": "Tokenizer BPE-marker handling is unrelated to multimodal chat-template assistant masks." }, { "left": "issue:36246", "right": "issue:41628", "accept": false, - "reason": "Missing model-specific processor import and missing top-level AutoImageProcessor import are distinct import/export issues." + "reason": "Specific missing class in qwen2_5_vl is not the same as the top-level AutoImageProcessor import failure." }, { "left": "issue:46082", "right": "issue:46121", "accept": false, - "reason": "Llama head_dim validation and rope-parameter JSON list handling are separate config-validation errors." + "reason": "Llama config head_dim validation is unrelated to rope parameter list typing." }, { "left": "issue:32090", "right": "issue:35141", "accept": false, - "reason": "Trainer GPU broadcast NoneType error is unrelated to token-embedding resizing and weight reinitialization." + "reason": "Trainer broadcast TypeError is unrelated to output embedding reinitialization after resize." }, { "left": "issue:43404", "right": "issue:45216", "accept": false, - "reason": "Weight tying in Mistral3 conditional generation and Qwen3.5 save_pretrained regression are different code paths and symptoms." + "reason": "Weight-tying bug in Mistral3ForConditionalGeneration is unrelated to Qwen3.5 save_pretrained checkpoint corruption." }, { "left": "issue:36246", "right": "issue:38617", "accept": false, - "reason": "Both are import errors, but they are for different missing symbols in different modules." + "reason": "Both are import errors, but for different symbols and different modules with no shared code path shown." }, { "left": "issue:43531", "right": "issue:45600", "accept": false, - "reason": "Qwen3-MoE sliding_window behavior and removed Sam3LiteText auto-mapping references are unrelated." + "reason": "Qwen3-MoE sliding_window behavior is unrelated to auto_mappings CI breakage." }, { - "left": "issue:41950", - "right": "issue:44016", + "left": "issue:44295", + "right": "issue:46506", "accept": false, - "reason": "Video-classification pipeline processor selection and a notebook syntax error are not the same bug." + "reason": "position_ids buffer access bug is unrelated to CPU/CUDA output divergence in PPDocLayoutV3." }, { - "left": "issue:44295", - "right": "issue:46506", + "left": "issue:41950", + "right": "issue:44016", "accept": false, - "reason": "position_ids buffer access and CPU/CUDA result divergence in PPDocLayoutV3 are different problems." + "reason": "Video-classification pipeline bug is unrelated to a notebook syntax error." }, { - "left": "issue:43673", - "right": "issue:46424", + "left": "issue:43582", + "right": "issue:46932", "accept": false, - "reason": "Cache-missing in chunked_prefill and max_cache_length only applying to quantized cache are separate cache semantics bugs." + "reason": "Apple Silicon warmup TypeError is unrelated to Gemma4 quantized embedder dtype casting." } ] }, @@ -12282,7 +12323,7 @@ { "label": "issue", "component_index": 1, - "component_count": 25, + "component_count": 26, "review_unit_index": 40, "review_unit_count": 61, "status": "reviewed", @@ -12300,7 +12341,6 @@ "issue:43441", "issue:43479", "issue:43526", - "issue:43582", "issue:43650", "issue:43673", "issue:43824", @@ -12308,10 +12348,12 @@ "issue:43976", "issue:44016", "issue:44062", + "issue:44075", "issue:44220", "issue:44279", "issue:44327", "issue:44336", + "issue:44368", "issue:44496", "issue:44521", "issue:44683", @@ -12338,13 +12380,12 @@ "issue:46506", "issue:46515", "issue:46693", - "issue:46752", - "issue:46932" + "issue:46752" ], "soft_pairs": [ - "issue:43582|issue:46932", - "issue:43650|issue:44220", + "issue:43673|issue:46424", "issue:43295|issue:45335", + "issue:43650|issue:44220", "issue:43976|issue:44987", "issue:45083|issue:46506", "issue:44016|issue:44220", @@ -12353,13 +12394,13 @@ "issue:44683|issue:45446", "issue:43901|issue:44496", "issue:42907|issue:43441", - "issue:43673|issue:44336", "issue:43381|issue:45593", + "issue:43673|issue:44336", "issue:43479|issue:45375", "issue:44496|issue:45759", "issue:43976|issue:44279", - "issue:44016|issue:44062", "issue:44792|issue:46424", + "issue:44016|issue:44062", "issue:43441|issue:45588", "issue:43824|issue:44933", "issue:44279|issue:46515", @@ -12371,230 +12412,240 @@ "issue:44279|issue:45405", "issue:46082|issue:46399", "issue:43295|issue:44779", - "issue:39692|issue:43901" + "issue:39692|issue:43901", + "issue:44075|issue:44368" ], - "prepared_review_unit_hash": "2c98f2ab0f8014b1590d082e4e767a0f7b41ae114858b859d191bbe034c9e049", + "prepared_review_unit_hash": "d2ce2975146acd803fd652eb0cb7376f94ab5597ec6c97b1130c737c48a95a54", "component_budget": { "node_count": 505, "item_count": 505, "soft_pair_count": 1854, - "serialized_chars": 962354, - "estimated_input_tokens": 240589, - "estimated_eval_tokens": 481434 + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 30, - "serialized_chars": 36590, - "estimated_input_tokens": 9148, - "estimated_eval_tokens": 18552 + "soft_pair_count": 31, + "serialized_chars": 36933, + "estimated_input_tokens": 9234, + "estimated_eval_tokens": 18724 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is heterogeneous: it mixes unrelated issues spanning docs, tokenizer/processor regressions, model-loading failures, config bugs, and performance/problems in different subsystems. I don\u2019t see a true duplicate group here, so there is no strong canonical item.", - "confidence": 0.98, - "canonical_issue_reason": null, + "summary": "Most items are unrelated single issues. The only likely duplicate pair is the FlashAttention crash on sink-less models, where the Ministral-3 report appears to be a model-specific manifestation of the same `s_aux=None` code-path failure.", + "confidence": 0.66, + "canonical_issue_reason": "issue:45588 is the most concrete and general report: it names the failing code path (`integrations/flash_attention.py`), the trigger (`s_aux=None`), and the affected class of models (sink-less).", "canonical_pr_reason": null, - "best_issue_reason": null, + "best_issue_reason": "issue:45588 is the best representative if these are clustered as duplicates, because it describes the underlying crash mechanism rather than a single model symptom.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:43582", - "right": "issue:46932", + "left": "issue:43673", + "right": "issue:46424", "accept": false, - "reason": "AppleSilicon caching_allocator_warmup TypeError vs Gemma4 quantized vision/audio dtype casting; different components and failures." + "reason": "Both mention cache behavior, but one is a GenerationMixin cache regression and the other is a cache-length config limitation; different bugs." }, { - "left": "issue:43650", - "right": "issue:44220", + "left": "issue:43295", + "right": "issue:45335", "accept": false, - "reason": "Placeholder 'ADD THE DATA' is not the same bug as _torch_extract_fbank_features(); no shared concrete code path." + "reason": "Multimodal processor/tokenizer regression vs T5Gemma embedding resize bug; no shared code-path or symptom." }, { - "left": "issue:43295", - "right": "issue:45335", + "left": "issue:43650", + "right": "issue:44220", "accept": false, - "reason": "Multimodal processor/tokenizer regression vs t5gemma resize_token_embeddings not updating decoder.embed_tokens; unrelated model paths." + "reason": "Completely unrelated titles and domains." }, { "left": "issue:43976", "right": "issue:44987", "accept": false, - "reason": "Python-version incompatibility in Transformers 5.1.0 vs loading a specific model package; different root causes." + "reason": "Python support/version issue vs model loading failure; not the same defect." }, { "left": "issue:45083", "right": "issue:46506", "accept": false, - "reason": "Helper output-length behavior in qwen3_omni_moe vs CPU/CUDA divergence in PPDocLayoutV3; different bugs." + "reason": "Different models and different failures: helper output-length logic vs CPU/CUDA divergence." }, { "left": "issue:44016", "right": "issue:44220", "accept": false, - "reason": "Notebook syntax error vs audio feature extraction helper issue; no overlap beyond being runtime failures." + "reason": "Notebook syntax error and audio feature extraction bug are unrelated." }, { "left": "issue:44496", "right": "issue:45042", "accept": false, - "reason": "Unrecognized model config vs PIL backend incorrectly requiring torchvision; different subsystems and symptoms." + "reason": "Unrecognized model config error vs PIL image processor torchvision requirement; different layers." }, { "left": "issue:43526", "right": "issue:43901", "accept": false, - "reason": "BeitImageProcessorFast reduce_labels bug vs TextClassificationPipeline docs mismatch; one is code behavior, the other docs." + "reason": "Image processor label reduction bug vs docs mismatch in a pipeline API." }, { "left": "issue:44683", "right": "issue:45446", "accept": false, - "reason": "torch>=2.9 flex_attention compilation failure vs incorrect version check for AuxRequest import; not the same code path." + "reason": "Both involve flex_attention, but one is a compiled-kernel failure and the other is a version-check/import bug; not clearly one bug." }, { "left": "issue:43901", "right": "issue:44496", "accept": false, - "reason": "Pipeline documentation mismatch vs model-loading/config error; unrelated issues." + "reason": "Docs/API wording issue vs model config loading failure; unrelated." }, { "left": "issue:42907", "right": "issue:43441", "accept": false, - "reason": "Saving dequantized Ministral/Devstral models vs FlashAttention failure for Ministral-3; different operations and failure modes." + "reason": "Both mention Ministral, but one is about saving dequantized weights and the other is a FlashAttention runtime crash." }, { - "left": "issue:43673", - "right": "issue:44336", + "left": "issue:43381", + "right": "issue:45593", "accept": false, - "reason": "GenerationMixin cache/chunked_prefill issue vs ANSI codes emitted in loading_report; unrelated subsystems." + "reason": "Gradient checkpointing eval-mode behavior vs D-FINE auxiliary-loss behavior; no overlap." }, { - "left": "issue:43381", - "right": "issue:45593", + "left": "issue:43673", + "right": "issue:44336", "accept": false, - "reason": "Gradient checkpointing/eval-mode issue vs D-FINE auxiliary-loss behavior; not the same bug." + "reason": "Cache-related regression and terminal ANSI formatting are unrelated." }, { "left": "issue:43479", "right": "issue:45375", "accept": false, - "reason": "Phi4MultimodalConfig default-init bug vs Qwen3_5MoeVisionConfig missing field/strict drop; different config classes and fields." + "reason": "Different multimodal config bugs in different model families; similar shape but not the same underlying change." }, { "left": "issue:44496", "right": "issue:45759", "accept": false, - "reason": "Unrecognized model config vs AutoModelForCausalLM not unwrapping text_config for composite Qwen multimodal configs; different failure points." + "reason": "One is unrecognized-model loading, the other is `from_config` unwrapping of `text_config`; not the same bug." }, { "left": "issue:43976", "right": "issue:44279", "accept": false, - "reason": "Python version support problem vs generic dependency issue; too broad and not the same underlying defect." + "reason": "Python version support issue vs generic dependency issue; no common code-path." }, { - "left": "issue:44016", - "right": "issue:44062", + "left": "issue:44792", + "right": "issue:46424", "accept": false, - "reason": "Notebook syntax error vs AddedToken special-argument TypeError; unrelated parser/runtime issues." + "reason": "Janus test failure is unrelated to cache-length config handling." }, { - "left": "issue:44792", - "right": "issue:46424", + "left": "issue:44016", + "right": "issue:44062", "accept": false, - "reason": "janus image-generation test failure vs cache_config max_cache_length behavior; different features." + "reason": "Notebook syntax error vs `AddedToken` keyword collision; unrelated." }, { "left": "issue:43441", "right": "issue:45588", - "accept": false, - "reason": "Ministral-3 FlashAttention failure vs flash_attention.py crash on sink-less models with s_aux=None; different branches and inputs." + "accept": true, + "reason": "Both point to the same FlashAttention failure on sink-less models; 43441 is a model-specific symptom, while 45588 names the underlying `s_aux=None` crash path." }, { "left": "issue:43824", "right": "issue:44933", "accept": false, - "reason": "Missing Qwen2.5 VL export vs nonexistent image_utils import; both import errors but for different symbols and causes." + "reason": "Different import failures from different modules; not duplicates." }, { "left": "issue:44279", "right": "issue:46515", "accept": false, - "reason": "Generic dependency issue vs torch 2.4.1 incompatibility; not enough evidence of one underlying bug." + "reason": "Generic dependency issue vs torch-version incompatibility; too broad and not the same bug." }, { "left": "issue:44521", "right": "issue:46693", "accept": false, - "reason": "apply_chat_template assistant_masks bug vs FA2 long-context performance regression; different behavior, not duplicate." + "reason": "Chat-template mask bug vs FA2 long-context performance regression; unrelated behavior." }, { "left": "issue:45335", "right": "issue:45584", "accept": false, - "reason": "Embedding resize bug in t5gemma vs Whisper empty-transcription failure after align_special_tokens; different models and code paths." + "reason": "Embedding resize bug vs Whisper empty-transcription generation failure; different code paths." }, { "left": "issue:46123", "right": "issue:46369", "accept": false, - "reason": "MaskGenerationPipeline partial-batch handling vs sam3-litetext memory overflow; unrelated generation issues." + "reason": "Partial-batch flag logic vs SAM memory overflow; unrelated." }, { "left": "issue:44327", "right": "issue:46752", "accept": false, - "reason": "QA decode_spans kth-out-of-bounds vs apply_chat_template empty-list IndexError; different APIs and failure conditions." + "reason": "QA span decoding top-k bug vs empty-list chat-template crash; different APIs and failure modes." }, { "left": "issue:43299", "right": "issue:45216", "accept": false, - "reason": "Qwen3VL MoE loading regression vs Qwen3.5 checkpoint save_pretrained regression; both Qwen-related but distinct code paths and symptoms." + "reason": "Different Qwen-related regressions affecting model loading vs checkpoint saving; not the same change." }, { "left": "issue:44279", "right": "issue:45405", "accept": false, - "reason": "Generic dependency issue vs PEFT minimum-version bump; no shared bug." + "reason": "Dependency issue title is too generic and does not match the PEFT version bump problem." }, { "left": "issue:46082", "right": "issue:46399", "accept": false, - "reason": "LlamaConfig head_dim validation vs Qwen3.5 apply_rotary_pos_emb registration/decorator issue; unrelated model/config problems." + "reason": "Different model/config issues with different triggers and subsystems." }, { "left": "issue:43295", "right": "issue:44779", "accept": false, - "reason": "Both are tokenizer-related regressions, but one is processor.tokenizer/image handling and the other is Deepseek tokenization correctness; not the same concrete bug." + "reason": "Both are v5 regressions around tokenizer/processor behavior, but they affect different models and report different symptoms." }, { "left": "issue:39692", "right": "issue:43901", "accept": false, - "reason": "SigLIP2 doc example errors vs TextClassificationPipeline docs mentioning return_all_scores; both docs-related, but different sections and issues." + "reason": "Documentation examples with model/processor mismatch vs pipeline docs wording mismatch; not the same issue." + }, + { + "left": "issue:44075", + "right": "issue:44368", + "accept": false, + "reason": "SGD optimizer args vs tie_word_embeddings warning; unrelated." } ] }, - "evaluator_result": null, - "evaluator_used": false, + "evaluator_result": { + "accept": true, + "feedback": "The clustering summary is conservative and stays within the packet. The only accepted edge, 43441 \u2194 45588, is phrased as a likely duplicate rather than a certainty, and the other pairs are rejected with appropriately narrow reasoning. No obvious overstatement of evidence." + }, + "evaluator_used": true, "retried": false, - "accepted_nontrivial_soft_edge": false, + "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null }, { "label": "issue", "component_index": 1, - "component_count": 25, + "component_count": 26, "review_unit_index": 41, "review_unit_count": 61, "status": "reviewed", @@ -12620,11 +12671,10 @@ "issue:43854", "issue:43976", "issue:44016", - "issue:44075", "issue:44242", "issue:44291", + "issue:44336", "issue:44351", - "issue:44368", "issue:44462", "issue:44466", "issue:44683", @@ -12646,14 +12696,15 @@ "issue:45812", "issue:45823", "issue:45834", + "issue:45859", "issue:45907", "issue:46082", "issue:46209", "issue:46512", + "issue:46851", "issue:47000" ], "soft_pairs": [ - "issue:44075|issue:44368", "issue:42796|issue:43650", "issue:43577|issue:45216", "issue:44351|issue:45278", @@ -12665,205 +12716,213 @@ "issue:44466|issue:45276", "issue:43381|issue:47000", "issue:43299|issue:43854", - "issue:44242|issue:45200", "issue:43653|issue:45593", - "issue:41950|issue:45823", + "issue:44242|issue:45200", "issue:43650|issue:46209", + "issue:41950|issue:45823", "issue:43479|issue:43525", "issue:44462|issue:45081", "issue:44016|issue:44291", - "issue:45588|issue:47000", "issue:44683|issue:45542", + "issue:45588|issue:47000", "issue:45081|issue:45907", "issue:43441|issue:45834", "issue:33290|issue:45656", "issue:42994|issue:43317", - "issue:45103|issue:45372" + "issue:45103|issue:45372", + "issue:44336|issue:46851", + "issue:45859|issue:46851" ], - "prepared_review_unit_hash": "bda8232e9ec1f4a88a8b35358667a4c6fa1e59661af29d65d24825bf8d0f97cf", + "prepared_review_unit_hash": "5bc45efc22a8b619ab3f86ea078cc4caae468e0808410a2316e0ba8d49527909", "component_budget": { "node_count": 505, "item_count": 505, "soft_pair_count": 1854, - "serialized_chars": 962354, - "estimated_input_tokens": 240589, - "estimated_eval_tokens": 481434 + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 }, "budget": { - "node_count": 47, - "item_count": 47, - "soft_pair_count": 26, - "serialized_chars": 35016, - "estimated_input_tokens": 8754, - "estimated_eval_tokens": 17764 + "node_count": 48, + "item_count": 48, + "soft_pair_count": 27, + "serialized_chars": 35856, + "estimated_input_tokens": 8964, + "estimated_eval_tokens": 18184 }, "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 (model-specific bugs, tokenizer regressions, config/load problems, and CI failures). None of the soft pairs look like the same underlying defect, so there is no clear canonical duplicate within this set.", - "confidence": 0.94, - "canonical_issue_reason": null, + "summary": "This set is mostly heterogeneous; the soft pairs are loose thematic similarities, not true duplicates. No PRs are present, and I would not merge any of the proposed pairs.", + "confidence": 0.13, + "canonical_issue_reason": "Issue 45278 is the broadest user-facing regression report in the set and the closest thing to an umbrella issue, but the overall cluster is too mixed to treat as one duplicate family.", "canonical_pr_reason": null, - "best_issue_reason": null, + "best_issue_reason": "45278 is the best representative issue because it captures a wide upgrade regression with multiple import failures, making it more suitable as a cluster anchor than the narrower, model-specific reports.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:44075", - "right": "issue:44368", - "accept": false, - "reason": "Both mention optimizer-related warnings/behavior, but one is about SGD args being ignored and the other is about a tied-embeddings warning during Qwen3.5 LoRA fine-tuning; different code paths and symptoms." - }, { "left": "issue:42796", "right": "issue:43650", "accept": false, - "reason": "Tokenizer BPE-marker handling and a vague 'ADD THE DATA' report are unrelated; no shared concrete bug or fix path." + "reason": "Different problems: tokenizer BPE-marker handling vs a vague data request." }, { "left": "issue:43577", "right": "issue:45216", "accept": false, - "reason": "Both involve dtype/state during model loading/saving, but they affect different models and different operations (Blip2 load dtype propagation vs Qwen3.5 save_pretrained checkpoint serialization)." + "reason": "Both involve dtype/serialization regressions, but they affect different models and different failure modes." }, { "left": "issue:44351", "right": "issue:45278", "accept": false, - "reason": "One is a missing HybridCache import, the other is broad import errors after upgrade; the latter is a generic upgrade breakage, not the same specific failure." + "reason": "Related to import breakage, but one is a specific missing symbol and the other is a broad upgrade regression; not clearly the same bug." }, { "left": "issue:43976", "right": "issue:45405", "accept": false, - "reason": "Python version compatibility for Transformers 5.1.0 is unrelated to an unreleased minimum PEFT version bump." + "reason": "Python version compatibility issue vs an unreleased dependency version bump; unrelated." }, { "left": "issue:46082", "right": "issue:46512", "accept": false, - "reason": "One is about invalid LlamaConfig head_dim validation; the other is stale model.config.dtype after from_config(dtype=...). Different model initialization issues." + "reason": "Different configs and symptoms: head_dim validation vs stale config dtype." }, { "left": "issue:41950", "right": "issue:45020", "accept": false, - "reason": "Video pipeline processor selection and remote_code breakage are unrelated subsystems with different failure modes." + "reason": "Pipeline processor lookup bug vs remote_code breakage; different code paths." }, { "left": "issue:44779", "right": "issue:45812", "accept": false, - "reason": "Both are tokenizer regressions, but one is DeepSeek tokenization and the other is Granite AutoTokenizer IDs; different models and likely different tokenizer logic." + "reason": "Both are tokenizer regressions, but for different model families and likely different root causes." }, { "left": "issue:43299", "right": "issue:45357", "accept": false, - "reason": "Both concern Qwen3.5/Qwen3VL loading or saving regressions, but one is a model-loading break in dev0 and the other is incorrect visual-encoder keys on save_pretrained; not the same bug." + "reason": "Loading failure for Qwen3VL MoE models vs save_pretrained key serialization for Qwen3.5; not the same issue." }, { "left": "issue:44466", "right": "issue:45276", "accept": false, - "reason": "Both are tied-weight/embedding serialization or resize behavior, but one is inconsistent lm_head.weight serialization while the other is Gemma4 resize_token_embeddings not updating per-layer/output embeddings." + "reason": "Serialization of tied weights vs resize_token_embeddings not updating Gemma4 embeddings; different behaviors." }, { "left": "issue:43381", "right": "issue:47000", "accept": false, - "reason": "Gradient checkpointing eval-mode misuse and RTDetrLoss infeasible cost matrix are unrelated training/runtime failures." + "reason": "Gradient checkpointing eval-mode restriction vs RTDetrLoss infeasible cost matrix; unrelated." }, { "left": "issue:43299", "right": "issue:43854", "accept": false, - "reason": "Both are model-loading failures, but they involve different models and causes (Qwen3VL MoE loading vs GLM-4.7-Flash unit-test loading)." - }, - { - "left": "issue:44242", - "right": "issue:45200", - "accept": false, - "reason": "Load-balancing loss missing when output_router_logits=False is an MoE/router-loss issue; mm_token_type_ids defaulting to zeros is a Gemma4 multimodal input handling issue." + "reason": "Different model-loading regressions with different failure points." }, { "left": "issue:43653", "right": "issue:45593", "accept": false, - "reason": "BigBirdTokenizer special-token registration/decoding and D-FINE auxiliary-loss handling are unrelated." + "reason": "Tokenizer special-token registration bug vs missing auxiliary losses in D-FINE." }, { - "left": "issue:41950", - "right": "issue:45823", + "left": "issue:44242", + "right": "issue:45200", "accept": false, - "reason": "Video pipeline image-processor lookup and Gemma4 PLE device mismatch are different problems in different parts of the stack." + "reason": "Router-load-balancing loss bug vs Gemma 4 text-only token type defaults; different subsystems." }, { "left": "issue:43650", "right": "issue:46209", "accept": false, - "reason": "A placeholder 'ADD THE DATA' issue is not the same as FP8 inference failure on B200 hardware." + "reason": "Generic data request vs FP8 inference failure; no overlap." + }, + { + "left": "issue:41950", + "right": "issue:45823", + "accept": false, + "reason": "Processor selection bug vs Gemma4 device-mapping mismatch; different failure modes." }, { "left": "issue:43479", "right": "issue:43525", "accept": false, - "reason": "Phi4Multimodal default config initialization and Llama4Config missing pad_token_id are distinct config-construction bugs." + "reason": "Different config initialization bugs in different model families." }, { "left": "issue:44462", "right": "issue:45081", "accept": false, - "reason": "AutoTokenizer ignoring tokenizer.json and a Mistral regex patch crash are different tokenizer-loading bugs with different failure points." + "reason": "AutoTokenizer ignoring tokenizer.json vs Mistral regex patch crash; distinct tokenizer-loading issues." }, { "left": "issue:44016", "right": "issue:44291", "accept": false, - "reason": "Notebook syntax error and init_empty_weights unexpected _is_hf_initialized argument are unrelated." + "reason": "Notebook syntax error vs init_empty_weights argument mismatch; unrelated." }, { - "left": "issue:45588", - "right": "issue:47000", + "left": "issue:44683", + "right": "issue:45542", "accept": false, - "reason": "flash_attention.py crashing on s_aux=None is unrelated to RTDetrLoss infeasible cost matrix." + "reason": "Compiled flex_attention torch-version breakage vs TensorBoard/TF backend packaging issue." }, { - "left": "issue:44683", - "right": "issue:45542", + "left": "issue:45588", + "right": "issue:47000", "accept": false, - "reason": "Compiled flex_attention on torch>=2.9 and tensorboard-installed-without-TensorFlow backend detection are unrelated." + "reason": "FlashAttention sink-less-model crash vs RTDetrLoss infeasible matrix; different code paths." }, { "left": "issue:45081", "right": "issue:45907", "accept": false, - "reason": "Mistral tokenizer regex patch crash and list_repo_files not forwarding revision are different code paths and not the same defect." + "reason": "Tokenizer regex crash vs repo file listing not forwarding revision; unrelated." }, { "left": "issue:43441", "right": "issue:45834", "accept": false, - "reason": "Ministral FlashAttention failure and Kosmos2.5 OCR index error are unrelated model-specific runtime bugs." + "reason": "FlashAttention incompatibility for Ministral-3 vs Kosmos2.5 OCR long-input index error." }, { "left": "issue:33290", "right": "issue:45656", "accept": false, - "reason": "Deepspeed/Adafactor OOM and deepspeed optimizer step being called twice are both Deepspeed-related but clearly different issues: memory blowup vs duplicate stepping." + "reason": "Both mention DeepSpeed/optimizer behavior, but one is OOM with Adafactor and the other is duplicate optimizer.step calls." }, { "left": "issue:42994", "right": "issue:43317", "accept": false, - "reason": "Quantized model saving failure and device_map=auto dequantized-model loading failure are separate quantization workflows with different breakages." + "reason": "Quantized model saving failure vs dequantized model loading/offload failure; opposite sides of the pipeline." }, { "left": "issue:45103", "right": "issue:45372", "accept": false, - "reason": "Auto docstring kwargs processing with future annotations and Gemma4 processor loading broken by ReasoningEffort import are unrelated failures." + "reason": "Auto-docstring annotation handling bug vs Gemma 4 processor import failure; unrelated." + }, + { + "left": "issue:44336", + "right": "issue:46851", + "accept": false, + "reason": "ANSI output formatting issue vs Qwen3.5 variable-length position_ids bug." + }, + { + "left": "issue:45859", + "right": "issue:46851", + "accept": false, + "reason": "CPU-offload incompatibility in rotary embedding vs variable-length position_ids behavior; not the same bug." } ] }, @@ -12877,7 +12936,7 @@ { "label": "issue", "component_index": 1, - "component_count": 25, + "component_count": 26, "review_unit_index": 42, "review_unit_count": 61, "status": "reviewed", @@ -12928,7 +12987,6 @@ "issue:45712", "issue:45800", "issue:45823", - "issue:45859", "issue:46291", "issue:46399", "issue:46655", @@ -12937,8 +12995,6 @@ "issue:47000" ], "soft_pairs": [ - "issue:44336|issue:46851", - "issue:45859|issue:46851", "issue:44623|issue:44749", "issue:44261|issue:44263", "issue:45600|issue:45712", @@ -12950,8 +13006,8 @@ "issue:43723|issue:44162", "issue:43525|issue:43881", "issue:44186|issue:47000", - "issue:43824|issue:43976", "issue:44016|issue:45823", + "issue:43824|issue:43976", "issue:42796|issue:43441", "issue:45542|issue:45800", "issue:43525|issue:43643", @@ -12969,214 +13025,208 @@ "issue:46399|issue:46851", "issue:45561|issue:45563" ], - "prepared_review_unit_hash": "f3aa718496b5523f391996b3ac92ffbe3768591b719a1575c2f6a1abe510805e", + "prepared_review_unit_hash": "29edde0f7f8ca07fced4c5ed4783bc6a29866842aae6f5bea829108efcee7dea", "component_budget": { "node_count": 505, "item_count": 505, "soft_pair_count": 1854, - "serialized_chars": 962354, - "estimated_input_tokens": 240589, - "estimated_eval_tokens": 481434 + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 }, "budget": { - "node_count": 48, - "item_count": 48, - "soft_pair_count": 31, - "serialized_chars": 36930, - "estimated_input_tokens": 9233, - "estimated_eval_tokens": 18722 + "node_count": 47, + "item_count": 47, + "soft_pair_count": 29, + "serialized_chars": 35759, + "estimated_input_tokens": 8940, + "estimated_eval_tokens": 18136 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This looks like a noisy bundle of mostly unrelated Transformers regressions; none of the soft-edge pairs read as true duplicates. The strongest representative issue is the Qwen2.5-VL import failure, since it has the most discussion/inbound references and a very concrete failure mode.", - "confidence": 0.86, - "canonical_issue_reason": "issue:43824 is the highest-signal report in the set: specific import failure, multiple inbound references, and substantial discussion make it the best stand-in for the cluster\u2019s overall regression theme.", + "summary": "This cluster is heterogeneous: the items span unrelated import, packaging, tokenizer, model-specific, and docs/runtime issues. I did not find any soft pair that clearly looks like the same underlying bug/change, so no soft edges are accepted.", + "confidence": 0.88, + "canonical_issue_reason": "issue:43824 is the strongest single representative: a high-engagement, user-facing import failure with a clear repro and broad impact, making it the most useful anchor in an otherwise mixed cluster.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43824 is a strong canonical candidate because it is well-scoped, reproducible, and clearly about a user-facing breakage rather than a doc/test-only or vaguely described problem.", + "best_issue_reason": "issue:43824 is the best issue candidate for triage because it is concrete, high-signal, and has stronger discussion/inbound references than most others here.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:44336", - "right": "issue:46851", - "accept": false, - "reason": "Unrelated symptoms: ANSI coloring in loading_report vs Qwen3.5 variable-length position_ids behavior." - }, - { - "left": "issue:45859", - "right": "issue:46851", - "accept": false, - "reason": "Both involve Qwen3.5, but one is CPU offload rotary embedding compatibility and the other is variable-length position_ids; different bugs." - }, { "left": "issue:44623", "right": "issue:44749", "accept": false, - "reason": "Processor save/load file omission is unrelated to a slowdown after upgrade while filtering data." + "reason": "Different bugs: missing save_pretrained files vs a tokenizer/data-filtering slowdown. No shared code-path." }, { "left": "issue:44261", "right": "issue:44263", "accept": false, - "reason": "Different code paths: MLA normalization precision issue vs torch.split return values in a MoE indexer." + "reason": "Unrelated model internals: RMS norm config precision issue vs torch.split return handling." }, { "left": "issue:45600", "right": "issue:45712", "accept": false, - "reason": "Both are maintenance/CI problems, but one is auto_mappings stale configs and the other is leftover dummy classes; not the same defect." + "reason": "Both are stale-artifact/CI-adjacent, but they target different files and failure modes (auto_mappings vs dummy_pt_objects)." }, { "left": "issue:42175", "right": "issue:43976", "accept": false, - "reason": "Backend dependency packaging issue vs Python version compatibility failure; no shared underlying bug." + "reason": "Different packaging problems: missing TensorFlow in torch extra vs Python-version compatibility." }, { "left": "issue:44671", "right": "issue:44779", "accept": false, - "reason": "Different models and failure modes: CamemBERT MLM predictions vs Deepseek tokenizer output regression." + "reason": "Different regressions: CamemBERT MLM outputs vs Deepseek tokenizer behavior." }, { "left": "issue:44623", "right": "issue:45397", "accept": false, - "reason": "Missing processor save artifacts is not the same as gemma-4 zero3 from_pretrained failure." + "reason": "Different model/load paths: processor save_pretrained artifact loss vs Gemma-4 ZeRO3 from_pretrained." }, { "left": "issue:43901", "right": "issue:44246", "accept": false, - "reason": "Docs mismatch for return_all_scores is unrelated to intermittent import slowness." + "reason": "Docs mismatch for a pipeline option vs intermittent slow import. Not the same issue." }, { "left": "issue:43673", "right": "issue:43761", "accept": false, - "reason": "Generation cache missing in chunked prefill and CLIPVisionModel hidden_states handling are different regressions." + "reason": "Different execution paths: generation cache during chunked prefill vs CLIPVisionModel hidden_states output." }, { "left": "issue:43723", "right": "issue:44162", "accept": false, - "reason": "Tokenizer loading in v5 and ESM2 breakage are distinct failures in different subsystems." + "reason": "Different failures: AutoTokenizer loading regression vs ESM2 model breakage." }, { "left": "issue:43525", "right": "issue:43881", "accept": false, - "reason": "Llama4Config missing pad_token_id and glm-4v-9b loading failed are separate model-specific issues." + "reason": "Both are model-loading failures, but they involve different models and different missing fields/errors." }, { "left": "issue:44186", "right": "issue:47000", "accept": false, - "reason": "LayoutLMv2 tokenizer crashes vs RTDetrLoss infeasible cost matrix are unrelated problems." + "reason": "Different subsystems: tokenizer batching/NER crash vs RTDetrLoss infeasible cost matrix." }, { - "left": "issue:43824", - "right": "issue:43976", + "left": "issue:44016", + "right": "issue:45823", "accept": false, - "reason": "ImportError for a model class is unrelated to a Python 3.9/3.10 compatibility report." + "reason": "Notebook syntax error and Gemma4 device-mismatch are unrelated." }, { - "left": "issue:44016", - "right": "issue:45823", + "left": "issue:43824", + "right": "issue:43976", "accept": false, - "reason": "Notebook syntax error and Gemma4 device mismatch do not share an underlying bug." + "reason": "ImportError for a specific class vs Python version support; unrelated root causes." }, { "left": "issue:42796", "right": "issue:43441", "accept": false, - "reason": "Same model family, but tokenizer BPE marker handling and FlashAttention failure are different code paths." + "reason": "Same model family, but tokenizer BPE handling and FlashAttention crash are distinct bugs." }, { "left": "issue:45542", "right": "issue:45800", "accept": false, - "reason": "Packaging missing TensorFlow backend vs a torch/transformers incompatibility are different issues." + "reason": "Different installation/runtime issues: TensorFlow backend exposure vs torch/transformers incompatibility." }, { "left": "issue:43525", "right": "issue:43643", "accept": false, - "reason": "Both mention missing fields, but one is a model config attribute bug and the other is trust_remote_code returning incomplete objects." + "reason": "Both mention missing config fields, but the model/config contexts and concrete failures differ." }, { "left": "issue:44188", "right": "issue:44336", "accept": false, - "reason": "Attention-kernel divergence and ANSI code emission are unrelated." + "reason": "Attention-kernel divergence on torch.compile vs ANSI code emission in loading_report are unrelated." }, { "left": "issue:44589", "right": "issue:45362", "accept": false, - "reason": "Float8 storage import/type error vs Qwen3.5 chat crash are not the same bug." + "reason": "Storage-object deserialization error vs Qwen3.5 chat loading crash; different code paths." }, { "left": "issue:45325", "right": "issue:46851", "accept": false, - "reason": "Both involve position_ids, but the model families and failure mechanisms differ." + "reason": "Both involve position_ids/rope in Qwen models, but they are different models and different behaviors." }, { "left": "issue:46291", "right": "issue:46399", "accept": false, - "reason": "Both touch kernels, but one is dependency-version import failure and the other is decorator/registration crash; not mergeable as one fix." + "reason": "Both are kernels-related, but one is an import-time dependency requirement and the other is a registration/decorator crash." }, { "left": "issue:44336", "right": "issue:44373", "accept": false, - "reason": "ANSI terminal codes vs a wrong docstring are unrelated." + "reason": "Loading-report ANSI formatting vs a wrong docstring; no shared bug." }, { "left": "issue:46655", "right": "issue:46726", "accept": false, - "reason": "Video padding calculation bug and missing vision backend error are different processor issues." + "reason": "Different processor bugs: video padding calculation vs missing vision-backend handling on AutoProcessor." }, { "left": "issue:44246", "right": "issue:44496", "accept": false, - "reason": "Intermittent import slowness is unrelated to unrecognized model/config loading failure." + "reason": "Intermittent slow import vs unrecognized model/config missing model_type. Not the same issue." + }, + { + "left": "issue:41628", + "right": "issue:43976", + "accept": false, + "reason": "ImportError for AutoImageProcessor vs Python version compatibility; unrelated." }, { "left": "issue:43784", "right": "issue:45542", "accept": false, - "reason": "Importing sentence-transformers fails for a different reason than missing TensorFlow backend installation." + "reason": "Importing sentence-transformers breaks for a different missing symbol than the TensorFlow backend/package issue." }, { "left": "issue:45083", "right": "issue:46851", "accept": false, - "reason": "Different Qwen variants and different failure contexts; not the same code-path defect." + "reason": "Both touch position-related helpers, but they are different models and different failure modes." }, { "left": "issue:44297", "right": "issue:45397", "accept": false, - "reason": "Tokenizer metadata mismatch is unrelated to gemma-4 zero3 loading." + "reason": "Tokenizer metadata saved incorrectly vs Gemma-4 loading under ZeRO3; unrelated." }, { "left": "issue:46399", "right": "issue:46851", "accept": false, - "reason": "Kernel registration crash vs variable-length position_ids bug are distinct issues." + "reason": "Different Qwen3.5 failures: kernels registration crash vs variable-length position_ids behavior." }, { "left": "issue:45561", "right": "issue:45563", "accept": false, - "reason": "xdist race on captured_info.txt is unrelated to a stale warning about num_return_sequences." + "reason": "pytest-xdist file race vs stale warning for num_return_sequences; no shared underlying bug." } ] }, @@ -13190,7 +13240,7 @@ { "label": "issue", "component_index": 1, - "component_count": 25, + "component_count": 26, "review_unit_index": 43, "review_unit_count": 61, "status": "reviewed", @@ -13249,8 +13299,8 @@ "issue:46851" ], "soft_pairs": [ - "issue:45103|issue:46399", "issue:43526|issue:45003", + "issue:45103|issue:46399", "issue:45072|issue:46829", "issue:43540|issue:46752", "issue:42491|issue:43299", @@ -13259,8 +13309,8 @@ "issue:44246|issue:44871", "issue:45588|issue:46851", "issue:44295|issue:45397", - "issue:44242|issue:45593", "issue:45003|issue:46821", + "issue:44242|issue:45593", "issue:44336|issue:44485", "issue:42886|issue:46132", "issue:43550|issue:46506", @@ -13278,70 +13328,70 @@ "issue:45375|issue:45759", "issue:44479|issue:46655" ], - "prepared_review_unit_hash": "8a7094650052a0c067fb85836a603818adec9e06fcd064d1f1c923589080b9df", + "prepared_review_unit_hash": "9bdfac12c1b123168b69fdbfd93d93d05f39d5268682d98200b6dfc18e445a89", "component_budget": { "node_count": 505, "item_count": 505, "soft_pair_count": 1854, - "serialized_chars": 962354, - "estimated_input_tokens": 240589, - "estimated_eval_tokens": 481434 + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 }, "budget": { "node_count": 47, "item_count": 47, "soft_pair_count": 28, - "serialized_chars": 35563, - "estimated_input_tokens": 8891, - "estimated_eval_tokens": 18038 + "serialized_chars": 35569, + "estimated_input_tokens": 8893, + "estimated_eval_tokens": 18042 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is heterogeneous: the soft pairs mostly share broad vocabulary (Qwen, video, import, offline, etc.) but not the same concrete bug or code path, so I reject all soft edges. If one issue must represent the cluster, the broadest and most actionable is the Qwen video regression report (#44479).", - "confidence": 0.73, - "canonical_issue_reason": "#44479 is the best representative because it describes a broad, version-specific regression spanning multiple Qwen VL/MoE variants, which is more central than the narrower single-model or one-off import/runtime reports.", + "summary": "The proposed pairs are mostly loose topical matches rather than true duplicates. I did not find any soft edge where the two issues clearly describe the same underlying bug or code path.", + "confidence": 0.91, + "canonical_issue_reason": "issue:44479 is the strongest canonical anchor in this set because it describes a concrete, version-specific regression that affects multiple related Qwen video models and is well scoped.", "canonical_pr_reason": null, - "best_issue_reason": "#44479 is the most suitable issue to anchor the cluster since it covers several related affected models and a clear regression surface, making it the broadest plausible canonical issue among these heterogeneous reports.", + "best_issue_reason": "issue:44479 is the best standalone issue here: it is specific, reproducible, and broad enough to represent a meaningful regression affecting several models, unlike the other mostly narrow or unrelated reports.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:45103", - "right": "issue:46399", + "left": "issue:43526", + "right": "issue:45003", "accept": false, - "reason": "Different subsystems: auto_docstring annotation parsing vs Qwen3.5 kernel registration." + "reason": "Different bugs: BeitImageProcessorFast label reduction vs unsafe sys.modules access in modeling_utils." }, { - "left": "issue:43526", - "right": "issue:45003", + "left": "issue:45103", + "right": "issue:46399", "accept": false, - "reason": "BeitImageProcessorFast label reduction bug is unrelated to modeling_utils sys.modules access." + "reason": "Both mention internal framework behavior, but one is auto-docstring parsing with future annotations and the other is Qwen3.5 kernel registration; different code paths." }, { "left": "issue:45072", "right": "issue:46829", "accept": false, - "reason": "Different model families and failures: dtype mismatch vs wrong hidden_states output." + "reason": "Unrelated failures: dtype mismatch in model inference versus wrong hidden_states output in a video classifier." }, { "left": "issue:43540", "right": "issue:46752", "accept": false, - "reason": "Qwen3OmniMoe video input validation is unrelated to empty chat template handling." + "reason": "Different areas: Qwen3OmniMoe video-input processing versus empty-list handling in apply_chat_template." }, { "left": "issue:42491", "right": "issue:43299", "accept": false, - "reason": "Both mention Qwen MoE loading, but the symptoms and affected paths differ enough that they are not the same bug." + "reason": "Both involve Qwen3/MoE loading regressions, but one is LoRA/accelerate/deepspeed compatibility and the other is direct Qwen3VL MoE loading; not the same bug." }, { "left": "issue:43531", "right": "issue:45362", "accept": false, - "reason": "Sliding window behavior in Qwen3-MoE is unrelated to Qwen3.5 chat crashes." + "reason": "Qwen3-MoE sliding_window behavior and Qwen3.5 chat crashes are separate model/runtime issues." }, { "left": "issue:46614", @@ -13353,139 +13403,127 @@ "left": "issue:44246", "right": "issue:44871", "accept": false, - "reason": "Import-time slowdown and Gemma eos_token_id mismatch are different problems." + "reason": "Different failures: missing auxiliary loss when router logits are off versus inconsistent eos_token_id config in Gemma-3." }, { "left": "issue:45588", "right": "issue:46851", "accept": false, - "reason": "Flash-attention sink handling is unrelated to variable-length position_ids in Qwen3.5." + "reason": "Both touch Qwen3.5, but one is flash_attention sink handling and the other is variable-length position_ids; not the same code path." }, { "left": "issue:44295", "right": "issue:45397", "accept": false, - "reason": "Buffer registration on position_ids does not match gemma-4 Zero3 from_pretrained issues." + "reason": "A position_ids buffer access error and a gemma-4 ZeRO3 from_pretrained issue are unrelated." }, { - "left": "issue:44242", - "right": "issue:45593", + "left": "issue:45003", + "right": "issue:46821", "accept": false, - "reason": "Missing load-balancing loss and missing auxiliary losses are similar in wording but occur in different models and training flows." + "reason": "Different subsystems entirely: sys.modules access in modeling_utils versus TimesFM window_size handling." }, { - "left": "issue:45003", - "right": "issue:46821", + "left": "issue:44242", + "right": "issue:45593", "accept": false, - "reason": "Unsafe sys.modules access is unrelated to the TimesFM window_size AttributeError." + "reason": "Both concern losses being skipped, but they are different models and different conditions; not one concrete defect." }, { "left": "issue:44336", "right": "issue:44485", "accept": false, - "reason": "ANSI output formatting and GLM-5 RoPE implementation are not the same bug." + "reason": "ANSI color output in loading_report is unrelated to GLM-5 RoPE implementation." }, { "left": "issue:42886", "right": "issue:46132", "accept": false, - "reason": "Offline tokenizer cache loading and AttentionInterface.register semantics are unrelated." + "reason": "Offline cache loading and AttentionInterface.register behavior changes are unrelated." }, { "left": "issue:43550", "right": "issue:46506", "accept": false, - "reason": "torch.compile SDPA failure in Bamba is unrelated to CPU/CUDA result divergence in PPDocLayoutV3." + "reason": "Different failures: torch.compile+SDPA on Bamba versus CPU/CUDA result divergence in PPDocLayoutV3." }, { "left": "issue:44779", "right": "issue:45292", "accept": false, - "reason": "Deepseek tokenizer correctness regression is unrelated to resize_token_embeddings not updating output embeddings." + "reason": "Tokenizer output regressions and resize_token_embeddings not updating output embeddings are separate issues." }, { "left": "issue:44451", "right": "issue:46752", "accept": false, - "reason": "Model loading failure for a specific checkpoint is unrelated to empty apply_chat_template input." + "reason": "ScandiBERT loading failure is unrelated to empty-list chat template handling." }, { "left": "issue:36246", "right": "issue:42757", "accept": false, - "reason": "Different import errors from different packages: qwen2_5_vl export vs huggingface_hub symbol." + "reason": "Both are import errors, but for different missing symbols in different packages; not the same underlying bug." }, { "left": "issue:36246", "right": "issue:44351", "accept": false, - "reason": "Both are import errors, but they concern different missing symbols and different modules." + "reason": "Missing Qwen2_5_VLImageProcessor export and missing HybridCache export are distinct symbol/import problems." }, { "left": "issue:43901", "right": "issue:45003", "accept": false, - "reason": "Documentation mismatch for TextClassificationPipeline is unrelated to modeling_utils internals." + "reason": "Docs mismatch for TextClassificationPipeline is unrelated to sys.modules access safety." }, { "left": "issue:43257", "right": "issue:45137", "accept": false, - "reason": "Qwen3 MoE weight conversion and DeepSpeed ZeRO3 deque errors are different failure modes." + "reason": "Both involve DeepSpeed, but one is Qwen3 MoE weight conversion and the other is a ZeRO3 deque error; different failures." }, { "left": "issue:41950", "right": "issue:44220", "accept": false, - "reason": "Video-classification image-processor lookup is unrelated to _torch_extract_fbank_features." + "reason": "Video-classification processor selection and audio fbank extraction are unrelated." }, { "left": "issue:43540", "right": "issue:46655", "accept": false, - "reason": "Both are video-related, but Qwen3OmniMoe video input errors and SmolVLM padding calculation are distinct code paths." + "reason": "Both are video-input bugs, but they affect different models and different processing logic." }, { "left": "issue:44496", "right": "issue:45003", "accept": false, - "reason": "Unrecognized model config and unsafe sys.modules access are not the same issue." + "reason": "Unrecognized-model config loading and unsafe sys.modules access are unrelated." }, { "left": "issue:45003", "right": "issue:45561", "accept": false, - "reason": "modeling_utils sys.modules access is unrelated to pytest-xdist captured_info.txt races." + "reason": "Generic import-safety issue versus pytest-xdist file race in testing utils." }, { "left": "issue:43262", "right": "issue:44246", "accept": false, - "reason": "Audio sampling-rate defaulting and import slowdown are unrelated." + "reason": "Audio chat-template sampling-rate default and expert load-balancing loss omission are unrelated." }, { "left": "issue:45375", "right": "issue:45759", "accept": false, - "reason": "Missing config field in Qwen3.5 MoE vision config is unrelated to text_config unwrapping in composite multimodal configs." - }, - { - "left": "issue:44479", - "right": "issue:46655", - "accept": false, - "reason": "Both involve video, but one is a broad Qwen version regression and the other is a SmolVLM padding bug." - }, - { - "left": "issue:44479", - "right": "issue:46655", - "accept": false, - "reason": "Same as above; different model families and different failure mechanisms." + "reason": "Both concern Qwen multimodal configs, but one is a missing field dropped by strict config handling and the other is text_config unwrapping in from_config; different bugs." }, { "left": "issue:44479", "right": "issue:46655", "accept": false, - "reason": "Same as above; not the same underlying code-path problem." + "reason": "Both are video regressions, but the affected models and failure mechanisms are different; not the same concrete bug." } ] }, @@ -13499,7 +13537,7 @@ { "label": "issue", "component_index": 1, - "component_count": 25, + "component_count": 26, "review_unit_index": 44, "review_unit_count": 61, "status": "reviewed", @@ -13567,198 +13605,198 @@ "issue:43475|issue:46369", "issue:43295|issue:45292", "issue:44466|issue:45335", - "issue:44877|issue:45375", "issue:36246|issue:43596", "issue:44242|issue:44857", + "issue:44877|issue:45375", "issue:44623|issue:46726", "issue:43526|issue:46897", "issue:44987|issue:45459", "issue:44038|issue:46506", "issue:29870|issue:30333", - "issue:42199|issue:46851", - "issue:44060|issue:44077", "issue:44479|issue:46170", "issue:43526|issue:46894", + "issue:44060|issue:44077", + "issue:42199|issue:46851", "issue:44368|issue:44977", "issue:45412|issue:46123", "issue:43653|issue:44162", "issue:43352|issue:45250", "issue:42886|issue:45229", - "issue:43475|issue:43540", - "issue:44877|issue:46170" + "issue:44877|issue:46170", + "issue:43475|issue:43540" ], - "prepared_review_unit_hash": "bac2c2d39d2325c3e4a7831610f841a4daadbe64711d9416466e898360c57aaa", + "prepared_review_unit_hash": "fc7648eb25efd8c97ba9f69664eaa4fb56ddcbadcb7dfebdc00bba28861c0ca7", "component_budget": { "node_count": 505, "item_count": 505, "soft_pair_count": 1854, - "serialized_chars": 962354, - "estimated_input_tokens": 240589, - "estimated_eval_tokens": 481434 + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 27, - "serialized_chars": 36040, - "estimated_input_tokens": 9010, - "estimated_eval_tokens": 18276 + "serialized_chars": 36034, + "estimated_input_tokens": 9009, + "estimated_eval_tokens": 18274 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The set is highly heterogeneous, and every soft pair appears to be at most a loose topical similarity rather than the same underlying bug or change. I would reject all soft edges and treat this cluster as not having a single clear duplicate canonical.", - "confidence": 0.96, - "canonical_issue_reason": "No single canonical issue fits well: these issues span unrelated bugs across many models, configs, tokenizers, and pipelines, so any representative would be arbitrary.", - "canonical_pr_reason": "No pull requests were provided in the input, so there is no canonical PR candidate.", - "best_issue_reason": "No issue is globally best for this cluster because the items do not share one concrete failure mode or code path.", - "best_pr_reason": "No pull requests were provided in the input, so there is no best PR candidate.", + "summary": "All soft edges look like false positives: the items span unrelated bugs across model families, tokenizer/config loading, and runtime errors. No pair is close enough to be the same concrete underlying issue, so none should be merged.", + "confidence": 0.94, + "canonical_issue_reason": "issue:45292 is the most general, self-contained API-level bug in the set and the best representative anchor if one issue must be chosen.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45292 is the cleanest model-agnostic report and the best dedupe anchor, even though it is not actually a duplicate of the others.", + "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44162", "right": "issue:45200", "accept": false, - "reason": "Different bugs: ESM2 is a model-specific breakage, while Gemma 4 mm_token_type_ids is a missing-default input handling issue." + "reason": "Different models and failures: ESM2 breakage vs Gemma4 mm_token_type_ids defaults." }, { "left": "issue:45859", "right": "issue:46399", "accept": false, - "reason": "Both involve Qwen3.5 rotary logic, but one is CPU-offload incompatibility and the other is a decorator/registration crash in apply_rotary_pos_emb; not the same fix." + "reason": "Both involve Qwen3.5, but one is CPU-offload compatibility and the other is a registration/decorator crash; different bugs." }, { "left": "issue:45292", "right": "issue:47000", "accept": false, - "reason": "Unrelated: token embedding resizing vs RT-DETR loss infeasibility." + "reason": "resize_token_embeddings vs RT-DETR cost matrix infeasible are unrelated subsystems and symptoms." }, { "left": "issue:44295", "right": "issue:44749", "accept": false, - "reason": "Different subsystems and symptoms: buffer/position_ids error versus a performance regression during data processing." + "reason": "Position_ids buffer access error is unrelated to the reported data-processing slowdown/regression." }, { "left": "issue:45137", "right": "issue:45600", "accept": false, - "reason": "DeepSpeed ZeRO3 deque failure and stale auto_mappings config references are unrelated." + "reason": "DeepSpeed ZeRO3 deque underflow and Sam3LiteText auto_mappings CI breakage are different code paths." }, { "left": "issue:43475", "right": "issue:46369", "accept": false, - "reason": "Both mention SAM3, but one is a missing attribute in video output and the other is a memory overflow in sam3-litetext; different code paths." + "reason": "Both mention SAM 3, but one is a missing attribute in video output and the other is a memory overflow in litetext." }, { "left": "issue:43295", "right": "issue:45292", "accept": false, - "reason": "Processor.tokenizer regression is unrelated to resize_token_embeddings not updating output embeddings." + "reason": "Custom processor/tokenizer regression is not the same bug as resize_token_embeddings not updating output embeddings." }, { "left": "issue:44466", "right": "issue:45335", "accept": false, - "reason": "Both touch embedding weights, but one is tied-weight serialization inconsistency and the other is resize_token_embeddings not propagating to decoder.embed_tokens." - }, - { - "left": "issue:44877", - "right": "issue:45375", - "accept": false, - "reason": "Both involve strict config handling, but they are different model configs and different failures: loading blocked versus a field silently dropped." + "reason": "Serialization inconsistency with tied weights is a different issue from resize_token_embeddings not affecting decoder.embed_tokens." }, { "left": "issue:36246", "right": "issue:43596", "accept": false, - "reason": "Completely different failures: missing Qwen2_5_VLImageProcessor import versus a DeepSpeed/BertModel index error." + "reason": "Qwen2_5_VL import failure and BertModel/deepspeed IndexError are unrelated." }, { "left": "issue:44242", "right": "issue:44857", "accept": false, - "reason": "Load-balancing loss gating and RT-DETR AMP crashes are different bugs in different models." + "reason": "MoE load-balancing loss gating is unrelated to LwDetrImageLoss AMP/CUDA crashes." + }, + { + "left": "issue:44877", + "right": "issue:45375", + "accept": false, + "reason": "Both touch config strictness, but they involve different models and different missing-field behaviors; not the same bug." }, { "left": "issue:44623", "right": "issue:46726", "accept": false, - "reason": "processor.save_pretrained missing files is not the same as AutoProcessor failing when a required vision backend is absent." + "reason": "Missing files on save_pretrained is different from AutoProcessor backend-missing handling." }, { "left": "issue:43526", "right": "issue:46897", "accept": false, - "reason": "These are different label-shifting bugs in different processors/models; similar pattern, not the same underlying defect." + "reason": "BEiT Fast label reduction bug and Florence2 double-shift training-loss bug are unrelated." }, { "left": "issue:44987", "right": "issue:45459", "accept": false, - "reason": "Loading a model on transformers>=5.1.0 and tokenizer protobuf error masking are unrelated." + "reason": "Model loading regression and protobuf-exception masking are different failure modes." }, { "left": "issue:44038", "right": "issue:46506", "accept": false, - "reason": "Qwen3-VL-Moe compatibility and CPU/CUDA result divergence in PPDocLayoutV3 are unrelated." + "reason": "Qwen3-VL-Moe regression and PPDocLayoutV3 CPU/CUDA divergence are unrelated." }, { "left": "issue:29870", "right": "issue:30333", "accept": false, - "reason": "Bart generate() UnboundLocalError has nothing to do with MLflow reporting after failed training jobs." + "reason": "Bart generate() UnboundLocalError is unrelated to MLFlow job-status reporting." }, { - "left": "issue:42199", - "right": "issue:46851", + "left": "issue:44479", + "right": "issue:46170", "accept": false, - "reason": "DETR cardinality/background-class handling and Qwen3.5 variable-length position_ids are different model-specific bugs." + "reason": "Video-input regression for Qwen models and Windows/ROCm torch.distributed import failure are different bugs." }, { - "left": "issue:44060", - "right": "issue:44077", + "left": "issue:43526", + "right": "issue:46894", "accept": false, - "reason": "One is a tied-weights warning in Qwen3-Next; the other is a patchtsmixer post_init policy issue." + "reason": "BEiT Fast label handling and CohereASR double-shift loss are unrelated." }, { - "left": "issue:44479", - "right": "issue:46170", + "left": "issue:44060", + "right": "issue:44077", "accept": false, - "reason": "Qwen video-input regression and Windows/ROCm torch.distributed import errors are unrelated." + "reason": "Tied-weights warning involving dt_bias and a patchtsmixer post_init restriction are not the same underlying issue." }, { - "left": "issue:43526", - "right": "issue:46894", + "left": "issue:42199", + "right": "issue:46851", "accept": false, - "reason": "Both mention training-loss shift patterns, but they are different models and not the same concrete bug or patch." + "reason": "DETR cardinality handling and Qwen3.5 variable-length position_ids are unrelated." }, { "left": "issue:44368", "right": "issue:44977", "accept": false, - "reason": "A tie_word_embeddings warning in Qwen3.5 is unrelated to flash-attention generation failure." + "reason": "Tie_word_embeddings warning during LoRA fine-tuning is different from flash-attention generation failure." }, { "left": "issue:45412", "right": "issue:46123", "accept": false, - "reason": "RT-DETR memory release and MaskGenerationPipeline partial-batch handling are unrelated." + "reason": "RT-DETR memory lifecycle bug is unrelated to MaskGenerationPipeline partial-batch handling." }, { "left": "issue:43653", "right": "issue:44162", "accept": false, - "reason": "BigBirdTokenizer special-token decoding and ESM2 breakage are unrelated." + "reason": "BigBirdTokenizer special-token registration and ESM2 breakage are unrelated." }, { "left": "issue:43352", "right": "issue:45250", "accept": false, - "reason": "A model-specific Flash Attention support error is not the same as a generic Flash Attention 2.0 topic issue." + "reason": "Model-specific Flash Attention 2 support gap is not the same as the generic flash-attention issue." }, { "left": "issue:42886", @@ -13767,16 +13805,16 @@ "reason": "Offline tokenizer cache loading and Gemma4 multi-GPU OOM are unrelated." }, { - "left": "issue:43475", - "right": "issue:43540", + "left": "issue:44877", + "right": "issue:46170", "accept": false, - "reason": "Different models and failures: SAM3 video output attribute missing versus Qwen3OmniMoe video-input ValueError." + "reason": "Strict config loading and Windows/ROCm dependency import failure are different code paths." }, { - "left": "issue:44877", - "right": "issue:46170", + "left": "issue:43475", + "right": "issue:43540", "accept": false, - "reason": "Strict config loading for granite_speech is unrelated to a Windows/ROCm distributed import/type error." + "reason": "SAM 3 Video attribute error and Qwen3OmniMoe video-input ValueError are different model bugs." } ] }, @@ -13790,7 +13828,7 @@ { "label": "issue", "component_index": 1, - "component_count": 25, + "component_count": 26, "review_unit_index": 45, "review_unit_count": 61, "status": "reviewed", @@ -13865,8 +13903,8 @@ "issue:44188|issue:46421", "issue:45685|issue:46431", "issue:45103|issue:46858", - "issue:44361|issue:44749", "issue:43065|issue:46123", + "issue:44361|issue:44749", "issue:44297|issue:44623", "issue:43994|issue:46726", "issue:44749|issue:45656", @@ -13876,219 +13914,219 @@ "issue:44623|issue:45656", "issue:43441|issue:46752", "issue:43475|issue:43643", - "issue:45375|issue:45478", "issue:42994|issue:43482", + "issue:45375|issue:45478", "issue:43475|issue:46821", "issue:28282|issue:30990" ], - "prepared_review_unit_hash": "0e76cb679ec993040a2dbec1e8cf518d6aa8f8fcc0e6c4dfff98e8785ca3568e", + "prepared_review_unit_hash": "3c84c15928ab1d5198adfe7cc031e8a357a709cd47029d44ee3c3d718c67d0b8", "component_budget": { "node_count": 505, "item_count": 505, "soft_pair_count": 1854, - "serialized_chars": 962354, - "estimated_input_tokens": 240589, - "estimated_eval_tokens": 481434 + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 30, - "serialized_chars": 36797, - "estimated_input_tokens": 9200, - "estimated_eval_tokens": 18656 + "serialized_chars": 36804, + "estimated_input_tokens": 9201, + "estimated_eval_tokens": 18658 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is heterogeneous: the items share a general Transformers/v5 ecosystem but not the same underlying bug. Most soft pairs are only loosely related by model family, backend, or warning/error type, so they should remain separate issues.", - "confidence": 0.93, - "canonical_issue_reason": "Issue 45656 is the closest anchor because it is a concrete, active regression and appears in multiple weak similarity links, but the set is still too diverse for a true duplicate canonical.", + "summary": "These issues span many unrelated regressions across model loading, caching, tokenization, pipelines, and backend-specific failures. None of the proposed soft pairs look like the same concrete bug/change under a strict duplicate triage standard.", + "confidence": 0.97, + "canonical_issue_reason": "issue:46858 is the most concrete and actionable report here: it names the exact failing path (_prefill), the broken contract (cache.get_seq_length()), and the affected scenario (static cache + torch.compile on non-CUDA backends).", "canonical_pr_reason": null, - "best_issue_reason": "Issue 45656 is the best single representative by being open, specific, and the most centrally connected in the candidate graph, though it does not subsume the others.", + "best_issue_reason": "issue:46858 has the clearest symptom-to-code-path mapping and is the strongest candidate for a canonical bug report among the issues shown.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:46170", "right": "issue:46832", "accept": false, - "reason": "Different problems: Windows/ROCm import dependency issue vs RT-DETR feature-level shape mismatch." + "reason": "Different failures: Windows/ROCm import/type dependency issue vs RTDetrModel forward feature-level mismatch." }, { "left": "issue:45397", "right": "issue:45656", "accept": false, - "reason": "Deepspeed optimizer-step duplication is unrelated to gemma-4 zero3 from_pretrained behavior." + "reason": "Different code paths: Gemma-4 Zero3 from_pretrained failure vs deepspeed causing optimizer.step() to run twice." }, { "left": "issue:43065", "right": "issue:43408", "accept": false, - "reason": "Both are SAM3-related, but one is a dummy conv layer issue and the other is a model-type/config warning." + "reason": "Same model family, but different bugs: dummy Conv2d implementation vs a model-type warning/misinstantiation issue." }, { "left": "issue:43723", "right": "issue:46710", "accept": false, - "reason": "Tokenizer loading failure and accuracy regression after upgrade are different code-paths and symptoms." + "reason": "Tokenizer loading problem vs post-upgrade accuracy regression; not the same underlying defect." }, { "left": "issue:44368", "right": "issue:45563", "accept": false, - "reason": "Both are warnings, but they concern unrelated APIs and stale warning behavior." + "reason": "Unrelated warnings in different features: tie_word_embeddings warning vs stale num_return_sequences warning in paged generate." }, { "left": "issue:44995", "right": "issue:45200", "accept": false, - "reason": "Second-forward cache crash is unrelated to Gemma 4 text-only mm_token_type_ids defaults." + "reason": "Different models and failures: stale cache/indexer crash vs Gemma 4 multimodal token type defaults." }, { "left": "issue:43931", "right": "issue:44368", "accept": false, - "reason": "Weight-shape mismatch in Qwen3-VL is not the same bug as a tie_word_embeddings warning." + "reason": "Weight-shape mismatch for Qwen3-VL is unrelated to the tie_word_embeddings warning on Qwen3.5 fine-tuning." }, { "left": "issue:44188", "right": "issue:46202", "accept": false, - "reason": "Torch.compile attention kernel divergence is unrelated to Gemma pipeline roundtrip tokenization." + "reason": "Different concrete problems: attention-kernel divergence under torch.compile vs Gemma pipeline roundtrip failure." }, { "left": "issue:45987", "right": "issue:46082", "accept": false, - "reason": "StaticCache return type bug and LlamaConfig head_dim validation are different issues." + "reason": "Cache API return-type bug vs LlamaConfig validation of head_dim; distinct issues." }, { "left": "issue:44483", "right": "issue:46202", "accept": false, - "reason": "API request rejection and Gemma pipeline roundtrip failure do not share the same failure mode." + "reason": "API request rejection in /v1/chat/completions is unrelated to Gemma pipeline roundtrip behavior." }, { "left": "issue:43906", "right": "issue:46202", "accept": false, - "reason": "Isolated reproduction of another issue and a Gemma pipeline text roundtrip bug are not the same problem." + "reason": "An isolated reproduction of another issue is not the same as the Gemma roundtrip failure." }, { "left": "issue:44995", "right": "issue:45588", "accept": false, - "reason": "Stale indexer cache on second forward is unrelated to flash_attention s_aux=None crashes." + "reason": "Different failures: GLM MoE stale index cache on second forward vs flash_attention crashing on s_aux=None." }, { "left": "issue:44188", "right": "issue:46421", "accept": false, - "reason": "Both mention cache/compile, but one is divergent kernels and the other is spurious recompilation from lazy init." + "reason": "torch.compile kernel divergence is unrelated to cache lazy initialization causing recompilation." }, { "left": "issue:45685", "right": "issue:46431", "accept": false, - "reason": "MPS histogram/int error is unrelated to missing chat template in processor.apply_chat_template." + "reason": "MPS histogram dtype error and missing chat template are unrelated bugs." }, { "left": "issue:45103", "right": "issue:46858", "accept": false, - "reason": "Auto-docstring annotation handling and cache.get_seq_length device-tensor arithmetic are different bugs." + "reason": "auto_docstring annotations crash is unrelated to StaticCache.get_seq_length returning a device tensor." }, { - "left": "issue:44361", - "right": "issue:44749", + "left": "issue:43065", + "right": "issue:46123", "accept": false, - "reason": "Tokenizer task AttributeError is unrelated to a data-filtering slowdown after upgrade." + "reason": "Different SAM3 pixel-decoder placeholder issue vs mask generation partial-batch handling bug." }, { - "left": "issue:43065", - "right": "issue:46123", + "left": "issue:44361", + "right": "issue:44749", "accept": false, - "reason": "SAM3 dummy conv and MaskGenerationPipeline partial-batch dropping results are different subsystems." + "reason": "Tokenizer AttributeError and a severe slowdown after upgrading are not the same defect." }, { "left": "issue:44297", "right": "issue:44623", "accept": false, - "reason": "Tokenizer save_pretrained metadata mismatch and processor save_pretrained missing files are related only at a high level, not the same concrete bug." + "reason": "Both involve save_pretrained, but one is tokenizer_config class metadata and the other is processor file omission; not one concrete bug." }, { "left": "issue:43994", "right": "issue:46726", "accept": false, - "reason": "Nonsensical SigLIP2 outputs and missing vision-backend image-processor errors are different loading paths." + "reason": "Nonsensical model outputs vs missing vision backend handling are different problems." }, { "left": "issue:44749", "right": "issue:45656", "accept": false, - "reason": "Performance regression in data filtering is unrelated to duplicate optimizer stepping under deepspeed." + "reason": "Performance regression after upgrade vs optimizer stepping twice under deepspeed are distinct issues." }, { "left": "issue:43701", "right": "issue:44749", "accept": false, - "reason": "Checkpoint key mismatch and data-filtering slowdown are unrelated." + "reason": "Checkpoint key mismatch is unrelated to the reported slowdown regression." }, { "left": "issue:43646", "right": "issue:44162", "accept": false, - "reason": "Custom model initialization breakage and ESM2-specific breakage are too broad and not the same concrete bug." + "reason": "Custom model initialization breakage in v5 is not the same as the ESM2 regression." }, { "left": "issue:44871", "right": "issue:46132", "accept": false, - "reason": "Gemma eos_token_id config inconsistency is unrelated to AttentionInterface.register behavior changes." + "reason": "Gemma eos_token_id mismatch is unrelated to AttentionInterface.register behavior changes." }, { "left": "issue:44623", "right": "issue:45656", "accept": false, - "reason": "Processor save_pretrained file emission and deepspeed optimizer stepping are different concerns." + "reason": "Processor save_pretrained file loss and duplicate optimizer steps are unrelated." }, { "left": "issue:43441", "right": "issue:46752", "accept": false, - "reason": "FlashAttention failure for Ministral-3 and empty-list apply_chat_template IndexError are unrelated." + "reason": "FlashAttention failure for Ministral-3 is unrelated to apply_chat_template on an empty list." }, { "left": "issue:43475", "right": "issue:43643", "accept": false, - "reason": "Missing fpn_position_embeddings in SAM3 video output is not the same as AutoConfig trust_remote_code missing fields." + "reason": "Missing Sam3VisionEncoderOutput field and AutoConfig trust_remote_code field loss are different code paths and symptoms." }, { - "left": "issue:45375", - "right": "issue:45478", + "left": "issue:42994", + "right": "issue:43482", "accept": false, - "reason": "Missing deepstack_visual_indexes in config and a broader Qwen3.5 Moe from_pretrained error are related by model family only, not the same bug." + "reason": "Quantized model saving failure vs Qwen2.5-GGUF loading failure are different operations and bugs." }, { - "left": "issue:42994", - "right": "issue:43482", + "left": "issue:45375", + "right": "issue:45478", "accept": false, - "reason": "Quantized model saving failure and Qwen2.5-GGUF loading failure are opposite directions with different failure points." + "reason": "Related model family, but different failures: dropped config field under @strict vs from_pretrained error on Qwen3.5 Moe." }, { "left": "issue:43475", "right": "issue:46821", "accept": false, - "reason": "SAM3 video output attribute error and TimesFM window_size AttributeError are unrelated." + "reason": "SAM3 vision encoder attribute error is unrelated to TimesFM window_size AttributeError." }, { "left": "issue:28282", "right": "issue:30990", "accept": false, - "reason": "PyTorch missing ImportError and Sentence Transformers loading stall are different installation/runtime problems." + "reason": "Missing PyTorch ImportError and Sentence Transformers stuck loading are different failure modes." } ] }, @@ -14102,7 +14140,7 @@ { "label": "issue", "component_index": 1, - "component_count": 25, + "component_count": 26, "review_unit_index": 46, "review_unit_count": 61, "status": "reviewed", @@ -14163,19 +14201,19 @@ ], "soft_pairs": [ "issue:44964|issue:45137", - "issue:43232|issue:43673", "issue:44188|issue:46851", - "issue:45310|issue:45375", + "issue:43232|issue:43673", "issue:41628|issue:44933", + "issue:45310|issue:45375", "issue:39692|issue:46726", "issue:46877|issue:47000", "issue:46463|issue:46877", - "issue:44869|issue:45356", "issue:43986|issue:46726", + "issue:44869|issue:45356", "issue:45230|issue:45397", "issue:44038|issue:44623", - "issue:44373|issue:46851", "issue:42994|issue:43525", + "issue:44373|issue:46851", "issue:43723|issue:46326", "issue:42371|issue:45092", "issue:43873|issue:44945", @@ -14185,245 +14223,245 @@ "issue:44945|issue:45753", "issue:45137|issue:47000", "issue:36010|issue:41628", - "issue:44857|issue:46877", "issue:45081|issue:45103", - "issue:46202|issue:46701", + "issue:44857|issue:46877", "issue:44038|issue:45310", "issue:45593|issue:46877", + "issue:46202|issue:46701", "issue:42994|issue:43824", "issue:44222|issue:45310", "issue:42673|issue:43704", "issue:42831|issue:44222" ], - "prepared_review_unit_hash": "8c9ae777775f79b5089872160d9eb3fd75df3be8bad5dcfd483595e325863ec2", + "prepared_review_unit_hash": "932272859dcce178ef32b0a385618e859d6ef473dab1041fb54685d598d8cb7b", "component_budget": { "node_count": 505, "item_count": 505, "soft_pair_count": 1854, - "serialized_chars": 962354, - "estimated_input_tokens": 240589, - "estimated_eval_tokens": 481434 + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 32, - "serialized_chars": 37000, - "estimated_input_tokens": 9250, - "estimated_eval_tokens": 18756 + "serialized_chars": 37005, + "estimated_input_tokens": 9252, + "estimated_eval_tokens": 18760 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly noisy: many items share broad themes like import/export regressions, tokenizer/model loading issues, or generation bugs, but they are not the same underlying defect. The strongest relationship is the Qwen3ForCausalLM VRAM leak pair (42673/43704), which looks like a likely duplicate / same issue based on title and symptom, with 42673 carrying the fuller discussion.", - "confidence": 0.83, - "canonical_issue_reason": "Issue 42673 is the best anchor because it matches 43704 on title and symptom and has the richer discussion trail, making it the strongest representative of that likely-duplicate pair.", + "summary": "Mostly unrelated issues; only the Qwen3ForCausalLM VRAM leak pair is an exact duplicate. Everything else differs by symbol, model family, or failure mode enough to keep separate.", + "confidence": 0.97, + "canonical_issue_reason": "Use issue:42673 as canonical: it is the earlier, clearer report of the same Qwen3ForCausalLM VRAM leak under multiple dataloader threads, and issue:43704 is a later duplicate of the same symptom.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 42673 is the strongest global issue choice because it appears to be the primary report for the Qwen3ForCausalLM VRAM leak and includes the most complete discussion for that likely duplicate.", + "best_issue_reason": "issue:42673 is the best representative overall because it is specific, reproducible, and has the clearest duplicate relationship in the set.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44964", "right": "issue:45137", "accept": false, - "reason": "Both are crashes, but one is a model-load/import problem and the other is a DeepSpeed ZeRO3 deque underflow; different code paths." - }, - { - "left": "issue:43232", - "right": "issue:43673", - "accept": false, - "reason": "Both touch generation internals, but sync_gpus kwargs handling and chunked_prefill cache handling are different bugs." + "reason": "Different failures: model loading with latest transformers vs DeepSpeed ZeRO3 deque error." }, { "left": "issue:44188", "right": "issue:46851", "accept": false, - "reason": "Attention-kernel divergence on torch.compile is unrelated to Qwen3.5 variable-length position_ids handling." + "reason": "Both involve generation internals, but one is attention-kernel divergence under torch.compile and the other is variable-length position_ids behavior in Qwen3.5." }, { - "left": "issue:45310", - "right": "issue:45375", + "left": "issue:43232", + "right": "issue:43673", "accept": false, - "reason": "Same model family, but one is a from_pretrained failure and the other is a missing config field being dropped; not clearly one bug." + "reason": "Generation flow issues, but one is sync_gpus/_update_model_kwargs_for_generation and the other is chunked_prefill cache availability in v5.0.0." }, { "left": "issue:41628", "right": "issue:44933", "accept": false, - "reason": "Both are import/export complaints, but they involve different symbols and modules, so they do not look like the same defect." + "reason": "Both are import problems, but they concern different missing names from different modules and code paths." + }, + { + "left": "issue:45310", + "right": "issue:45375", + "accept": false, + "reason": "Same model family, but one is a from_pretrained error and the other is a config field being dropped by @strict." }, { "left": "issue:39692", "right": "issue:46726", "accept": false, - "reason": "A doc example mismatch/quantization issue is unrelated to AutoProcessor backend-missing error handling." + "reason": "Doc-example/model mismatch and quantization errors are not the same as AutoProcessor's missing-vision-backend error." }, { "left": "issue:46877", "right": "issue:47000", "accept": false, - "reason": "auto_docstring stdout diagnostics and RTDetrLoss cost-matrix failure are unrelated." + "reason": "Unrelated issues: import-time doc-lint stdout noise vs RTDetrLoss infeasible cost matrix crash." }, { "left": "issue:46463", "right": "issue:46877", "accept": false, - "reason": "Incorrect T5 decoder-input validation messaging is unrelated to auto_docstring import-time diagnostics." + "reason": "Different subsystems entirely: T5 decoder-input validation messaging vs auto_docstring import-time diagnostics." }, { - "left": "issue:44869", - "right": "issue:45356", + "left": "issue:43986", + "right": "issue:46726", "accept": false, - "reason": "Whisper timestamp decoding crash and Kimi tokenizer regression are distinct tokenizer bugs." + "reason": "Both mention processor/backend availability, but one is a video-model load crash without torchvision and the other is an AutoProcessor image-processor error." }, { - "left": "issue:43986", - "right": "issue:46726", + "left": "issue:44869", + "right": "issue:45356", "accept": false, - "reason": "Both mention missing vision dependencies, but one is a video-model AutoProcessor crash and the other is an image-processor backend lookup issue; too different to merge." + "reason": "Both are tokenizer/decoding bugs, but they affect different models and different concrete failure modes." }, { "left": "issue:45230", "right": "issue:45397", "accept": false, - "reason": "Generic bug report is not the same as the Gemma-4 zero3 from_pretrained issue." + "reason": "A generic bug report is not the same as the specific gemma-4 Zero3 from_pretrained issue." }, { "left": "issue:44038", "right": "issue:44623", "accept": false, - "reason": "Qwen3-VL-Moe compatibility and processor.save_pretrained missing files are different subsystems." + "reason": "Different symptoms and likely different paths: Qwen3-VL-Moe bug vs processor.save_pretrained missing files." }, { - "left": "issue:44373", - "right": "issue:46851", + "left": "issue:42994", + "right": "issue:43525", "accept": false, - "reason": "Docstring text for position_ids is not the same issue as the runtime variable-length position_ids bug." + "reason": "Quantized model saving failure is unrelated to Llama4Config missing pad_token_id." }, { - "left": "issue:42994", - "right": "issue:43525", + "left": "issue:44373", + "right": "issue:46851", "accept": false, - "reason": "Quantized model saving failure and missing pad_token_id on Llama4Config are unrelated." + "reason": "Docstring wording for position_ids is not the same as Qwen3.5 variable-length input behavior." }, { "left": "issue:43723", "right": "issue:46326", "accept": false, - "reason": "Tokenizer loading in v5 and continue_final_message for mistral_common are different code paths." + "reason": "Tokenizer loading failure in v5 is not the same as continue_final_message behavior for mistral_common tokenizers." }, { "left": "issue:42371", "right": "issue:45092", "accept": false, - "reason": "TF32 settings guidance is unrelated to InternVL2 meta-initialization incompatibility." + "reason": "TF32 API guidance and InternVL2 v5 meta-init incompatibility are unrelated." }, { "left": "issue:43873", "right": "issue:44945", "accept": false, - "reason": "Quantization/offloading behavior and pipeline-parallel output corruption are separate bugs." + "reason": "Quantization offloading behavior and pipeline-parallel output corruption are different bugs." }, { "left": "issue:43784", "right": "issue:46399", "accept": false, - "reason": "sentence-transformers import failure and Qwen3.5 kernel registration error are unrelated." + "reason": "Sentence-transformers import NameError and Qwen3.5 kernel registration crash are different code paths." }, { "left": "issue:45406", "right": "issue:46202", "accept": false, - "reason": "Gemma4Processor missing _tokenizer and Gemma \u2581 roundtrip failure are different symptoms and likely different fixes." + "reason": "Gemma4Processor missing _tokenizer and Gemma pipeline roundtrip on \u2581 are different issues." }, { "left": "issue:41628", "right": "issue:44351", "accept": false, - "reason": "Both are import errors, but they are about different public symbols and different modules." + "reason": "Both are import errors from transformers, but they are for different symbols and causes." }, { "left": "issue:44945", "right": "issue:45753", "accept": false, - "reason": "Pipeline-parallel incorrect output and a Qwen3.5 image-specific infinite loop are not the same bug." + "reason": "Pipeline-parallel incorrect output and Qwen3_5 infinite loop on one image are different runtime failures." }, { "left": "issue:45137", "right": "issue:47000", "accept": false, - "reason": "DeepSpeed ZeRO3 deque underflow and RTDetrLoss infeasible cost matrix are unrelated failures." + "reason": "DeepSpeed ZeRO3 deque underflow is unrelated to RTDetrLoss cost-matrix infeasibility." }, { "left": "issue:36010", "right": "issue:41628", "accept": false, - "reason": "Both are import regressions, but GenerationMixin import failure and AutoImageProcessor import failure are separate API breaks." - }, - { - "left": "issue:44857", - "right": "issue:46877", - "accept": false, - "reason": "LwDetrLoss AMP crash and auto_docstring stdout diagnostics have no shared code path." + "reason": "Different missing imports: GenerationMixin from transformers.generation vs AutoImageProcessor from transformers." }, { "left": "issue:45081", "right": "issue:45103", "accept": false, - "reason": "Mistral tokenizer regex crash and auto_docstring future-annotations crash are unrelated." + "reason": "Tokenizer regex patch crash and auto_docstring annotation handling are unrelated." }, { - "left": "issue:46202", - "right": "issue:46701", + "left": "issue:44857", + "right": "issue:46877", "accept": false, - "reason": "Both involve Gemma-related pipelines, but the failing paths are different (token roundtrip vs image-text-to-text generation output)." + "reason": "Float16 AMP CUDA crash in LwDetrImageLoss is unrelated to doc-lint stdout logging." }, { "left": "issue:44038", "right": "issue:45310", "accept": false, - "reason": "Both are Qwen MoE model-loading issues, but they target different model variants and failure modes." + "reason": "Both mention Qwen3 MoE, but the concrete crash conditions and failure modes are different." }, { "left": "issue:45593", "right": "issue:46877", "accept": false, - "reason": "D-FINE auxiliary-loss behavior is unrelated to auto_docstring diagnostics." + "reason": "D-FINE auxiliary-loss behavior is unrelated to auto_docstring diagnostics output." + }, + { + "left": "issue:46202", + "right": "issue:46701", + "accept": false, + "reason": "Both touch Gemma pipelines, but one is a roundtrip tokenizer edge case and the other is an image-text-to-text pipeline failure." }, { "left": "issue:42994", "right": "issue:43824", "accept": false, - "reason": "Quantized saving failure and Qwen2_5_VL import failure are unrelated." + "reason": "Quantized saving failure and Qwen2_5_VL import error are different problems." }, { "left": "issue:44222", "right": "issue:45310", "accept": false, - "reason": "FP8 save_pretrained on MoE and a Qwen3.5 from_pretrained error are different bugs." + "reason": "FP8 save_pretrained moe and Qwen3.5 MoE from_pretrained error are different bugs." }, { "left": "issue:42673", "right": "issue:43704", "accept": true, - "reason": "Same title and symptom (Qwen3ForCausalLM VRAM leak with multiple dataloader threads); 42673 is the richer report and 43704 looks like the same underlying bug." + "reason": "Same underlying bug: Qwen3ForCausalLM leaks VRAM when used across multiple dataloader threads; the later issue is a duplicate of the same symptom." }, { "left": "issue:42831", "right": "issue:44222", "accept": false, - "reason": "FineGrainedFP8 accuracy issues and FP8 save_pretrained failures are different concrete problems." + "reason": "Both relate to FP8, but one is an accuracy issue and the other is a save_pretrained bug." } ] }, "evaluator_result": { - "accept": false, - "feedback": "The cluster summary is a bit too strong for the evidence in the packet. In particular, calling 42673/43704 an \"exact duplicate match\" is not grounded by any explicit duplicate link in the packet; the evidence only supports that they have the same title/symptom and that 42673 has a richer discussion trail. The negative soft-edge verdicts are generally conservative, but the top-line duplicate conclusion should be softened to something like \"likely duplicate / same issue\" rather than exact duplicate." + "accept": true, + "feedback": "Grounded overall. The Qwen3ForCausalLM pair (issue:42673 / issue:43704) is strongly supported as a duplicate by the identical titles and earlier/later timestamps, and the rest of the soft-edge rejections stay conservative based on the packet titles alone. Minor note: avoid implying reproducibility unless it\u2019s explicit in the issue body." }, "evaluator_used": true, - "retried": true, + "retried": false, "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null @@ -14431,7 +14469,7 @@ { "label": "issue", "component_index": 1, - "component_count": 25, + "component_count": 26, "review_unit_index": 47, "review_unit_count": 61, "status": "reviewed", @@ -14491,8 +14529,8 @@ ], "soft_pairs": [ "issue:43232|issue:44792", - "issue:43504|issue:45561", "issue:42175|issue:44933", + "issue:43504|issue:45561", "issue:45561|issue:45600", "issue:44496|issue:45561", "issue:45405|issue:46877", @@ -14515,220 +14553,220 @@ "issue:43784|issue:43824", "issue:27957|issue:30333", "issue:42898|issue:43295", - "issue:43874|issue:45092", "issue:30333|issue:34689", + "issue:43874|issue:45092", "issue:44734|issue:45103", "issue:42898|issue:45356", "issue:42222|issue:45200" ], - "prepared_review_unit_hash": "161cc6835dfd7f01820dc89228d0306ba2bffb6872312926645a2e4df042b82e", + "prepared_review_unit_hash": "af7d2ca785f1e43a657ffdea1a0ef40f1589edf71445078b865cb15e8d20bd19", "component_budget": { "node_count": 505, "item_count": 505, "soft_pair_count": 1854, - "serialized_chars": 962354, - "estimated_input_tokens": 240589, - "estimated_eval_tokens": 481434 + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 }, "budget": { "node_count": 47, "item_count": 47, "soft_pair_count": 30, - "serialized_chars": 36381, - "estimated_input_tokens": 9096, - "estimated_eval_tokens": 18448 + "serialized_chars": 36378, + "estimated_input_tokens": 9095, + "estimated_eval_tokens": 18446 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The cluster is heterogeneous. The soft pairs mostly share broad themes like tokenization, multimodal loading, or docs/build issues, but not the same concrete bug or change, so I would not merge any of them as duplicates.", - "confidence": 0.92, - "canonical_issue_reason": "issue:45812 is the strongest standalone anchor: it describes a broad, silent v4\u2192v5 tokenizer regression affecting all Granite models, which is more concrete and representative than the narrower or unrelated reports.", + "summary": "This cluster is heterogeneous: it mixes unrelated issue types across tokenization, model loading, multimodal configs, docs, and test infra. No soft pair looks like a true duplicate or the same fixable code-path problem, and there are no PRs here. If one issue must anchor the cluster, the open and more active #45561 is the best representative, but it is not a real duplicate of the others.", + "confidence": 0.95, + "canonical_issue_reason": "#45561 is the strongest cluster representative because it is open, has multiple inbound references, and has the most current triage signal among an otherwise mixed set of issues.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45812 is the best overall issue candidate because it is a clearly scoped, user-facing regression with broad impact and a crisp reproduction class.", + "best_issue_reason": "#45561 is the most suitable single issue to surface for manual review, but only as a representative anchor; the cluster does not form a coherent duplicate set.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43232", "right": "issue:44792", "accept": false, - "reason": "Generation kwargs sync and a Janus image-generation test failure are unrelated bug paths." + "reason": "Different generation bug vs janus image-generation test failure; no shared code path." }, { - "left": "issue:43504", - "right": "issue:45561", + "left": "issue:42175", + "right": "issue:44933", "accept": false, - "reason": "A legacy preset-loading failure in BEiT segmentation does not match an xdist race on captured_info.txt." + "reason": "Backend dependency packaging issue vs non-existent import from image_utils; unrelated causes." }, { - "left": "issue:42175", - "right": "issue:44933", + "left": "issue:43504", + "right": "issue:45561", "accept": false, - "reason": "Missing TensorFlow in a torch extra and a nonexistent image_utils import are different packaging/API problems." + "reason": "Legacy pretrained-load bug vs pytest-xdist file-write race; different subsystems." }, { "left": "issue:45561", "right": "issue:45600", "accept": false, - "reason": "A test-worker file race and stale auto_mappings entries are distinct CI failures." + "reason": "Captured-info race in tests vs stale auto_mappings config references; not the same bug." }, { "left": "issue:44496", "right": "issue:45561", "accept": false, - "reason": "An unrecognized-model config error and a pytest-xdist capture race are unrelated." + "reason": "Unrecognized model/config loading error vs test utility race; distinct failure modes." }, { "left": "issue:45405", "right": "issue:46877", "accept": false, - "reason": "A dependency-version pin issue and auto_docstring stdout logging are separate repo hygiene problems." + "reason": "Dependency version bump issue vs docstring import-time stderr/stdout behavior; unrelated." }, { "left": "issue:43976", "right": "issue:44933", "accept": false, - "reason": "Python-version compatibility and an invalid import from image_utils are not the same defect." + "reason": "Python version compatibility vs missing import symbol; different problems." }, { "left": "issue:30333", "right": "issue:33290", "accept": false, - "reason": "MLflow reporting on training failure and a DeepSpeed Adafactor OOM are different training issues." + "reason": "MLflow reporting on failed jobs vs DeepSpeed adafactor OOM; no shared underlying defect." }, { "left": "issue:30990", "right": "issue:35141", "accept": false, - "reason": "Sentence-Transformers loading hangs and embedding resizing reinitialization are unrelated model-loading vs initialization bugs." + "reason": "Loading hang in sentence-transformers vs embedding resize/post_init reinit bug; different code paths." }, { "left": "issue:42898", "right": "issue:44568", "accept": false, - "reason": "Both involve tokenizers in v5, but one is about clean_up_tokenization_spaces and the other about add_special_tokens behavior for a specific tokenizer." + "reason": "Both are tokenizer regressions in v5, but they affect different behaviors and fixes." }, { "left": "issue:42175", "right": "issue:45712", "accept": false, - "reason": "A missing TensorFlow backend dependency is not the same as leftover dummy PT classes leaking into dir(transformers)." + "reason": "Optional TensorFlow backend packaging vs leftover dummy PT objects; unrelated repository hygiene issues." }, { "left": "issue:43479", "right": "issue:46829", "accept": false, - "reason": "Default multimodal config initialization and a VideoPrism output-field mixup are different model bugs." + "reason": "Config default initialization bug vs wrong output tensor field in video classification output; different models and logic." }, { "left": "issue:44623", "right": "issue:45310", "accept": false, - "reason": "Processor save_pretrained file omissions and Qwen3.5 MoE from_pretrained errors do not look like the same change." + "reason": "Processor save_pretrained missing files vs Qwen3.5 MoE from_pretrained error; not the same defect." }, { "left": "issue:46077", "right": "issue:46123", "accept": false, - "reason": "Causal masks for Gemma4 and a batch-finalization bug in MaskGenerationPipeline are unrelated pipeline behaviors." + "reason": "Attention-mask construction bug vs batch-finalization logic in a generation pipeline; different paths." }, { "left": "issue:42673", "right": "issue:46693", "accept": false, - "reason": "A Qwen3 VRAM leak in dataloader threads and an FA2 long-context performance regression are both performance-related but not the same code-path bug." + "reason": "VRAM leak from multiple dataloader threads vs long-context FA2 performance regression; distinct symptoms and fixes." }, { "left": "issue:42898", "right": "issue:45812", "accept": false, - "reason": "Both are tokenizer regressions, but one concerns cleanup-space behavior while the other is wrong token IDs for Granite models." + "reason": "Both are tokenization regressions, but one is clean-up spacing and the other is wrong Granite token IDs." }, { "left": "issue:43479", "right": "issue:44869", "accept": false, - "reason": "Multimodal config defaults and Whisper timestamp decoding crashes are unrelated." + "reason": "Multimodal config defaults vs Whisper timestamp decoding crash; unrelated components." }, { "left": "issue:43723", "right": "issue:44568", "accept": false, - "reason": "A general AutoTokenizer.from_pretrained v5 loading issue is not the same as missing BOS/EOS insertion for one tokenizer." + "reason": "AutoTokenizer load issue in v5 vs add_special_tokens BOS/EOS regression; different tokenizer failures." }, { "left": "issue:44016", "right": "issue:46231", "accept": false, - "reason": "A notebook syntax error and a documentation parameter-name mismatch are different documentation problems." + "reason": "Notebook syntax error vs documentation parameter-name mismatch; different artifact types." }, { "left": "issue:39692", "right": "issue:46909", "accept": false, - "reason": "Broken examples for different docs pages are not the same underlying issue." + "reason": "SigLIP2 example with model/processor mismatch and quantization failure vs broken DIA docs example; separate docs issues." }, { "left": "issue:42898", "right": "issue:44779", "accept": false, - "reason": "Both are v5 tokenizer regressions, but the reported failures and likely code paths differ." + "reason": "General tokenization cleanup regression vs Deepseek tokenizer codec regression; different concrete bugs." }, { "left": "issue:43479", "right": "issue:44079", "accept": false, - "reason": "A config-initialization bug is not the same as ModelOutput key assignment with None values." + "reason": "Multimodal config defaulting bug vs ModelOutput key assignment bug; unrelated internals." }, { "left": "issue:43784", "right": "issue:43824", "accept": false, - "reason": "A sentence-transformers import NameError and a missing Qwen2.5-VL symbol are unrelated import failures." + "reason": "sentence-transformers import NameError vs missing Qwen2.5-VL symbol export; different import failures." }, { "left": "issue:27957", "right": "issue:30333", "accept": false, - "reason": "FlashAttention2 model support and MLflow reporting on job failure are unrelated." + "reason": "Flash Attention 2 support for XLM-Roberta vs MLflow reporting after failed training; unrelated." }, { "left": "issue:42898", "right": "issue:43295", "accept": false, - "reason": "A tokenizer cleanup regression and a processor.tokenizer/image-passing regression are different v5 breakages." + "reason": "Both are v5 regressions, but one is token cleanup behavior and the other is custom processor/tokenizer access." }, { - "left": "issue:43874", - "right": "issue:45092", + "left": "issue:30333", + "right": "issue:34689", "accept": false, - "reason": "A missing image-processor method and a remote-code meta-init incompatibility are different multimodal loading issues." + "reason": "Failed-job MLflow tracking vs model loading breakage for Llama 3.2 Vision; different code paths." }, { - "left": "issue:30333", - "right": "issue:34689", + "left": "issue:43874", + "right": "issue:45092", "accept": false, - "reason": "MLflow job-state reporting and Llama 3.2 Vision model loading are unrelated." + "reason": "Missing image-patch helper in GLM46V processor vs InternVL2 meta-init incompatibility; different multimodal bugs." }, { "left": "issue:44734", "right": "issue:45103", "accept": false, - "reason": "KV-cache continuation indexing in serve and auto_docstring crashes under future annotations are unrelated." + "reason": "KV-cache continuation crash in serving vs auto_docstring annotation parsing crash; unrelated." }, { "left": "issue:42898", "right": "issue:45356", "accept": false, - "reason": "Both are tokenizer regressions, but one is cleanup-space behavior and the other is codec handling plus a warning issue in Kimi-K2.5." + "reason": "Tokenizer cleanup behavior change vs Kimi-K2.5 codec/warning regression; distinct tokenizer issues." }, { "left": "issue:42222", "right": "issue:45200", "accept": false, - "reason": "Broken vitpose model files and Gemma 4 mm_token_type_ids defaults are different multimodal defects." + "reason": "Broken VitPose model files vs Gemma 4 token-type defaulting; separate multimodal/model-loading defects." } ] }, @@ -14742,7 +14780,7 @@ { "label": "issue", "component_index": 1, - "component_count": 25, + "component_count": 26, "review_unit_index": 48, "review_unit_count": 61, "status": "reviewed", @@ -14811,12 +14849,12 @@ "issue:44075|issue:46231", "issue:44060|issue:45125", "issue:44933|issue:45712", - "issue:44448|issue:46132", "issue:42796|issue:43723", + "issue:44448|issue:46132", "issue:42831|issue:44623", "issue:44297|issue:45656", - "issue:44222|issue:44623", "issue:43479|issue:43644", + "issue:44222|issue:44623", "issue:45245|issue:46132", "issue:44360|issue:46762", "issue:46132|issue:46710", @@ -14831,23 +14869,23 @@ "issue:44190|issue:46082", "issue:44263|issue:45092", "issue:43232|issue:46424", - "issue:43232|issue:45834", - "issue:46209|issue:46431" + "issue:46209|issue:46431", + "issue:43232|issue:45834" ], - "prepared_review_unit_hash": "c791b384f09827f09cbcc540885dab89313b2c74b495232f515ece12271e505a", + "prepared_review_unit_hash": "ee6e7a63cf8275982f705feb69df3ad0176eb4797baa4ccda2b76b73216912a2", "component_budget": { "node_count": 505, "item_count": 505, "soft_pair_count": 1854, - "serialized_chars": 962354, - "estimated_input_tokens": 240589, - "estimated_eval_tokens": 481434 + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 31, - "serialized_chars": 36653, + "serialized_chars": 36655, "estimated_input_tokens": 9164, "estimated_eval_tokens": 18584 }, @@ -14856,186 +14894,192 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This set is highly heterogeneous: it mixes unrelated import errors, tokenizer/config regressions, training-example bugs, FP8/cache issues, and model-specific numerical problems. None of the proposed soft edges look like true duplicates of the same underlying bug or change.", - "confidence": 0.74, - "canonical_issue_reason": "No true duplicate cluster is evident here; if one issue must serve as a representative label, #43723 is the broadest user-facing loading regression (`AutoTokenizer.from_pretrained` in v5), but it does not subsume the other unrelated reports.", + "summary": "This is a loose similarity cluster, not a true duplicate set. The items mostly share broad themes like import errors, tokenizer/model save_pretrained behavior, or v5 regressions, but they describe different code paths and fixes. All soft edges should be rejected. No PRs are present.", + "confidence": 0.86, + "canonical_issue_reason": "issue:44297 is the most central concrete bug in the set and sits near several serialization-related similarities, but it is still too specific to serve as a real duplicate umbrella for the cluster.", "canonical_pr_reason": null, - "best_issue_reason": "#43723 is the clearest standalone, high-signal bug report with a public API failure and enough generality to act as a representative issue for the cluster, even though the cluster itself is not cohesive.", + "best_issue_reason": "issue:44297 is a well-scoped, actionable runtime bug with clear reproduction around tokenizer serialization, making it the best representative issue among these heterogeneous reports.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44295", "right": "issue:45656", "accept": false, - "reason": "Different bugs: buffer/position_ids read failure vs DeepSpeed calling optimizer.step twice." + "reason": "Different subsystems and failures: buffer access bug vs deepspeed optimizer stepping." }, { "left": "issue:30333", "right": "issue:36331", "accept": false, - "reason": "MLFlow reporting after failed training vs CustomTrainer signature mismatch are unrelated." + "reason": "Unrelated training-reporting/MLflow issue versus a custom trainer signature regression." }, { "left": "issue:45405", "right": "issue:45712", "accept": false, - "reason": "PEFT version pinning and leftover dummy classes are separate packaging/repo-maintenance issues." + "reason": "PEFT version pinning and dummy class cleanup are unrelated." }, { "left": "issue:22355", "right": "issue:28282", "accept": false, - "reason": "Both are import errors, but one is missing `transformers.onnx` and the other is missing PyTorch dependency handling." + "reason": "Both are import/dependency errors, but they concern different missing modules and causes." }, { "left": "issue:43116", "right": "issue:44190", "accept": false, - "reason": "Different example-script failures: empty multi-label outputs vs local dataset loading." + "reason": "Both touch example scripts, but one is multi-label classification output and the other is local dataset loading." }, { "left": "issue:44222", "right": "issue:44297", "accept": false, - "reason": "FP8 MoE save_pretrained bug and tokenizer_config class mismatch are different serialization paths." + "reason": "Both involve save_pretrained, but one is FP8 model serialization and the other is tokenizer config metadata." }, { "left": "issue:44075", "right": "issue:46231", "accept": false, - "reason": "Optimizer argument handling vs documentation parameter-name typo; not the same bug." + "reason": "Optimizer argument usage and a docstring parameter-name mismatch are not the same bug." }, { "left": "issue:44060", "right": "issue:45125", "accept": false, - "reason": "Incorrect tied-weights warning and missing tensor-parallel plan are distinct Qwen issues." + "reason": "Both are Qwen3_5Moe-related, but one is a tied-weights warning and the other is missing tensor-parallel plan support." }, { "left": "issue:44933", "right": "issue:45712", "accept": false, - "reason": "Missing import from `image_utils` vs dummy-class leakage in no-torch mode are different namespace problems." + "reason": "Different export/import hygiene problems with different targets and failure modes." }, { - "left": "issue:44448", - "right": "issue:46132", + "left": "issue:42796", + "right": "issue:43723", "accept": false, - "reason": "Pegasus output regression and AttentionInterface behavior change are unrelated." + "reason": "Tokenizer BPE-marker handling and tokenizer loading failures are distinct issues." }, { - "left": "issue:42796", - "right": "issue:43723", + "left": "issue:44448", + "right": "issue:46132", "accept": false, - "reason": "Tokenizer BPE-marker handling and tokenizer loading failure are distinct tokenizer bugs." + "reason": "Different regressions: model output drift versus attention registration behavior." }, { "left": "issue:42831", "right": "issue:44623", "accept": false, - "reason": "FP8 accuracy regression and processor save_pretrained missing files are unrelated." + "reason": "FP8 accuracy and processor save_pretrained file omission are unrelated." }, { "left": "issue:44297", "right": "issue:45656", "accept": false, - "reason": "Tokenizer config serialization mismatch is unrelated to optimizer double stepping." + "reason": "Tokenizer serialization metadata bug versus optimizer stepping bug; no shared code path." }, { - "left": "issue:44222", - "right": "issue:44623", + "left": "issue:43479", + "right": "issue:43644", "accept": false, - "reason": "Both mention save_pretrained, but they affect different objects and code paths." + "reason": "Config default initialization and non-persistent buffer corruption are separate problems." }, { - "left": "issue:43479", - "right": "issue:43644", + "left": "issue:44222", + "right": "issue:44623", "accept": false, - "reason": "Config default initialization and non-persistent-buffer junk are unrelated." + "reason": "Both mention save_pretrained, but they affect different objects and different serialization paths." }, { "left": "issue:45245", "right": "issue:46132", "accept": false, - "reason": "Category-limit runtime error and AttentionInterface registration behavior are different problems." + "reason": "Category-count validation is unrelated to attention registry behavior changes." }, { "left": "issue:44360", "right": "issue:46762", "accept": false, - "reason": "DSA indexer ReLU discussion and Minimax attention numerical differences are unrelated." + "reason": "Different model-specific numerical/architecture issues with no clear shared root cause." }, { "left": "issue:46132", "right": "issue:46710", "accept": false, - "reason": "Registered-function behavior change and model accuracy regression are not the same code-path bug." + "reason": "Both are regressions, but they affect different attention/model behavior and fixes." }, { "left": "issue:41628", "right": "issue:42757", "accept": false, - "reason": "Both are import failures, but they concern different symbols and different packages." + "reason": "Both are import errors, but for different symbols from different packages." }, { "left": "issue:44297", "right": "issue:45310", "accept": false, - "reason": "Tokenizer save_pretrained mismatch and Qwen3.5 MoE from_pretrained error are unrelated." + "reason": "Tokenizer save_pretrained metadata mismatch is not the same as Qwen3.5 MoE from_pretrained failure." }, { "left": "issue:44373", "right": "issue:44485", "accept": false, - "reason": "Docstring mismatch for position_ids and GLM-5 RoPE implementation are different issue types." + "reason": "Docstring mismatch is not the same as a RoPE implementation bug." }, { "left": "issue:43723", "right": "issue:44977", "accept": false, - "reason": "Tokenizer loading regression and flash-attention generation failure are unrelated." + "reason": "Tokenizer loading and flash-attention generation failures are unrelated." }, { "left": "issue:46614", "right": "issue:46932", "accept": false, - "reason": "Stale phonemizer backend state and Gemma4 dtype casting are different subsystems." + "reason": "Stale phonemizer backend state is unrelated to Gemma4 quantized embedder dtype casting." }, { "left": "issue:45071", "right": "issue:46082", "accept": false, - "reason": "Both are config-related, but type-checking breakage and explicit head_dim validation are distinct." + "reason": "General type-checking breakage and a specific head_dim validation edge case are different issues." }, { - "left": "issue:44190", - "right": "issue:46082", + "left": "issue:44222", + "right": "issue:44484", "accept": false, - "reason": "Local dataset loading and LlamaConfig validation are unrelated." + "reason": "A concrete serialization bug is not the same as a question about shard size defaults." + }, + { + "left": "issue:43122", + "right": "issue:46209", + "accept": false, + "reason": "Tokenizer output drift and FP8 inference failure are unrelated." }, { "left": "issue:44263", "right": "issue:45092", "accept": false, - "reason": "torch.split return handling in an indexer and remote-code meta-init incompatibility are unrelated." + "reason": "torch.split indexing behavior and meta-initialization incompatibility are different code paths." }, { "left": "issue:43232", "right": "issue:46424", "accept": false, - "reason": "Generation kwarg syncing and cache length semantics are different mechanisms." + "reason": "Generation kwargs updates and cache length config handling are not the same bug." }, { - "left": "issue:43232", - "right": "issue:45834", + "left": "issue:46209", + "right": "issue:46431", "accept": false, - "reason": "Generation-state update bug and Kosmos2.5 OCR index error are unrelated." + "reason": "FP8 inference on B200 and missing chat template validation are unrelated." }, { - "left": "issue:46209", - "right": "issue:46431", + "left": "issue:43232", + "right": "issue:45834", "accept": false, - "reason": "FP8 inference on B200 and missing chat template are unrelated." + "reason": "Generation sync behavior and OCR long-input indexing are unrelated." } ] }, @@ -15049,7 +15093,7 @@ { "label": "issue", "component_index": 1, - "component_count": 25, + "component_count": 26, "review_unit_index": 49, "review_unit_count": 61, "status": "reviewed", @@ -15113,10 +15157,10 @@ "issue:43475|issue:46305", "issue:46305|issue:46655", "issue:44295|issue:44623", - "issue:44263|issue:44373", - "issue:44263|issue:44360", "issue:46369|issue:46909", + "issue:44263|issue:44373", "issue:42907|issue:46050", + "issue:44263|issue:44360", "issue:45092|issue:45357", "issue:42831|issue:45310", "issue:44279|issue:46328", @@ -15124,10 +15168,10 @@ "issue:44483|issue:45406", "issue:22355|issue:27957", "issue:44829|issue:46002", - "issue:44261|issue:45092", - "issue:43644|issue:46305", "issue:41628|issue:43824", "issue:42831|issue:44038", + "issue:44261|issue:45092", + "issue:43644|issue:46305", "issue:43408|issue:46402", "issue:22355|issue:28218", "issue:42831|issue:45753", @@ -15142,20 +15186,20 @@ "issue:43232|issue:45593", "issue:45381|issue:46909" ], - "prepared_review_unit_hash": "58c44852a68a95789693990536cb23b2172bfd373f7ffe6191dbbfc180fa6117", + "prepared_review_unit_hash": "fcaeeb798fabb08fcb4a4a39b3c1a2d12aa8415ced280c8ad78f61d457120f60", "component_budget": { "node_count": 505, "item_count": 505, "soft_pair_count": 1854, - "serialized_chars": 962354, - "estimated_input_tokens": 240589, - "estimated_eval_tokens": 481434 + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 32, - "serialized_chars": 36803, + "serialized_chars": 36802, "estimated_input_tokens": 9201, "estimated_eval_tokens": 18658 }, @@ -15164,204 +15208,204 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly heterogeneous; the soft pairs share broad themes like imports, tokenizers, or save/load regressions, but none are clearly the same concrete bug or change. I would not merge any of the soft edges.", - "confidence": 0.79, - "canonical_issue_reason": "If one issue must anchor the set, 45278 is the broadest representative: it describes a generalized post-upgrade import regression rather than a narrow model-specific symptom.", - "canonical_pr_reason": null, - "best_issue_reason": "45278 is the best global representative because it is the most umbrella-like user-facing regression report in the group; the rest are narrower and mostly unrelated.", - "best_pr_reason": null, + "summary": "The set is heterogeneous and the soft-similarity links are mostly superficial (same subsystem or related theme, but not the same concrete bug/change). I would not merge any of the candidate pairs as duplicates.", + "confidence": 0.62, + "canonical_issue_reason": "issue:46305 is the most actionable single issue in the set: it describes a precise, reproducible off-by-one bug in a specific code path and is an active/open report, making it the best canonical target among these mostly unrelated issues.", + "canonical_pr_reason": "No pull requests are present in the provided items.", + "best_issue_reason": "issue:46305 is the strongest issue candidate overall because it is concrete, narrow, and clearly states the failure mode; it is a better duplicate anchor than the broader or more ambiguous reports here.", + "best_pr_reason": "No pull requests are present in the provided items.", "soft_edge_verdicts": [ { "left": "issue:30990", "right": "issue:38175", "accept": false, - "reason": "Different symptoms and domains: loading hang in Sentence Transformers vs zero probabilities in SigLIP2." + "reason": "Both are model/runtime problems, but one is a loading hang and the other is a zero-probability accuracy issue; they are not the same bug." }, { "left": "issue:43475", "right": "issue:46305", "accept": false, - "reason": "SAM3 vision-encoder attribute error vs MoonshineStreaming window off-by-one; unrelated bugs." + "reason": "These affect different SAM3 vs MoonshineStreaming code paths and describe unrelated failures." }, { "left": "issue:46305", "right": "issue:46655", "accept": false, - "reason": "Both are video-processing issues, but one is a streaming window bug and the other is a padding-calculation bug." + "reason": "Both mention video-related processing, but one is a streaming mask off-by-one and the other is a video padding calculation bug." }, { "left": "issue:44295", "right": "issue:44623", "accept": false, - "reason": "One is a buffer read/position_ids issue; the other is missing files on processor save. Different code paths." + "reason": "Registering position_ids as a buffer and missing files on processor save_pretrained are distinct issues." }, { - "left": "issue:44263", - "right": "issue:44373", + "left": "issue:46369", + "right": "issue:46909", "accept": false, - "reason": "DSA indexer torch.split behavior vs a wrong docstring; the docstring issue is purely documentation." + "reason": "Memory overflow in sam3-litetext and a broken DIA documentation example are unrelated." }, { "left": "issue:44263", - "right": "issue:44360", + "right": "issue:44373", "accept": false, - "reason": "Different concrete problems in the DSA indexer: split output handling vs missing ReLU." + "reason": "A torch.split indexing bug and a wrong docstring are not the same underlying problem." }, { - "left": "issue:46369", - "right": "issue:46909", + "left": "issue:42907", + "right": "issue:46050", "accept": false, - "reason": "Memory overflow in sam3-litetext vs a broken docs example for DIA; not the same bug." + "reason": "Both concern saving quantized models, but they target different quantization flows and failure modes." }, { - "left": "issue:42907", - "right": "issue:46050", + "left": "issue:44263", + "right": "issue:44360", "accept": false, - "reason": "Both involve quantized model saving/loading, but they target different quantization flows and failure modes." + "reason": "A return-value mismatch in DSA indexing is different from the missing ReLU discussion." }, { "left": "issue:45092", "right": "issue:45357", "accept": false, - "reason": "Remote-code/meta-init incompatibility vs incorrect visual encoder keys on save_pretrained; different regressions." + "reason": "Meta-initialization incompatibility for old InternVL2 checkpoints is not the same as incorrect visual encoder keys in Qwen3.5 save_pretrained." }, { "left": "issue:42831", "right": "issue:45310", "accept": false, - "reason": "FineGrainedFP8 accuracy degradation vs Qwen3.5 from_pretrained error; unrelated issues." + "reason": "FineGrainedFP8 accuracy drift and Qwen3.5 from_pretrained errors are different bugs." }, { "left": "issue:44279", "right": "issue:46328", "accept": false, - "reason": "Generic dependency issue vs a SequenceFeatureExtractor.pad performance problem; not the same bug." + "reason": "A general dependency issue is unrelated to a performance inefficiency in SequenceFeatureExtractor.pad." }, { "left": "issue:45381", "right": "issue:46305", "accept": false, - "reason": "Qwen2.5-VL video vision_position_ids bug vs MoonshineStreaming mask off-by-one; unrelated models and paths." + "reason": "They are both off-by-one style bugs, but in different models and different masking/positioning logic." }, { "left": "issue:44483", "right": "issue:45406", "accept": false, - "reason": "Chat-completions request rejection vs Gemma4Processor attribute crash; different surfaces and causes." + "reason": "An API request rejection and a Gemma4Processor tokenizer attribute crash are unrelated." }, { "left": "issue:22355", "right": "issue:27957", "accept": false, - "reason": "Module import failure for transformers.onnx vs an XLMRoberta Flash Attention 2 issue; unrelated." + "reason": "Missing transformers.onnx and XLMRoberta Flash Attention 2 are completely different issues." }, { "left": "issue:44829", "right": "issue:46002", "accept": false, - "reason": "Degenerate training with flash_attention_3 vs NaNs after checkout/init in classification heads; not the same failure." + "reason": "Flash Attention 3 training degeneration and NaNs after checkout/init are different failure modes." }, { - "left": "issue:44261", - "right": "issue:45092", + "left": "issue:41628", + "right": "issue:43824", "accept": false, - "reason": "MLA layernorm epsilon precision bug vs meta-initialization incompatibility for remote-code checkpoints." + "reason": "These are separate missing-import cases for different symbols, not one concrete bug." }, { - "left": "issue:43644", - "right": "issue:46305", + "left": "issue:42831", + "right": "issue:44038", "accept": false, - "reason": "Non-persistent buffer junk in Transformers 5.0 vs a streaming mask off-by-one; different subsystems." + "reason": "A FineGrainedFP8 accuracy report and a Qwen3-VL-Moe bug report are too different to be duplicates." }, { - "left": "issue:41628", - "right": "issue:43824", + "left": "issue:44261", + "right": "issue:45092", "accept": false, - "reason": "Both are import errors, but for different missing symbols and likely different fixes; too broad to treat as duplicate." + "reason": "Missing rms_norm_eps in MLA q_a_layernorm and remote-code meta-init incompatibility are distinct root causes." }, { - "left": "issue:42831", - "right": "issue:44038", + "left": "issue:43644", + "right": "issue:46305", "accept": false, - "reason": "FineGrainedFP8 accuracy issue vs Qwen3-VL-Moe bug; unrelated." + "reason": "Non-persistent buffer junk in Transformers 5.0.0 is unrelated to a streaming mask off-by-one bug." }, { "left": "issue:43408", "right": "issue:46402", "accept": false, - "reason": "Model-type warning on instantiation vs missing text-encoder weights load; different SAM3 issues." + "reason": "A model-type warning and missing text encoder weight loading are different problems." }, { "left": "issue:22355", "right": "issue:28218", "accept": false, - "reason": "Import error for transformers.onnx vs tokenizer adding an extra space; unrelated." + "reason": "Importing transformers.onnx and tokenizer spacing after added tokens are unrelated." }, { "left": "issue:42831", "right": "issue:45753", "accept": false, - "reason": "FP8 accuracy regression vs Qwen3_5 infinite loop on an image; different bug classes." + "reason": "FP8 accuracy issues are not the same as Qwen3_5 looping on a specific image." }, { "left": "issue:39401", "right": "issue:44568", "accept": false, - "reason": "Wrong offset mapping in Qwen3 tokenizer vs add_special_tokens not adding BOS/EOS in mDeBERTa; different tokenizer behaviors." + "reason": "Wrong offset_mapping and missing BOS/EOS insertion are both tokenizer bugs, but not the same concrete defect." }, { "left": "issue:42886", "right": "issue:43066", "accept": false, - "reason": "Offline cache load failure vs wrong tokenizer decoder type in v5; different failure modes." + "reason": "Offline cache loading failure and wrong tokenizer decoder type are different issues." }, { "left": "issue:42886", "right": "issue:42994", "accept": false, - "reason": "Tokenizer offline cache bug vs quantized model saving failure; unrelated." + "reason": "Tokenizer offline-cache loading and quantized model saving are unrelated." }, { "left": "issue:42757", "right": "issue:45278", "accept": false, - "reason": "Specific missing `is_offline_mode` import is not the same concrete bug as the broader 'many import errors after update' report." + "reason": "A missing huggingface_hub symbol and broad import errors after upgrade overlap only at a very high level, not as one concrete bug." }, { "left": "issue:27957", "right": "issue:30990", "accept": false, - "reason": "Flash Attention 2 on XLMRoberta vs Sentence Transformers loading hang; unrelated." + "reason": "Flash Attention 2 with XLMRoberta and a Sentence Transformers loading stall are distinct." }, { "left": "issue:43408", "right": "issue:46369", "accept": false, - "reason": "SAM3 type warning vs memory overflow in sam3-litetext; different symptoms and likely causes." + "reason": "SAM3 model-type mismatch and SAM3-litetext memory overflow are different failures." }, { "left": "issue:44188", "right": "issue:44373", "accept": false, - "reason": "torch.compile attention-kernel divergence vs a wrong docstring; not comparable bugs." + "reason": "A torch.compile attention-kernel divergence bug is unrelated to a docstring issue." }, { "left": "issue:42796", "right": "issue:43122", "accept": false, - "reason": "Tokenizer BPE-marker handling bug vs general tokenization changes across versions; not the same underlying defect." + "reason": "Ministral BPE-marker handling and generic tokenization differences across versions are related themes but not the same bug." }, { "left": "issue:43232", "right": "issue:45593", "accept": false, - "reason": "Generation kwarg update bug after sync_gpus vs D-FINE auxiliary-loss behavior; unrelated." + "reason": "Generation kwargs after sync_gpus and D-FINE auxiliary-loss behavior are unrelated." }, { "left": "issue:45381", "right": "issue:46909", "accept": false, - "reason": "Qwen2.5-VL video position IDs bug vs DIA documentation example breakage; no overlap." + "reason": "A Qwen2.5-VL video position_ids bug and a DIA documentation example problem are not duplicates." } ] }, @@ -15375,7 +15419,7 @@ { "label": "issue", "component_index": 1, - "component_count": 25, + "component_count": 26, "review_unit_index": 50, "review_unit_count": 61, "status": "reviewed", @@ -15444,20 +15488,20 @@ "issue:43756|issue:44060", "issue:39401|issue:44488", "issue:44295|issue:44297", - "issue:42371|issue:44263", "issue:41628|issue:45278", "issue:43576|issue:43784", + "issue:42371|issue:44263", "issue:42907|issue:43937", - "issue:46172|issue:46762", "issue:43784|issue:45600", + "issue:46172|issue:46762", "issue:44117|issue:44704", "issue:39401|issue:44521", "issue:44625|issue:44829", "issue:44704|issue:45290", "issue:43122|issue:44521", "issue:43408|issue:43475", - "issue:43122|issue:46326", "issue:44360|issue:44485", + "issue:43122|issue:46326", "issue:42907|issue:43452", "issue:44038|issue:44297", "issue:22355|issue:29159", @@ -15467,220 +15511,220 @@ "issue:44038|issue:44222", "issue:43824|issue:44351" ], - "prepared_review_unit_hash": "f423b1d143ffeb4d4515cd6f29dd161b06ff550674a148b63a98dde97adeff0f", + "prepared_review_unit_hash": "38dc7670a268f95145de96930ea5ee3fc62cf79da15636095a40f36bab423d87", "component_budget": { "node_count": 505, "item_count": 505, "soft_pair_count": 1854, - "serialized_chars": 962354, - "estimated_input_tokens": 240589, - "estimated_eval_tokens": 481434 + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 31, - "serialized_chars": 36536, - "estimated_input_tokens": 9134, - "estimated_eval_tokens": 18524 + "serialized_chars": 36529, + "estimated_input_tokens": 9133, + "estimated_eval_tokens": 18522 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly a grab-bag of unrelated Transformers bugs and regressions (imports, tokenizers, model-specific failures, serialization, multimodal/chat-template issues). None of the soft pairs look like the same underlying defect, so all soft edges are rejected.", - "confidence": 0.64, - "canonical_issue_reason": "issue:45278 is the broadest, most umbrella-like report in the set: many import errors after upgrading Transformers. It best matches the cluster\u2019s import-regression subset, even though the overall cluster is heterogeneous.", + "summary": "The cluster is mostly a collection of unrelated bug reports that only look superficially similar by model family, version, or subsystem. None of the soft pairs appear to describe the same concrete bug/change, so I would not merge any of them as duplicates. The broadest umbrella-style report is the 5.5.0 import-regression issue.", + "confidence": 0.93, + "canonical_issue_reason": "issue:45278 is the closest thing to a canonical issue because it is the broadest umbrella report in the set: a general import-regression after upgrading to 5.5.0, which is more representative than the model-specific or single-symptom tickets.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45278 is the best single issue for triage because it is broad, user-facing, and likely to attract related import-regression reports; however, it should not be treated as a duplicate hub for the whole cluster.", + "best_issue_reason": "issue:45278 is the best overall representative because it describes a wider failure mode and could subsume related import-breakage reports better than the narrower Qwen/SAM/tokenizer issues.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43540", "right": "issue:46909", "accept": false, - "reason": "One is a Qwen3OmniMoe video-input ValueError; the other is a broken documentation example for DIA. Different area and failure mode." + "reason": "Different bugs: Qwen3OmniMoe video-input validation vs a broken DIA documentation example." }, { "left": "issue:43646", "right": "issue:46002", "accept": false, - "reason": "Custom model initialization breakage and NaNs in classification heads are different bugs with different symptoms and likely different code paths." + "reason": "Both concern model initialization, but one is a Transformers 5.0 custom-model init breakage and the other is NaNs after checkout/init; not the same code-path bug." }, { "left": "issue:30990", "right": "issue:33357", "accept": false, - "reason": "Sentence-Transformers loading hangs and a MacOS bus error with a CLIP model are unrelated runtime failures." + "reason": "Unrelated symptoms: Sentence-Transformers loading hang vs a MacOS bus error with a CLIP model." }, { "left": "issue:42907", "right": "issue:44945", "accept": false, - "reason": "Saving dequantized Ministrals/Devstrals and incorrect output under pipeline parallelism are unrelated issues." + "reason": "Different model/pathology: dequantized save failure vs incorrect output under pipeline parallelism." }, { "left": "issue:42907", "right": "issue:44964", "accept": false, - "reason": "Dequantized save failure and inability to load Phi-4 multimodal instruct are different problems." + "reason": "One is about saving dequantized Ministrals/Devstrals, the other about loading Phi-4 multimodal; no concrete overlap." }, { "left": "issue:43976", "right": "issue:44945", "accept": false, - "reason": "Python version incompatibility and wrong LLM output under pipeline parallelism do not share the same underlying bug." + "reason": "Python-version compatibility breakage is not the same bug as pipeline-parallel output corruption." }, { "left": "issue:43756", "right": "issue:44060", "accept": false, - "reason": "These are both model-specific, but one is about dropped RoPE layers and the other about a tied-weights warning; not the same defect." + "reason": "Different model implementations and defects: Smollm3 RoPE-layer loss vs Qwen3-Next tied-weights warning." }, { "left": "issue:39401", "right": "issue:44488", "accept": false, - "reason": "Qwen3 tokenizer offset mapping and failure to load cjvt/sleng-bert are unrelated tokenizer/model issues." + "reason": "Qwen3 tokenizer offset_mapping bug is unrelated to inability to load cjvt/sleng-bert." }, { "left": "issue:44295", "right": "issue:44297", "accept": false, - "reason": "A position_ids buffer read error and tokenizer_class mismatch on save_pretrained are different serialization bugs." - }, - { - "left": "issue:42371", - "right": "issue:44263", - "accept": false, - "reason": "TF32 API guidance and a torch.split return-value issue in GlmMoeDsaIndexer are unrelated." + "reason": "One is a position_ids buffer access problem, the other is tokenizer_class metadata mismatch on save_pretrained." }, { "left": "issue:41628", "right": "issue:45278", "accept": false, - "reason": "Both involve imports, but one is a specific AutoImageProcessor import failure while the other is a broader multi-import regression after upgrade; not enough to call them the same bug." + "reason": "Both involve imports, but one is a specific missing AutoImageProcessor import and the other is a broad post-upgrade import-regression report; not the same concrete defect." }, { "left": "issue:43576", "right": "issue:43784", "accept": false, - "reason": "Broken env command in v5 and a NameError when importing sentence-transformers are different failures." + "reason": "Broken CLI/env command vs an import NameError in sentence-transformers; different failure modes." }, { - "left": "issue:42907", - "right": "issue:43937", + "left": "issue:42371", + "right": "issue:44263", "accept": false, - "reason": "Saving dequantized Ministrals/Devstrals and invalid GenerationConfig for GLM-5 are unrelated." + "reason": "TF32 API guidance issue and a torch.split return-value complaint in a DSA indexer are unrelated." }, { - "left": "issue:46172", - "right": "issue:46762", + "left": "issue:42907", + "right": "issue:43937", "accept": false, - "reason": "The DeepSeek-V4 tensor operation bug and Minimax MSA attention numerical mismatch are different model/code-path problems." + "reason": "Dequantized save failure and invalid GenerationConfig in GLM-5 are separate bugs." }, { "left": "issue:43784", "right": "issue:45600", "accept": false, - "reason": "Import NameError with sentence-transformers and stale auto_mappings references to removed Sam3LiteText configs are separate breakages." + "reason": "Sentence-transformers import NameError vs stale auto_mappings references to removed SAM3 configs; different root causes." + }, + { + "left": "issue:46172", + "right": "issue:46762", + "accept": false, + "reason": "Both are numerical/model correctness complaints, but they target different models and code paths." }, { "left": "issue:44117", "right": "issue:44704", "accept": false, - "reason": "TOKENIZER_MAPPING_NAMES returning None and AutoProcessor not passing kwargs to cached_file are different from_pretrained-related issues." + "reason": "TOKENIZER_MAPPING_NAMES None-handling is not the same as AutoProcessor not forwarding kwargs to cached_file." }, { "left": "issue:39401", "right": "issue:44521", "accept": false, - "reason": "Wrong offset_mapping in Qwen3 tokenizer and zero assistant masks for multimodal chat templates are not the same bug." + "reason": "Qwen3 tokenizer offset mapping and multimodal assistant_mask generation are different tokenizer/template bugs." }, { "left": "issue:44625", "right": "issue:44829", "accept": false, - "reason": "num_labels propagation in Qwen3.5 and degenerate training with flash_attention_3 are unrelated." + "reason": "Qwen3.5 num_labels propagation and flash_attention_3 degenerate training are unrelated." }, { "left": "issue:44704", "right": "issue:45290", "accept": false, - "reason": "Missing kwargs in AutoProcessor.from_pretrained and tool-call chat-template crashes are different code paths." + "reason": "AutoProcessor kwargs forwarding and chat-template crashes with tool-call messages are separate code paths." }, { "left": "issue:43122", "right": "issue:44521", "accept": false, - "reason": "Version-dependent tokenization changes and multimodal assistant-mask failures are not the same underlying issue." + "reason": "Version-to-version tokenization differences are not the same bug as multimodal assistant_masks being all zeros." }, { "left": "issue:43408", "right": "issue:43475", "accept": false, - "reason": "SAM3 video/model-type warning and missing fpn_position_embeddings are distinct SAM3 bugs." + "reason": "Both are SAM 3 Video related, but one is a model-type warning and the other is a missing attribute on encoder output; different defects." }, { - "left": "issue:43122", - "right": "issue:46326", + "left": "issue:44360", + "right": "issue:44485", "accept": false, - "reason": "General tokenization changes across versions and continue_final_message behavior for mistral_common tokenizers are different defects." + "reason": "DSA ReLU discussion and GLM-5 RoPE implementation are not the same issue." }, { - "left": "issue:44360", - "right": "issue:44485", + "left": "issue:43122", + "right": "issue:46326", "accept": false, - "reason": "A DSA indexer ReLU discussion and GLM-5 RoPE implementation issue are unrelated model-internal changes." + "reason": "General tokenization changes across versions are not the same as the mistral_common continue_final_message bug." }, { "left": "issue:42907", "right": "issue:43452", "accept": false, - "reason": "Dequantized-save failure and gguf_file breaking AutoTokenizer/AutoModelForCausalLM loading are different bugs." + "reason": "Dequantized save failures and gguf_file breaking from_pretrained are different loading/saving paths." }, { "left": "issue:44038", "right": "issue:44297", "accept": false, - "reason": "Qwen3-VL-Moe issues and tokenizer_class mismatch on save_pretrained are unrelated." + "reason": "Qwen3-VL-Moe bug and tokenizer_class metadata mismatch are unrelated." }, { "left": "issue:22355", "right": "issue:29159", "accept": false, - "reason": "Missing transformers.onnx module and inconsistent slow/fast tokenizer behavior are different, unrelated problems." + "reason": "ONNX missing-module import error and slow-vs-fast tokenizer inconsistency are unrelated." }, { "left": "issue:45216", "right": "issue:45478", "accept": false, - "reason": "Qwen3.5 save_pretrained checkpoint corruption and Qwen3.5 from_pretrained failure are both around the same model family, but they describe different operations and distinct failures." + "reason": "Both mention Qwen3.5 and 5.4.0+, but one is a bad saved checkpoint and the other is a from_pretrained error; not enough evidence for the same concrete bug." }, { "left": "issue:44038", "right": "issue:44295", "accept": false, - "reason": "Qwen3-VL-Moe regressions and a position_ids buffer error are unrelated." + "reason": "Qwen3-VL-Moe bug and position_ids buffer access bug do not share the same failure mode." }, { "left": "issue:39401", "right": "issue:46614", "accept": false, - "reason": "Qwen3 tokenizer offset mapping and stale espeak backend selection in Wav2Vec2PhonemeCTCTokenizer are unrelated." + "reason": "Qwen3 tokenizer offset mapping and stale phonemizer language state are unrelated tokenizer bugs." }, { "left": "issue:44038", "right": "issue:44222", "accept": false, - "reason": "Qwen3-VL-Moe and FP8 save_pretrained for MoE share only a broad MoE theme, not the same concrete bug." + "reason": "Qwen3-VL-Moe loading issue and FP8 save_pretrained moe issue are separate model/save-path problems." }, { "left": "issue:43824", "right": "issue:44351", "accept": false, - "reason": "These are both import errors, but for different exported symbols/models; not the same underlying defect." + "reason": "Both are import errors, but they involve different missing symbols and likely different export/mapping problems." } ] }, @@ -15694,7 +15738,7 @@ { "label": "issue", "component_index": 1, - "component_count": 25, + "component_count": 26, "review_unit_index": 51, "review_unit_count": 61, "status": "reviewed", @@ -15767,30 +15811,30 @@ "issue:42898|issue:46614", "issue:38617|issue:45278", "issue:43937|issue:44485", - "issue:43066|issue:44779", "issue:42673|issue:45072", + "issue:43066|issue:44779", "issue:43611|issue:46002", "issue:44038|issue:45478", "issue:43976|issue:45446", "issue:45357|issue:45478", - "issue:44355|issue:46231", "issue:43116|issue:43493", + "issue:44355|issue:46231", "issue:44625|issue:46002", "issue:44610|issue:46909", - "issue:42371|issue:43874", "issue:45083|issue:45446", + "issue:42371|issue:43874", "issue:46726|issue:46981", "issue:44075|issue:44936", "issue:43618|issue:46829" ], - "prepared_review_unit_hash": "ad9367acf7b433a520236f7dc23073dbb5a95c8f73bd2db8cd2946023a338a5f", + "prepared_review_unit_hash": "68acbf5715a9611cb1a95b651d59e1f313caa3b71d83cd8b1fa1acd0aab14c80", "component_budget": { "node_count": 505, "item_count": 505, "soft_pair_count": 1854, - "serialized_chars": 962354, - "estimated_input_tokens": 240589, - "estimated_eval_tokens": 481434 + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 }, "budget": { "node_count": 48, @@ -15805,180 +15849,180 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly heterogeneous: the issues span unrelated import errors, tokenizer regressions, multimodal/model-specific bugs, training/runtime problems, and documentation nitpicks. None of the soft pairs look like the same underlying bug or change.", - "confidence": 0.93, - "canonical_issue_reason": "No single issue cleanly represents the cluster because the items are not duplicates of one another. If an umbrella representative is needed, issue:45278 is the broadest import-regression report, but it is not a true canonical duplicate target.", + "summary": "This cluster is a loose collection of mostly unrelated Transformers issues with some superficial overlap around v5 regressions, imports, tokenization, and model-specific bugs. I did not find any soft-edge pair that is confidently the same underlying defect.", + "confidence": 0.57, + "canonical_issue_reason": "issue:45278 is the broadest and most umbrella-like report in the set, covering multiple import failures after upgrading to 5.5.0.", "canonical_pr_reason": null, - "best_issue_reason": "Issue:45278 is the best umbrella issue because it is broad, regression-oriented, and closest to grouping the import-related subset. Still, the cluster is too diverse for a strong canonical choice.", + "best_issue_reason": "issue:45278 is the best representative issue for this cluster because it is general, recent, and plausibly subsumes several import-related regression reports.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43116", "right": "issue:45200", "accept": false, - "reason": "Different bugs: example script multi-label output vs Gemma 4 multimodal token defaults." + "reason": "Different bugs: example script multi-label output handling vs Gemma 4 text-only mm_token_type_ids defaulting." }, { "left": "issue:43122", "right": "issue:44779", "accept": false, - "reason": "Both are tokenizer-related, but one is version-wide tokenization drift and the other is a DeepSeek-specific tokenizer regression." + "reason": "Both are tokenizer regressions, but they describe different models and different failure modes; not enough evidence of one underlying bug." }, { "left": "issue:42757", "right": "issue:44351", "accept": false, - "reason": "Both are import errors, but for different missing symbols from different packages/modules." + "reason": "Both are import errors, but for different symbols and different modules; likely separate breakages." }, { "left": "issue:43278", "right": "issue:44945", "accept": false, - "reason": "Training/eval dtype mismatch is unrelated to incorrect outputs under pipeline parallelism." + "reason": "BF16/FP32 dtype drift in evaluation is unrelated to incorrect LLM output under pipeline parallelism." }, { "left": "issue:36296", "right": "issue:43116", "accept": false, - "reason": "Tensor-parallel training bug and multi-label classification output bug affect different code paths." + "reason": "Tensor-parallel training bug and multi-label classification output issue are unrelated." }, { "left": "issue:22355", "right": "issue:29127", "accept": false, - "reason": "Unrelated import failure vs LayoutLMv3 error-message clarity issue." + "reason": "One is a missing module import, the other is a LayoutLMv3 error message clarity issue." }, { "left": "issue:42175", "right": "issue:44938", "accept": false, - "reason": "Different platform/install problems: missing TensorFlow backend in a torch install vs Python 3.14 load failures." + "reason": "Environment/backend packaging issue vs Python 3.14 load failure; different problems." }, { "left": "issue:43493", "right": "issue:45753", "accept": false, - "reason": "Both mention Qwen3.5-family models, but one is SigLIP2 implementation mismatch and the other is an image-triggered infinite loop." + "reason": "SigLIP2 implementation discrepancy is not the same as Qwen3_5 infinite-loop behavior on a specific image." }, { "left": "issue:46231", "right": "issue:46877", "accept": false, - "reason": "Documentation parameter-name mismatch is unrelated to noisy stdout from auto_docstring import-time diagnostics." + "reason": "Documentation parameter mismatch vs docstring tooling printing diagnostics at import time; related area, but not the same bug." }, { "left": "issue:44188", "right": "issue:46328", "accept": false, - "reason": "Attention-kernel divergence under torch.compile is unrelated to SequenceFeatureExtractor.pad performance overhead." + "reason": "Attention-kernel divergence under torch.compile is unrelated to SequenceFeatureExtractor.pad performance." }, { "left": "issue:42898", "right": "issue:46614", "accept": false, - "reason": "Tokenization-spaces behavior change in v5 is unrelated to stale espeak language backend state." + "reason": "General tokenization-space behavior change and stale phonemizer backend state are different tokenizer issues." }, { "left": "issue:38617", "right": "issue:45278", "accept": false, - "reason": "Different import regressions involving different symbols and upgrade contexts; not the same failure." + "reason": "Specific missing import and broad import-error regression may be related in theme, but not the same concrete code-path problem." }, { "left": "issue:43937", "right": "issue:44485", "accept": false, - "reason": "Invalid GenerationConfig for GLM-5 is a different bug from GLM-5 RoPE implementation discussion." + "reason": "GenerationConfig validation failure is distinct from GLM-5 RoPE implementation discussion." }, { - "left": "issue:43066", - "right": "issue:44779", + "left": "issue:42673", + "right": "issue:45072", "accept": false, - "reason": "Wrong tokenizer decoder type in v5 is not the same as DeepSeek tokenizer output differences." + "reason": "VRAM leak in dataloader-threaded Qwen3ForCausalLM is unrelated to bfloat16 dtype mismatches in CI." }, { - "left": "issue:42673", - "right": "issue:45072", + "left": "issue:43066", + "right": "issue:44779", "accept": false, - "reason": "VRAM leak in multi-threaded Qwen3ForCausalLM is unrelated to dtype mismatch CI failures." + "reason": "Wrong decoder-type regression and Deepseek tokenizer incorrect output are too indirect to treat as the same bug." }, { "left": "issue:43611", "right": "issue:46002", "accept": false, - "reason": "base_model_prefix loading regression is unrelated to NaNs in classification heads after checkout/init." + "reason": "Base-model-prefix loading breakage is unrelated to NaNs in classification heads after checkout/init." }, { "left": "issue:44038", "right": "issue:45478", "accept": false, - "reason": "Both concern Qwen3.5/Qwen3.5-Moe, but the concrete failure modes differ: one is a general bug report, the other is a from_pretrained error." + "reason": "Both mention Qwen Moe models, but they target different model variants and different load-time failures." }, { "left": "issue:43976", "right": "issue:45446", "accept": false, - "reason": "Python version compatibility issue is unrelated to the AuxRequest import version check bug." + "reason": "Python version support regression is unrelated to the PyTorch version check for AuxRequest import." }, { "left": "issue:45357", "right": "issue:45478", "accept": false, - "reason": "Saving incorrect visual encoder keys and from_pretrained load errors are different code paths and failure modes." + "reason": "Saving incorrect visual encoder keys and from_pretrained load errors are different directions of failure with no clear shared code-path." }, { - "left": "issue:44355", - "right": "issue:46231", + "left": "issue:43116", + "right": "issue:43493", "accept": false, - "reason": "Compiled Python file errors are unrelated to a docstring parameter-name documentation mismatch." + "reason": "Multi-label classification example behavior and SigLIP2 implementation mismatch are unrelated." }, { - "left": "issue:43116", - "right": "issue:43493", + "left": "issue:44355", + "right": "issue:46231", "accept": false, - "reason": "Multi-label example script output bug and SigLIP2 model discrepancy are unrelated." + "reason": "Running compiled Python files errors and a documentation parameter name mismatch are not the same issue." }, { "left": "issue:44625", "right": "issue:46002", "accept": false, - "reason": "Qwen3.5 num_labels propagation bug is unrelated to NaNs in classification heads." + "reason": "num_labels propagation bug in Qwen3.5 is unrelated to NaNs in classification heads." }, { "left": "issue:44610", "right": "issue:46909", "accept": false, - "reason": "OmDet-Turbo processor input-size mismatch is unrelated to a broken documentation example for DIA." + "reason": "Processor input-size mismatch and a broken documentation example are different kinds of defects." }, { - "left": "issue:42371", - "right": "issue:43874", + "left": "issue:45083", + "right": "issue:45446", "accept": false, - "reason": "TF32 behavior settings warning and missing image-patch method are different issues." + "reason": "Feature-extractor length helper behavior and a version-check import bug are unrelated." }, { - "left": "issue:45083", - "right": "issue:45446", + "left": "issue:42371", + "right": "issue:43874", "accept": false, - "reason": "Feature-length helper behavior in qwen3_omni_moe is unrelated to PyTorch version gating for AuxRequest." + "reason": "TF32 API guidance and missing image-patch-count method are unrelated." }, { "left": "issue:46726", "right": "issue:46981", "accept": false, - "reason": "Missing vision-backend error messaging is unrelated to repeated HTTP 429 retry behavior in AutoProcessor.from_pretrained." + "reason": "Missing-vision-backend error handling and repeated 429 retries under HF_ENDPOINT are different AutoProcessor behaviors." }, { "left": "issue:44075", "right": "issue:44936", "accept": false, - "reason": "SGD optimizer args not used is unrelated to trainer.evaluate failing after trainer.train." + "reason": "Unused SGD args and trainer.evaluate() failing after train() are not the same bug." }, { "left": "issue:43618", "right": "issue:46829", "accept": false, - "reason": "Missing CLIPOutput attentions assignment is not the same as VideoPrism hidden_states being populated incorrectly." + "reason": "Both are output-field assignment problems, but they affect different model outputs and different fields; too weak to merge." } ] }, @@ -15992,7 +16036,7 @@ { "label": "issue", "component_index": 1, - "component_count": 25, + "component_count": 26, "review_unit_index": 52, "review_unit_count": 61, "status": "reviewed", @@ -16053,8 +16097,8 @@ ], "soft_pairs": [ "issue:43116|issue:43825", - "issue:43066|issue:46710", "issue:43992|issue:44704", + "issue:43066|issue:46710", "issue:46614|issue:46752", "issue:44995|issue:46697", "issue:28348|issue:46640", @@ -16064,14 +16108,14 @@ "issue:43827|issue:45446", "issue:33453|issue:39401", "issue:43064|issue:43673", - "issue:42222|issue:43441", "issue:43937|issue:45563", "issue:43824|issue:46981", + "issue:42222|issue:43441", + "issue:42673|issue:46655", "issue:42222|issue:47000", "issue:43673|issue:46328", - "issue:42673|issue:46655", - "issue:44291|issue:45071", "issue:45405|issue:45446", + "issue:44291|issue:45071", "issue:42222|issue:46752", "issue:44117|issue:45588", "issue:44380|issue:46077", @@ -16082,202 +16126,208 @@ "issue:43450|issue:43986", "issue:45071|issue:46512" ], - "prepared_review_unit_hash": "34f79a3b774b05273a20ad6fb19f92cd9debf0d967108857db5a1626965abf48", + "prepared_review_unit_hash": "b08157c483c3ca8d759d5859e63185d82a5eef3bf0b410536255191c81d2ef03", "component_budget": { "node_count": 505, "item_count": 505, "soft_pair_count": 1854, - "serialized_chars": 962354, - "estimated_input_tokens": 240589, - "estimated_eval_tokens": 481434 + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 29, - "serialized_chars": 36284, - "estimated_input_tokens": 9071, - "estimated_eval_tokens": 18398 + "serialized_chars": 36278, + "estimated_input_tokens": 9070, + "estimated_eval_tokens": 18396 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This is a heterogeneous cluster of Transformers issues spanning tokenizers, processors, attention backends, loading/config, and training. None of the proposed soft pairs looks like the same concrete bug or change, so I would not collapse them together.", - "confidence": 0.94, - "canonical_issue_reason": "None of these is a safe canonical; the items span distinct bugs, so I would not merge the cluster around a single issue.", + "summary": "No soft pair looks like the same underlying bug or change. Most are only loosely related by subsystem (tokenizers, generation, flash-attn, v5 regressions) but point to distinct code paths or model-specific failures, so they should stay separate issues.", + "confidence": 0.87, + "canonical_issue_reason": "issue:33453 is the broadest umbrella-style report in the set ('Regression in tokenizer loading') and is the most plausible canonical representative among the issue-only cluster.", "canonical_pr_reason": null, - "best_issue_reason": "If forced, 33453 is the broadest umbrella for tokenizer-loading regressions, but it is still too generic to represent the rest of the cluster well.", + "best_issue_reason": "issue:33453 is the best global representative because it is broad, concrete, and names the common failure class without overfitting to a single model.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43116", "right": "issue:43825", "accept": false, - "reason": "Different failures: a multi-label example script bug vs a v5 translation-task error message issue." + "reason": "Both touch example/pipeline behavior, but one is a multi-label classification script bug and the other is a v5 pipeline translation-message mismatch." }, { - "left": "issue:43066", - "right": "issue:46710", + "left": "issue:43992", + "right": "issue:44704", "accept": false, - "reason": "Tokenizer decoder-type regression and a DeepSeek accuracy regression are different concrete problems." + "reason": "Both concern loading APIs, but they fail in different layers: missing model weights vs AutoProcessor forwarding kwargs to cached_file." }, { - "left": "issue:43992", - "right": "issue:44704", + "left": "issue:43066", + "right": "issue:46710", "accept": false, - "reason": "Missing model weight key on load vs AutoProcessor kwarg passthrough are unrelated code paths." + "reason": "These are unrelated: tokenizer decoder type in v5 versus a DeepSeek accuracy regression after upgrading Transformers." }, { "left": "issue:46614", "right": "issue:46752", "accept": false, - "reason": "Phonemizer backend state bug and empty chat-template handling are not the same issue." + "reason": "Tokenizer backend state leakage and apply_chat_template crashing on an empty list are different components and failure modes." }, { "left": "issue:44995", "right": "issue:46697", "accept": false, - "reason": "GlmMoeDsa stale cache crash vs Swin2SR image over-padding are distinct bugs." + "reason": "A stale cache crash in GlmMoeDsa is unrelated to Swin2SR image over-padding in the processor." }, { "left": "issue:28348", "right": "issue:46640", "accept": false, - "reason": "Feature request to add GPT2 flash-attention support is not the same as a vague Flash Attention 2 bug report." + "reason": "One is a feature request for GPT2 FlashAttention 2.0 support; the other is a separate generic Flash Attention 2 issue." }, { "left": "issue:42673", "right": "issue:42796", "accept": false, - "reason": "VRAM leak under multithreaded loading and tokenizer BPE-marker handling are unrelated." + "reason": "VRAM leakage in Qwen3ForCausalLM and tokenizer BPE marker handling in Ministral are different bugs." }, { "left": "issue:44263", "right": "issue:46762", "accept": false, - "reason": "torch.split return-value handling in GlmMoeDsaIndexer is different from Minimax attention numerical differences." + "reason": "GlmMoeDsaIndexer split return values and Minimax MSA attention numerical differences are not the same bug." }, { "left": "issue:45092", "right": "issue:46762", "accept": false, - "reason": "Meta-init incompatibility for remote-code checkpoints is not the same as Minimax MSA numerical mismatch." + "reason": "Remote-code checkpoint/meta-init incompatibility is unrelated to attention numerical drift." }, { "left": "issue:43827", "right": "issue:45446", "accept": false, - "reason": "Documentation still referencing pipeline() and a flex_attention version check bug are different issues." + "reason": "Docs still referencing pipeline() and a PyTorch version check bug in flex_attention are different issues." }, { "left": "issue:33453", "right": "issue:39401", "accept": false, - "reason": "Generic tokenizer-loading regression and Qwen3 offset_mapping bug are not the same underlying defect." + "reason": "Both are tokenizer regressions, but one is broad loading regression while the other is a Qwen3 offset_mapping bug." }, { "left": "issue:43064", "right": "issue:43673", "accept": false, - "reason": "FSDP2/PEFT optimizer-state corruption and chunked_prefill cache missing are distinct training vs generation bugs." - }, - { - "left": "issue:42222", - "right": "issue:43441", - "accept": false, - "reason": "Vitpose model breakage and Ministral flash-attention failure are unrelated model-specific issues." + "reason": "Optimizer-state corruption under FSDP2/PEFT and missing GenerationMixin cache during chunked_prefill are unrelated training vs generation issues." }, { "left": "issue:43937", "right": "issue:45563", "accept": false, - "reason": "Invalid GenerationConfig and a stale warning about num_return_sequences are separate problems." + "reason": "An invalid GenerationConfig error and a stale warning for num_return_sequences are different generation-path problems." }, { "left": "issue:43824", "right": "issue:46981", "accept": false, - "reason": "Missing Qwen2_5_VL import and repeated HTTP 429 retries in AutoProcessor are unrelated." + "reason": "An import failure for a specific model class and repeated 429 retries in AutoProcessor are not the same underlying bug." }, { "left": "issue:42222", - "right": "issue:47000", + "right": "issue:43441", "accept": false, - "reason": "Vitpose loading breakage and RTDetrLoss infeasible cost matrix are different code paths." + "reason": "Vitpose model breakage and Ministral FlashAttention failure are separate model-specific issues." }, { - "left": "issue:43673", - "right": "issue:46328", + "left": "issue:42673", + "right": "issue:46655", "accept": false, - "reason": "Missing generation cache and SequenceFeatureExtractor.pad performance waste are not the same bug." + "reason": "A VRAM leak in model usage threads is unrelated to a video processor padding calculation bug." }, { - "left": "issue:42673", - "right": "issue:46655", + "left": "issue:42222", + "right": "issue:47000", "accept": false, - "reason": "VRAM leak in threaded Qwen3 code and incorrect video padding in SmolVLM are unrelated." + "reason": "Vitpose breakage and RTDetrLoss infeasible cost matrix errors do not share a concrete code path." }, { - "left": "issue:44291", - "right": "issue:45071", + "left": "issue:43673", + "right": "issue:46328", "accept": false, - "reason": "Unexpected _is_hf_initialized arg during init_empty_weights and PretrainedConfig type-checking breakage are different failures." + "reason": "Missing cache in generation and inefficient numpy-to-list conversion in padding are different performance/correctness issues." }, { "left": "issue:45405", "right": "issue:45446", "accept": false, - "reason": "PEFT version pinning and a PyTorch version check bug in flex_attention are unrelated." + "reason": "A dependency version bump problem and a flex_attention import-version check bug are unrelated." + }, + { + "left": "issue:44291", + "right": "issue:45071", + "accept": false, + "reason": "Unexpected _is_hf_initialized argument handling and PretrainedConfig type-checking regressions are distinct v5 compatibility issues." }, { "left": "issue:42222", "right": "issue:46752", "accept": false, - "reason": "Vitpose model breakage and empty-list chat template handling are not the same issue." + "reason": "Vitpose model breakage and empty-list chat template crashing are unrelated." }, { "left": "issue:44117", "right": "issue:45588", "accept": false, - "reason": "Tokenizer mapping None on from_pretrained and flash_attention s_aux=None AttributeError are separate bugs." + "reason": "Tokenizer mapping assumptions and flash_attention integration crashing on s_aux=None are different subsystems." }, { "left": "issue:44380", "right": "issue:46077", "accept": false, - "reason": "GPT2 attention-scaling being ignored and Gemma4 bidirectional mask construction are different model/backend problems." + "reason": "GPT2 attention scaling being ignored and Gemma4 building causal masks are both attention-related but not the same bug." }, { "left": "issue:44466", "right": "issue:46002", "accept": false, - "reason": "lm_head serialization inconsistency and NaNs in classification heads are unrelated symptoms." + "reason": "Serialization of lm_head.weight and NaNs in classification heads are unrelated training/serialization issues." }, { "left": "issue:39401", "right": "issue:44779", "accept": false, - "reason": "Qwen3 offset_mapping is distinct from DeepSeek tokenizer regression." + "reason": "Both are tokenizer correctness regressions, but they affect different models and behaviors: offset_mapping vs incorrect tokenization results." }, { "left": "issue:43937", "right": "issue:44977", "accept": false, - "reason": "Invalid GenerationConfig and Qwen3.5 flash-attention generation failure are different issues." + "reason": "Invalid GenerationConfig handling and Qwen3.5 flash-attention generation failure are distinct generation problems." }, { "left": "issue:44410", "right": "issue:45310", "accept": false, - "reason": "Missing qwen3next projections and Qwen3.5 MoE from_pretrained error do not look like one underlying bug." + "reason": "Missing projections in qwen3next layer 0 and a Qwen3.5 MoE from_pretrained error are different model-loading/model-architecture bugs." + }, + { + "left": "issue:43450", + "right": "issue:43986", + "accept": false, + "reason": "Incorrect batched video-processor output shape and an AutoProcessor crash without torchvision are separate video-processing issues." }, { "left": "issue:45071", "right": "issue:46512", "accept": false, - "reason": "PretrainedConfig type-checking breakage and stale config.dtype after from_config are separate configuration bugs." + "reason": "PretrainedConfig type checking breakage and stale model.config.dtype after from_config(dtype=...) are different v5 API regressions." } ] }, @@ -16291,7 +16341,7 @@ { "label": "issue", "component_index": 1, - "component_count": 25, + "component_count": 26, "review_unit_index": 53, "review_unit_count": 61, "status": "reviewed", @@ -16312,6 +16362,7 @@ "issue:43122", "issue:43208", "issue:43408", + "issue:43673", "issue:43742", "issue:43937", "issue:44075", @@ -16321,7 +16372,6 @@ "issue:44261", "issue:44263", "issue:44410", - "issue:44483", "issue:44484", "issue:44485", "issue:44493", @@ -16335,7 +16385,6 @@ "issue:45310", "issue:45341", "issue:45362", - "issue:45464", "issue:45478", "issue:45561", "issue:45632", @@ -16352,9 +16401,9 @@ ], "soft_pairs": [ "issue:45341|issue:45561", - "issue:45632|issue:46981", - "issue:45341|issue:46328", "issue:44829|issue:45216", + "issue:45341|issue:46328", + "issue:45632|issue:46981", "issue:44117|issue:45081", "issue:42617|issue:45341", "issue:43408|issue:43742", @@ -16362,220 +16411,220 @@ "issue:43122|issue:46298", "issue:45362|issue:46328", "issue:39401|issue:46209", - "issue:44877|issue:46231", "issue:42222|issue:46178", + "issue:44877|issue:46231", "issue:44484|issue:45310", "issue:42371|issue:44261", "issue:43937|issue:44496", + "issue:44933|issue:45278", "issue:42371|issue:46617", "issue:44261|issue:46762", - "issue:44933|issue:45278", "issue:44410|issue:45478", "issue:44075|issue:45656", "issue:38617|issue:44220", - "issue:43208|issue:46894", "issue:43066|issue:43122", - "issue:44263|issue:44485", + "issue:43208|issue:46894", "issue:43097|issue:44493", + "issue:44263|issue:44485", "issue:44077|issue:46981", - "issue:44483|issue:45464" + "issue:43673|issue:45341" ], - "prepared_review_unit_hash": "a20f8e7f875d51a481a983dcfc8436beb61aee062aa3cbc9bf6faa3192855c7f", + "prepared_review_unit_hash": "cd9f3a58f16a4d9bb9d5bf0f812fbae2f639aecbcade3b57915289cfaa431456", "component_budget": { "node_count": 505, "item_count": 505, "soft_pair_count": 1854, - "serialized_chars": 962354, - "estimated_input_tokens": 240589, - "estimated_eval_tokens": 481434 + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 }, "budget": { - "node_count": 48, - "item_count": 48, + "node_count": 47, + "item_count": 47, "soft_pair_count": 28, - "serialized_chars": 35713, - "estimated_input_tokens": 8929, - "estimated_eval_tokens": 18114 + "serialized_chars": 35144, + "estimated_input_tokens": 8786, + "estimated_eval_tokens": 17828 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "No clear duplicate cluster emerges here: the issues span many unrelated regressions across tokenization, model loading, training, and utilities. I did not find any soft pair that looks like the same underlying bug/change closely enough to merge.", - "confidence": 0.89, - "canonical_issue_reason": "issue:45561 is the best anchor because it is open, recent, and has the strongest discussion/inbound signal around a concrete, reproducible testing-utils race.", + "summary": "This cluster is mostly heterogeneous issue reports. The soft-similarity pairs share broad subsystems (tokenizers, imports, models, optimizers, config loading) but not the same concrete bug or fixable code path, so none should be merged as duplicates.", + "confidence": 0.33, + "canonical_issue_reason": "If a single issue must represent the cluster, issue 45278 is the broadest umbrella-style regression report (many import errors after upgrading) and is the closest to a cluster anchor.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45561 is the most suitable global representative: it has the clearest actionable bug report and the highest triage signal among the listed issues.", + "best_issue_reason": "Issue 45278 is the most general and triage-friendly report in this set: it describes a broad post-upgrade import regression rather than a narrow model-specific symptom.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:45341", "right": "issue:45561", "accept": false, - "reason": "Same file area, but the reports describe different bugs: a vague testing_utils issue vs a specific xdist race on captured_info.txt." + "reason": "Same file area, but different bug class: a small testing_utils bug vs a xdist race on captured_info.txt." }, { - "left": "issue:45632", - "right": "issue:46981", + "left": "issue:44829", + "right": "issue:45216", "accept": false, - "reason": "Both involve remote-code/model loading, but one is a local cache-path collision and the other is repeated 429 retry behavior." + "reason": "Different problems: flash-attention training degeneration vs incorrect checkpoint saving for Qwen3.5." }, { "left": "issue:45341", "right": "issue:46328", "accept": false, - "reason": "Unrelated failure modes: a generic testing_utils bug versus a SequenceFeatureExtractor.pad performance issue." + "reason": "Unrelated: testing_utils vs a SequenceFeatureExtractor.pad performance issue." }, { - "left": "issue:44829", - "right": "issue:45216", + "left": "issue:45632", + "right": "issue:46981", "accept": false, - "reason": "Different problems: flash_attention_3 training degeneration vs Qwen3.5 save_pretrained checkpoint correctness." + "reason": "Both involve loading/caching, but one is cache-path collision and the other is repeated HTTP 429 retry behavior." }, { "left": "issue:44117", "right": "issue:45081", "accept": false, - "reason": "Both are tokenizer-related, but one is a mapping lookup returning None and the other is a Mistral regex patch crash." + "reason": "Both are tokenizer-related, but one is a mapping lookup bug and the other is a Mistral regex patch crash." }, { "left": "issue:42617", "right": "issue:45341", "accept": false, - "reason": "No shared underlying bug; one is a 3d_parallel.py runtime issue and the other is a testing_utils bug report." + "reason": "Unrelated: a 3d_parallel.py runtime failure vs a testing_utils bug." }, { "left": "issue:43408", "right": "issue:43742", "accept": false, - "reason": "Different model-loading issues: a sam3_video/sam3_tracker type mismatch versus a KeyError on MobileLLM-125M." + "reason": "Different model-loading issues affecting different model families and failure modes." }, { "left": "issue:38617", "right": "issue:41628", "accept": false, - "reason": "Both are import errors, but they concern different missing symbols and likely different code paths." + "reason": "Both are import errors, but for different missing symbols and likely different removal paths." }, { "left": "issue:43122", "right": "issue:46298", "accept": false, - "reason": "Tokenizer-behavior regression versus an unrelated noisy logging complaint." + "reason": "Different tokenization regression vs a noisy logging message; no shared bug." }, { "left": "issue:45362", "right": "issue:46328", "accept": false, - "reason": "Qwen3.5 chat crash is unrelated to a SequenceFeatureExtractor.pad conversion inefficiency." + "reason": "Completely different areas: chat generation crash vs feature-extractor padding performance." }, { "left": "issue:39401", "right": "issue:46209", "accept": false, - "reason": "Tokenizer offset-mapping bug and FP8/B200 inference failure are unrelated subsystems and symptoms." + "reason": "Tokenizer offset mapping bug vs FP8 inference failure on B200; unrelated." }, { - "left": "issue:44877", - "right": "issue:46231", + "left": "issue:42222", + "right": "issue:46178", "accept": false, - "reason": "Strict config loading for granite_speech is unrelated to a doc parameter-name mismatch." + "reason": "Different model family breakage vs unsupported backend mapping validation." }, { - "left": "issue:42222", - "right": "issue:46178", + "left": "issue:44877", + "right": "issue:46231", "accept": false, - "reason": "Vitpose model breakage and BACKENDS_MAPPING validation are different underlying loading/backend issues." + "reason": "Config-loading strictness vs a documentation parameter-name mismatch." }, { "left": "issue:44484", "right": "issue:45310", "accept": false, - "reason": "Different issues: save_pretrained shard-size question versus Qwen3.5 MoE from_pretrained failure." + "reason": "Save-pretrained shard-size question vs Qwen3.5 from_pretrained error; unrelated." }, { "left": "issue:42371", "right": "issue:44261", "accept": false, - "reason": "TF32 API-setting guidance is unrelated to MLA q_a_layernorm missing rms_norm_eps precision behavior." + "reason": "TF32 API behavior change vs a model precision/eps bug in MLA layernorm." }, { "left": "issue:43937", "right": "issue:44496", "accept": false, - "reason": "GenerationConfig validation failure and unrecognized-model config loading are different code paths." + "reason": "Both are validation failures, but one is GenerationConfig invalidity and the other is missing model_type in config." + }, + { + "left": "issue:44933", + "right": "issue:45278", + "accept": false, + "reason": "A specific missing import vs a broad import-regression report; not the same concrete bug." }, { "left": "issue:42371", "right": "issue:46617", "accept": false, - "reason": "TF32 behavior settings and trust_remote_code cache FileNotFoundError are unrelated." + "reason": "TF32 API control message vs trust_remote_code cache FileNotFoundError; unrelated." }, { "left": "issue:44261", "right": "issue:46762", "accept": false, - "reason": "Both mention numerical quality, but one is a missing config epsilon and the other is a broader attention-vs-reference discrepancy." - }, - { - "left": "issue:44933", - "right": "issue:45278", - "accept": false, - "reason": "An import from image_utils versus many import errors after upgrade; overlapping theme but not the same concrete bug." + "reason": "Both are numerical/model-behavior issues, but they affect different models and code paths." }, { "left": "issue:44410", "right": "issue:45478", "accept": false, - "reason": "qwen3next missing projections and Qwen3.5 MoE from_pretrained error are distinct model-specific failures." + "reason": "Different model architectures and failure modes; no shared concrete fix." }, { "left": "issue:44075", "right": "issue:45656", "accept": false, - "reason": "Different optimizer bugs: SGD args ignored versus optimizer.step being called twice with deepspeed." + "reason": "Both mention optimizer behavior, but one is unused SGD args and the other is double optimizer stepping under deepspeed." }, { "left": "issue:38617", "right": "issue:44220", "accept": false, - "reason": "Different failures entirely: missing configuration_utils export versus _torch_extract_fbank_features() behavior." - }, - { - "left": "issue:43208", - "right": "issue:46894", - "accept": false, - "reason": "xLSTM training blockers and CohereASR double-shift loss bug are separate model-training issues." + "reason": "ImportError for a missing symbol is unrelated to an audio feature extraction issue." }, { "left": "issue:43066", "right": "issue:43122", "accept": false, - "reason": "Both are tokenizer regressions in v5, but they describe different concrete symptoms and likely different fixes." + "reason": "Both concern tokenizer regressions around v5, but the reported symptoms are different enough that they do not clearly share one concrete code-path bug." }, { - "left": "issue:44263", - "right": "issue:44485", + "left": "issue:43208", + "right": "issue:46894", "accept": false, - "reason": "A torch.split indexing issue and a RoPE implementation discussion are not the same bug." + "reason": "Different model/training bugs, even if both mention loss/training behavior." }, { "left": "issue:43097", "right": "issue:44493", "accept": false, - "reason": "Removed tie_embeddings API and unexpected position-id keys are unrelated regressions." + "reason": "Removed tie_embeddings API vs unexpected position-id keys; unrelated." + }, + { + "left": "issue:44263", + "right": "issue:44485", + "accept": false, + "reason": "Different GLM-5 issues: torch.split return handling vs RoPE implementation." }, { "left": "issue:44077", "right": "issue:46981", "accept": false, - "reason": "PatchTSMixer post_init policy and AutoProcessor 429 retry behavior are unrelated." + "reason": "Post-init policy issue vs HTTP retry behavior; unrelated." }, { - "left": "issue:44483", - "right": "issue:45464", + "left": "issue:43673", + "right": "issue:45341", "accept": false, - "reason": "Both hit chat/completions, but one is a general request-acceptance problem and the other is a model-specific streaming inference failure." + "reason": "Generation cache regression vs a testing_utils bug; different code paths." } ] }, @@ -16589,7 +16638,7 @@ { "label": "issue", "component_index": 1, - "component_count": 25, + "component_count": 26, "review_unit_index": 54, "review_unit_count": 61, "status": "reviewed", @@ -16624,6 +16673,7 @@ "issue:44315", "issue:44360", "issue:44410", + "issue:44483", "issue:44484", "issue:44485", "issue:44829", @@ -16634,6 +16684,7 @@ "issue:45372", "issue:45375", "issue:45440", + "issue:45464", "issue:45478", "issue:45542", "issue:45588", @@ -16641,24 +16692,22 @@ "issue:46050", "issue:46123", "issue:46172", - "issue:46298", "issue:46369", - "issue:46752", "issue:46897", "issue:46981" ], "soft_pairs": [ - "issue:43673|issue:45341", + "issue:44483|issue:45464", "issue:43931|issue:44410", "issue:41628|issue:45071", "issue:45071|issue:45542", "issue:44484|issue:45478", - "issue:43874|issue:44261", "issue:43931|issue:44038", "issue:43012|issue:43577", - "issue:44263|issue:46172", + "issue:43874|issue:44261", "issue:43441|issue:44829", "issue:27957|issue:43352", + "issue:44263|issue:46172", "issue:44360|issue:46172", "issue:43097|issue:43452", "issue:43986|issue:45341", @@ -16668,246 +16717,239 @@ "issue:45753|issue:46369", "issue:44117|issue:44987", "issue:43531|issue:45341", - "issue:43874|issue:44263", "issue:43097|issue:45588", + "issue:43874|issue:44263", "issue:43673|issue:46123", "issue:45440|issue:46172", "issue:44410|issue:45375", "issue:43986|issue:45372", - "issue:44038|issue:44410", "issue:43097|issue:46050", + "issue:44038|issue:44410", "issue:38617|issue:42757", "issue:43064|issue:44315", - "issue:43208|issue:46897", "issue:43986|issue:44484", - "issue:46298|issue:46752" + "issue:43208|issue:46897" ], - "prepared_review_unit_hash": "534631c8a7e85b16f4a9e1a2c112a869f8f5bf4b9994dbb604a0cd5c0251f885", + "prepared_review_unit_hash": "169be7f8ba93925432a7b9e9755811edb2ccecd544d7cfaa00b65018a89ce3ed", "component_budget": { "node_count": 505, "item_count": 505, "soft_pair_count": 1854, - "serialized_chars": 962354, - "estimated_input_tokens": 240589, - "estimated_eval_tokens": 481434 + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 }, "budget": { "node_count": 47, "item_count": 47, - "soft_pair_count": 33, - "serialized_chars": 36826, - "estimated_input_tokens": 9207, - "estimated_eval_tokens": 18670 + "soft_pair_count": 32, + "serialized_chars": 36567, + "estimated_input_tokens": 9142, + "estimated_eval_tokens": 18540 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly a grab-bag of unrelated Transformers regressions (imports, model loading, quantization, attention, config, and training bugs). None of the soft pairs look like the same underlying defect, so I would not merge any of them as duplicates.", - "confidence": 0.78, - "canonical_issue_reason": "There is no strong duplicate hub here; the issues are too heterogeneous. If forced to pick one anchor, 43097 is the broadest release-regression style report, but it does not represent the whole cluster well.", + "summary": "This cluster does not look like a true duplicate set; it spans many unrelated Transformers/HF bugs across imports, model loading, FlashAttention, training, tokenization, and infra behavior. I rejected all soft edges.", + "confidence": 0.93, + "canonical_issue_reason": "No clear canonical issue: the items are a heterogeneous mix of unrelated bugs, so there is no single duplicate target that the rest collapse into.", "canonical_pr_reason": null, - "best_issue_reason": "No single issue is a clean canonical representative for the entire cluster. 43097 is the closest to a broadly relevant v5 compatibility regression, but the rest are mostly separate bugs.", + "best_issue_reason": "No single issue is a strong global representative; any choice would be arbitrary because the cluster lacks one underlying problem.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:43673", - "right": "issue:45341", + "left": "issue:44483", + "right": "issue:45464", "accept": false, - "reason": "GenerationMixin cache/chunked_prefill regression vs a testing_utils bug; different code paths and symptoms." + "reason": "Both mention chat/completions failures, but one is a generic API complaint and the other is a Qwen3.5 streaming issue; not the same bug." }, { "left": "issue:43931", "right": "issue:44410", "accept": false, - "reason": "Qwen3-VL model loading shape mismatch vs Qwen3Next missing attention projections; related family, but not the same bug." + "reason": "Different model families and different failure modes: weight-shape mismatch vs missing projections." }, { "left": "issue:41628", "right": "issue:45071", "accept": false, - "reason": "AutoImageProcessor import failure vs PretrainedConfig type-checking regression; unrelated failures." + "reason": "Unrelated import error vs config type-checking regression." }, { "left": "issue:45071", "right": "issue:45542", "accept": false, - "reason": "Type checking breakage in PretrainedConfig vs undefined tf backend error from TensorBoard-only installs." + "reason": "Different subsystems: config typing vs TensorBoard/TF backend handling." }, { "left": "issue:44484", "right": "issue:45478", "accept": false, - "reason": "save_pretrained shard-size question vs Qwen3.5 MoE from_pretrained error; not the same issue." - }, - { - "left": "issue:43874", - "right": "issue:44261", - "accept": false, - "reason": "Missing image-processor method vs MLA q_a_layernorm epsilon mismatch; different subsystems." + "reason": "A save_pretrained shard-size question is unrelated to a Qwen3.5-MoE from_pretrained failure." }, { "left": "issue:43931", "right": "issue:44038", "accept": false, - "reason": "Both are Qwen3-VL-related, but one is a weight-shape load error and the other is a different v5/MoE failure." + "reason": "Both are Qwen3-VL-related, but one is a shape mismatch and the other is a broader v5/MoE load bug." }, { "left": "issue:43012", "right": "issue:43577", "accept": false, - "reason": "PyTorch bfloat16 compile warning vs Blip2 dtype not propagating; separate bugs." + "reason": "Different dtype issues: compile-time precision warning vs model loading leaving dtype at float32." }, { - "left": "issue:44263", - "right": "issue:46172", + "left": "issue:43874", + "right": "issue:44261", "accept": false, - "reason": "GlmMoeDsaIndexer split return handling vs deepseek-v4 comb transpose logic; unrelated code paths." + "reason": "Different models and code paths: multimodal token counting vs MLA layernorm config precision." }, { "left": "issue:43441", "right": "issue:44829", "accept": false, - "reason": "FlashAttention support error for a model vs degenerate training under flash_attention_3; different concrete problems." + "reason": "Both involve FlashAttention, but one is a model-support/load failure and the other is training degeneration under flash_attention_3." }, { "left": "issue:27957", "right": "issue:43352", "accept": false, - "reason": "Two different models lacking FlashAttention 2 support; same broad topic, but not the same bug." + "reason": "Unsupported FlashAttention 2 on different models is not enough to treat these as the same underlying bug." + }, + { + "left": "issue:44263", + "right": "issue:46172", + "accept": false, + "reason": "Different DeepSeek/GLM implementation bugs with no shared concrete code path." }, { "left": "issue:44360", "right": "issue:46172", "accept": false, - "reason": "DSA indexer activation issue vs deepseek-v4 tensor transpose bug; not duplicates." + "reason": "One is a missing ReLU in a DSA indexer; the other is a tensor dtype/transpose issue." }, { "left": "issue:43097", "right": "issue:43452", "accept": false, - "reason": "Removed tie_embeddings API vs gguf_file pretraining loader breakage; distinct regressions." + "reason": "Removed API semantics vs gguf_file/from_pretrained breakage are unrelated." }, { "left": "issue:43986", "right": "issue:45341", "accept": false, - "reason": "AutoProcessor video-model crash without torchvision vs a testing_utils bug; unrelated." + "reason": "Video AutoProcessor import/runtime crash is unrelated to a testing_utils bug." }, { "left": "issue:45003", "right": "issue:45341", "accept": false, - "reason": "Unsafe sys.modules access in modeling_utils vs testing_utils bug; different modules and failures." + "reason": "Unsafe sys.modules access is a different problem from the testing_utils issue." }, { "left": "issue:42886", "right": "issue:46981", "accept": false, - "reason": "Offline cache loading failure vs repeated 429 retries via HF_ENDPOINT; both network-ish, but different behaviors." + "reason": "Both are hub/network related, but offline cache loading and HTTP 429 retry behavior are distinct bugs." }, { "left": "issue:44261", "right": "issue:44485", "accept": false, - "reason": "LayerNorm epsilon precision mismatch vs GLM-5 RoPE implementation; separate model-specific issues." + "reason": "RMS norm epsilon precision discussion is unrelated to GLM-5 RoPE implementation." }, { "left": "issue:45753", "right": "issue:46369", "accept": false, - "reason": "Qwen3_5 infinite loop on one image vs sam3-litetext memory overflow; different models and failure modes." + "reason": "An infinite-loop image bug and a memory overflow are different failure classes." }, { "left": "issue:44117", "right": "issue:44987", "accept": false, - "reason": "Tokenizer mapping None handling vs loading physical-intelligence/fast failure; not the same loader bug." + "reason": "Tokenizer mapping None handling and a specific model load failure are not clearly the same code-path bug." }, { "left": "issue:43531", "right": "issue:45341", "accept": false, - "reason": "Qwen3-MoE sliding_window issue vs testing_utils bug; unrelated." + "reason": "Qwen3-MoE sliding-window behavior is unrelated to testing_utils." }, { - "left": "issue:43874", - "right": "issue:44263", + "left": "issue:43097", + "right": "issue:45588", "accept": false, - "reason": "Missing image patch-count method vs torch.split return handling; different GLM bugs." + "reason": "A removed config/helper API and a flash_attention s_aux=None crash are different changes." }, { - "left": "issue:43097", - "right": "issue:45588", + "left": "issue:43874", + "right": "issue:44263", "accept": false, - "reason": "API removal in v5 vs flash_attention.py crash on s_aux=None; different code paths." + "reason": "Different GLM46V multimodal-token bug vs GLM MoE indexer split behavior." }, { "left": "issue:43673", "right": "issue:46123", "accept": false, - "reason": "Missing cache in GenerationMixin vs MaskGenerationPipeline final-batch handling; unrelated pipeline vs generation bugs." + "reason": "Generation cache handling and mask generation batching are different pipeline bugs." }, { "left": "issue:45440", "right": "issue:46172", "accept": false, - "reason": "DeepseekV3MoE divergence from remote implementation vs deepseek-v4 tensor transpose bug; same general family, but not the same defect." + "reason": "Both are DeepSeek-related, but they describe different divergences and different concrete code paths." }, { "left": "issue:44410", "right": "issue:45375", "accept": false, - "reason": "Missing attn projections in qwen3next layer 0 vs dropped deepstack_visual_indexes field; different model/config bugs." + "reason": "Qwen3next missing projections and missing vision-config fields are distinct model-definition issues." }, { "left": "issue:43986", "right": "issue:45372", "accept": false, - "reason": "AutoProcessor crash without torchvision vs missing ReasoningEffort import from mistral_common; unrelated dependency/import failures." + "reason": "AutoProcessor without torchvision and a mistral_common import error are unrelated." }, { - "left": "issue:44038", - "right": "issue:44410", + "left": "issue:43097", + "right": "issue:46050", "accept": false, - "reason": "Qwen3-VL-Moe v5 bug vs qwen3next projection-missing bug; different models and symptoms." + "reason": "API removal and quantized-model save/load corruption are not the same underlying bug." }, { - "left": "issue:43097", - "right": "issue:46050", + "left": "issue:44038", + "right": "issue:44410", "accept": false, - "reason": "Removed tie_embeddings API vs sinq quantization save/load corruption; distinct regressions." + "reason": "Different Qwen variants and different missing-attribute/model-structure problems." }, { "left": "issue:38617", "right": "issue:42757", "accept": false, - "reason": "Two separate import errors for different symbols/packages; not the same bug." + "reason": "Separate import errors from different modules/packages." }, { "left": "issue:43064", "right": "issue:44315", "accept": false, - "reason": "FSDP2/PEFT optimizer-state corruption vs Liger Kernel not applied with model_init; different training-stack issues." - }, - { - "left": "issue:43208", - "right": "issue:46897", - "accept": false, - "reason": "xLSTM training blockers vs Florence2 double-shift loss bug; unrelated models and fixes." + "reason": "Optimizer-state corruption under FSDP2/PEFT is unrelated to Liger Kernel not being applied." }, { "left": "issue:43986", "right": "issue:44484", "accept": false, - "reason": "AutoProcessor video-model dependency crash vs max_shard_size default discussion; unrelated." + "reason": "No shared bug: video processor crash vs a shard-size configuration question." }, { - "left": "issue:46298", - "right": "issue:46752", + "left": "issue:43208", + "right": "issue:46897", "accept": false, - "reason": "Noisy logging message vs apply_chat_template empty-list crash; not the same underlying defect." + "reason": "Both are training bugs, but they affect different models with different concrete symptoms and fixes." } ] }, @@ -16921,7 +16963,7 @@ { "label": "issue", "component_index": 1, - "component_count": 25, + "component_count": 26, "review_unit_index": 55, "review_unit_count": 61, "status": "reviewed", @@ -16931,7 +16973,6 @@ "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-505", "nodes": [ - "issue:36683", "issue:38617", "issue:42890", "issue:42907", @@ -16942,7 +16983,6 @@ "issue:43352", "issue:43502", "issue:43550", - "issue:43582", "issue:43825", "issue:43873", "issue:43906", @@ -16978,16 +17018,18 @@ "issue:46664", "issue:46701", "issue:46710", + "issue:46752", "issue:46821" ], "soft_pairs": [ + "issue:46298|issue:46752", "issue:46655|issue:46664", "issue:42890|issue:46298", "issue:44062|issue:44117", - "issue:38617|issue:44351", "issue:43097|issue:43873", - "issue:46172|issue:46710", + "issue:38617|issue:44351", "issue:43012|issue:46821", + "issue:46172|issue:46710", "issue:44117|issue:46701", "issue:43502|issue:43992", "issue:44936|issue:46365", @@ -16998,236 +17040,235 @@ "issue:42907|issue:44655", "issue:42907|issue:45753", "issue:43012|issue:43550", - "issue:43825|issue:44655", "issue:45712|issue:46298", "issue:43012|issue:44368", "issue:43208|issue:43550", + "issue:43825|issue:44655", "issue:43992|issue:44484", "issue:43352|issue:43976", "issue:43352|issue:44380", "issue:43066|issue:43906", "issue:43097|issue:46298", - "issue:44162|issue:46172", "issue:45584|issue:46365", + "issue:44162|issue:46172", "issue:44380|issue:46132", "issue:44077|issue:46365", - "issue:44222|issue:45478", - "issue:36683|issue:43582" + "issue:44222|issue:45478" ], - "prepared_review_unit_hash": "d23f1877d4ad2e0e6971daffe25f011177145bbb070fe7c0feb44e30d82e9831", + "prepared_review_unit_hash": "5e1831300dfa895fc78e474dd7b225282c60882248f9a83e8eb570e853d3fc2e", "component_budget": { "node_count": 505, "item_count": 505, "soft_pair_count": 1854, - "serialized_chars": 962354, - "estimated_input_tokens": 240589, - "estimated_eval_tokens": 481434 + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 }, "budget": { - "node_count": 48, - "item_count": 48, + "node_count": 47, + "item_count": 47, "soft_pair_count": 32, - "serialized_chars": 37018, - "estimated_input_tokens": 9255, - "estimated_eval_tokens": 18766 + "serialized_chars": 36469, + "estimated_input_tokens": 9118, + "estimated_eval_tokens": 18492 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "These items are a heterogeneous mix of unrelated Transformers issues rather than one duplicate cluster. None of the soft pairs describe the same underlying bug or change closely enough to merge.", + "summary": "The cluster is heterogeneous: the soft-similarity pairs mostly share broad Transformers subsystem keywords, but they point to different bugs, regressions, or features. I did not find any pair that clearly describes the same underlying issue/change, so no soft edges are accepted.", "confidence": 0.94, - "canonical_issue_reason": null, + "canonical_issue_reason": "No single canonical issue stands out because the items are about different code paths and failure modes rather than one underlying defect.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43873 is the strongest standalone anchor by discussion/inbound references, but it is not a duplicate of the other items and the cluster is not coherent enough to have a true canonical issue.", + "best_issue_reason": "No issue is a strong cluster representative; the set is too mixed across unrelated bugs, so a single best issue would be misleading.", "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:46298", + "right": "issue:46752", + "accept": false, + "reason": "Different problems: noisy logging vs an empty-list crash in apply_chat_template." + }, { "left": "issue:46655", "right": "issue:46664", "accept": false, - "reason": "Both are SmolVLM video-processor bugs, but one is padding calculation and the other is `num_frames` semantics; related subsystem, different defects." + "reason": "Both are SmolVLM video-processor bugs, but one is padding math and the other is num_frames semantics." }, { "left": "issue:42890", "right": "issue:46298", "accept": false, - "reason": "One is missing `set_seed()` in a test suite; the other is a noisy logging message. No shared underlying bug." + "reason": "Test seed flakiness is unrelated to a logging message." }, { "left": "issue:44062", "right": "issue:44117", "accept": false, - "reason": "Both concern tokenizer paths, but one is an `AddedToken` argument conflict and the other is `TOKENIZER_MAPPING_NAMES` returning `None`; different failure modes." - }, - { - "left": "issue:38617", - "right": "issue:44351", - "accept": false, - "reason": "Both are import errors, but they involve different missing symbols from different code paths." + "reason": "One is an AddedToken keyword-argument error; the other is a tokenizer mapping None-handling bug." }, { "left": "issue:43097", "right": "issue:43873", "accept": false, - "reason": "One is about a removed config attribute in v5; the other is quantization/offloading behavior. Not the same bug." + "reason": "Token/embedding config removal is unrelated to quantization offloading behavior." }, { - "left": "issue:46172", - "right": "issue:46710", + "left": "issue:38617", + "right": "issue:44351", "accept": false, - "reason": "One is a specific dtype/transpose issue in DeepSeek-V4; the other is an accuracy regression after upgrading Transformers. Different problems." + "reason": "Both are import errors, but for different missing symbols and likely different root causes." }, { "left": "issue:43012", "right": "issue:46821", "accept": false, - "reason": "A bfloat16 precision warning from compilation is unrelated to a TimesFM `window_size` attribute error." + "reason": "A PyTorch precision warning is unrelated to a TimesFM AttributeError." + }, + { + "left": "issue:46172", + "right": "issue:46710", + "accept": false, + "reason": "DeepSeek model code failure and a separate accuracy regression are not the same bug." }, { "left": "issue:44117", "right": "issue:46701", "accept": false, - "reason": "Tokenizer mapping `None` handling is unrelated to the DiffusionGemma pipeline failure." + "reason": "Tokenizer mapping fallback logic is unrelated to the DiffusionGemma image-text pipeline failure." }, { "left": "issue:43502", "right": "issue:43992", "accept": false, - "reason": "`local_files_only=True` still triggering network access is unrelated to missing `embed_tokens.weight` during `from_pretrained`." + "reason": "Local-files-only network leakage is unrelated to a missing embed_tokens weight on load." }, { "left": "issue:44936", "right": "issue:46365", "accept": false, - "reason": "`trainer.evaluate()` after training and failing to import `Trainer` with `datasets` installed are different issues." + "reason": "Trainer evaluate/train state bug is unrelated to an import-time datasets conflict." }, { "left": "issue:44704", "right": "issue:45103", "accept": false, - "reason": "Passing kwargs to `cached_file` and docstring parsing with `from __future__ import annotations` do not share a code-path." + "reason": "Passing kwargs to cached_file is unrelated to the auto_docstring annotation crash." }, { "left": "issue:44380", "right": "issue:46421", "accept": false, - "reason": "GPT2 attention scaling under SDPA/FlashAttention is unrelated to cache lazy initialization and recompilation." + "reason": "GPT-2 attention scaling under SDPA/FlashAttention is unrelated to chunked-prefill recompilation." }, { "left": "issue:44938", "right": "issue:46178", "accept": false, - "reason": "Python 3.14 import/load failure is unrelated to an invalid backend name in `BACKENDS_MAPPING`." + "reason": "Python 3.14 import failure and a BACKENDS_MAPPING validation error are different compatibility issues." }, { "left": "issue:43986", "right": "issue:44016", "accept": false, - "reason": "Missing `torchvision` for `AutoProcessor` video models is unrelated to a notebook syntax error." + "reason": "AutoProcessor without torchvision and a notebook syntax error are unrelated." }, { "left": "issue:42907", "right": "issue:44655", "accept": false, - "reason": "Saving dequantized Ministral/Devstral models is not the same as saving `Pipeline` objects with `save_pretrained`." + "reason": "Saving dequantized models and saving Pipeline objects are different save_pretrained problems." }, { "left": "issue:42907", "right": "issue:45753", "accept": false, - "reason": "A save/dequantization bug is unrelated to an infinite-loop generation issue on a specific image." + "reason": "A save failure for dequantized models is unrelated to an image-specific infinite loop." }, { "left": "issue:43012", "right": "issue:43550", "accept": false, - "reason": "A PyTorch warning about bfloat16 precision is unrelated to Bamba failing under `torch.compile` with SDPA." - }, - { - "left": "issue:43825", - "right": "issue:44655", - "accept": false, - "reason": "The pipeline translation error message and pipeline `save_pretrained` support are distinct issues." + "reason": "A bfloat16 precision warning is unrelated to a torch.compile SDPA failure." }, { "left": "issue:45712", "right": "issue:46298", "accept": false, - "reason": "Leftover dummy classes leaking into `dir(transformers)` is unrelated to noisy logging." + "reason": "Dummy object leakage in repo checks is unrelated to a logging message." }, { "left": "issue:43012", "right": "issue:44368", "accept": false, - "reason": "These are both warnings, but they come from different configs and code paths (`bfloat16` compile warning vs `tie_word_embeddings`)." + "reason": "These are distinct warning cases with different triggers and model/config paths." }, { "left": "issue:43208", "right": "issue:43550", "accept": false, - "reason": "xLSTM training bugs for small models are unrelated to Bamba's SDPA/`torch.compile` failure." + "reason": "xLSTM training blockers are unrelated to the Bamba SDPA compile issue." + }, + { + "left": "issue:43825", + "right": "issue:44655", + "accept": false, + "reason": "A pipeline task-support error message is unrelated to Pipeline save_pretrained failures." }, { "left": "issue:43992", "right": "issue:44484", "accept": false, - "reason": "A missing embedding weight in `from_pretrained` is unrelated to the default 50GB shard size discussion." + "reason": "Missing weight loading and the 50GB shard-size question are unrelated save_pretrained topics." }, { "left": "issue:43352", "right": "issue:43976", "accept": false, - "reason": "Flash Attention 2 unsupported for NemotronH is unrelated to the Python 3.9/3.10 compatibility report." + "reason": "A model-specific Flash Attention limitation and a Python version compatibility issue are different problems." }, { "left": "issue:43352", "right": "issue:44380", "accept": false, - "reason": "Unsupported FA2 for NemotronH and GPT2 attention scaling under SDPA are different model/backend bugs." + "reason": "Nemotron Flash Attention support and GPT-2 attention scaling are unrelated model bugs." }, { "left": "issue:43066", "right": "issue:43906", "accept": false, - "reason": "Wrong tokenizer decoder type in v5 is not the same as an isolated reproduction of another issue." + "reason": "Tokenizer decoder type mismatch and an isolated reproduction of another issue are not the same bug." }, { "left": "issue:43097", "right": "issue:46298", "accept": false, - "reason": "Removed `tie_embeddings_and_encoder_decoder` and a noisy logging message are unrelated." - }, - { - "left": "issue:44162", - "right": "issue:46172", - "accept": false, - "reason": "ESM2 being broken and a DeepSeek-V4 transpose/dtype bug are separate model-specific failures." + "reason": "Removed tie_embeddings behavior and a noisy log message are unrelated." }, { "left": "issue:45584", "right": "issue:46365", "accept": false, - "reason": "Whisper empty-transcription failure is unrelated to the `Trainer` import problem." + "reason": "Whisper empty-transcription generation failure is unrelated to Trainer import conflicts." }, { - "left": "issue:44077", - "right": "issue:46365", + "left": "issue:44162", + "right": "issue:46172", "accept": false, - "reason": "`patchtsmixer` `post_init` policy and `Trainer` import failure are unrelated." + "reason": "ESM2 breakage and a DeepSeek tensor operation bug are unrelated." }, { - "left": "issue:44222", - "right": "issue:45478", + "left": "issue:44380", + "right": "issue:46132", "accept": false, - "reason": "FP8 MoE save issues are not the same as the Qwen3.5 MoE `from_pretrained` regression." + "reason": "GPT-2 backend scaling bug is unrelated to AttentionInterface.register changing function behavior." }, { - "left": "issue:36683", - "right": "issue:43582", + "left": "issue:44222", + "right": "issue:45478", "accept": false, - "reason": "Gemma3Config missing `vocab_size` and Apple Silicon `caching_allocator_warmup` TypeError are unrelated bugs." + "reason": "FP8 save_pretrained behavior and a MoE from_pretrained regression are related only by model family, not by the same underlying bug." } ] }, @@ -17241,7 +17282,7 @@ { "label": "issue", "component_index": 1, - "component_count": 25, + "component_count": 26, "review_unit_index": 56, "review_unit_count": 61, "status": "reviewed", @@ -17269,6 +17310,7 @@ "issue:43502", "issue:43525", "issue:43526", + "issue:43582", "issue:43611", "issue:43618", "issue:43756", @@ -17300,6 +17342,7 @@ "issue:46877" ], "soft_pairs": [ + "issue:36683|issue:43582", "issue:44380|issue:45479", "issue:42175|issue:46365", "issue:42757|issue:43824", @@ -17314,8 +17357,8 @@ "issue:43756|issue:44485", "issue:42994|issue:44655", "issue:36010|issue:42757", - "issue:43299|issue:43867", "issue:45276|issue:46077", + "issue:43299|issue:43867", "issue:44625|issue:46365", "issue:44483|issue:45290", "issue:43611|issue:44655", @@ -17334,232 +17377,238 @@ "issue:29870|issue:36296", "issue:43064|issue:44492" ], - "prepared_review_unit_hash": "4a75aa68110ea9ec433366328c9cbd73708e86a59f215dfdbe1581b39357a6cb", + "prepared_review_unit_hash": "b86b86580e6d3af60f1608d285a9dc43869536bcfa3128aaa44778d030061e72", "component_budget": { "node_count": 505, "item_count": 505, "soft_pair_count": 1854, - "serialized_chars": 962354, - "estimated_input_tokens": 240589, - "estimated_eval_tokens": 481434 + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 }, "budget": { - "node_count": 47, - "item_count": 47, - "soft_pair_count": 33, - "serialized_chars": 36816, - "estimated_input_tokens": 9204, - "estimated_eval_tokens": 18664 + "node_count": 48, + "item_count": 48, + "soft_pair_count": 34, + "serialized_chars": 37684, + "estimated_input_tokens": 9421, + "estimated_eval_tokens": 19098 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This is a mixed bag of unrelated Transformers issues. The soft links mostly reflect superficial similarities like import errors, config attributes, model loading, or save/pretrained regressions, but they do not look like the same underlying bug.", - "confidence": 0.95, - "canonical_issue_reason": "No canonical issue: the items span different bugs and regressions across unrelated subsystems and model families.", + "summary": "The cluster is highly heterogeneous: most pairs are only superficially related by broad themes like import errors, model loading/saving, or model-family regressions, but they do not describe the same underlying bug or change. No soft edge looks strong enough to merge.", + "confidence": 0.96, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "No single issue cleanly subsumes the cluster; at most 36683 is a discussion-heavy config-attribute bug, but it still does not represent the rest.", + "best_issue_reason": null, "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:36683", + "right": "issue:43582", + "accept": false, + "reason": "Both are runtime/configuration failures, but one is Gemma3 missing vocab_size and the other is an Apple Silicon TypeError in allocator warmup; unrelated bugs." + }, { "left": "issue:44380", "right": "issue:45479", "accept": false, - "reason": "Different problems: GPT-2 attention scaling backend behavior vs a sequence-classification loss degeneracy." + "reason": "Both involve training/inference behavior, but one is GPT2 attention scaling being ignored in SDPA/FlashAttention and the other is zero loss for single-label classification; different code paths." }, { "left": "issue:42175", "right": "issue:46365", "accept": false, - "reason": "Both are packaging/import-related, but one is missing TensorFlow in the torch extra and the other is a Trainer import conflict with datasets." + "reason": "Both mention import trouble, but one is a packaging/backend dependency issue and the other is Trainer import conflict with datasets; not the same problem." }, { "left": "issue:42757", "right": "issue:43824", "accept": false, - "reason": "Both are import errors, but for different missing symbols from different packages and unrelated code paths." + "reason": "Both are ImportErrors, but for different symbols and different causes: huggingface_hub is_offline_mode vs transformers Qwen2_5_VLForConditionalGeneration." }, { "left": "issue:43064", "right": "issue:44380", "accept": false, - "reason": "Training-state corruption under FSDP/PEFT is unrelated to GPT-2 attention scaling being ignored by SDPA/FlashAttention." + "reason": "One is an FSDP2/PEFT optimizer-state mismatch, the other is attention-scaling config ignored in backend kernels; unrelated regressions." }, { "left": "issue:42490", "right": "issue:43012", "accept": false, - "reason": "A local-path import warning is not the same bug as a PyTorch precision warning during bfloat16 compilation." + "reason": "A local path trailing-slash import warning is not the same as a bfloat16 compilation precision warning." }, { "left": "issue:44483", "right": "issue:44514", "accept": false, - "reason": "One is an API endpoint request-validation failure; the other is a chat template crash on batched inputs." + "reason": "Both touch chat/template flows, but one is an API /v1/chat/completions acceptance issue and the other is a batched apply_chat_template crash with padding=False." }, { "left": "issue:43618", "right": "issue:44380", "accept": false, - "reason": "CLIP attentions regression and GPT-2 attention scaling are different model/backends and different failure modes." + "reason": "CLIPOutput attentions missing is a model output regression; GPT2 attention scaling ignored is a backend-specific config bug." }, { "left": "issue:38617", "right": "issue:43502", "accept": false, - "reason": "Importing a moved helper is unrelated to unwanted network/API requests when local_files_only=True." + "reason": "Importing layer_type_validation and local_files_only still making API requests are different subsystems and failure modes." }, { "left": "issue:43208", "right": "issue:43526", "accept": false, - "reason": "xLSTM training bugs and BeitImageProcessorFast label reduction are unrelated subsystems." + "reason": "xLSTM training bugs and BeitImageProcessorFast reduce_labels returning one label are unrelated model/processor issues." }, { "left": "issue:44038", "right": "issue:44230", "accept": false, - "reason": "Same model family theme, but one is a loading regression and the other is fp8 inference support; not the same change." + "reason": "Both are Qwen3-VL/Qwen3.5-related, but one is a loading regression in transformers 5.0 and the other is FP8 inference support; not the same fix." }, { "left": "issue:43618", "right": "issue:43761", "accept": false, - "reason": "Both involve CLIP regressions, but one is missing attentions and the other is hidden_states being None; different code paths." + "reason": "Both are CLIP vision regressions, but one loses attentions and the other loses hidden_states under output_hidden_states=True; distinct outputs and code paths." }, { "left": "issue:43756", "right": "issue:44485", "accept": false, - "reason": "Both mention RoPE, but they concern different models and different implementation bugs." + "reason": "Smollm3 RoPE-layer loss and GLM-5 RoPE implementation discussion are different models and different issues." }, { "left": "issue:42994", "right": "issue:44655", "accept": false, - "reason": "Quantized model save/load breakage is not the same as inability to save Pipeline objects." + "reason": "Quantized model saving failure and Pipeline save_pretrained failure are both serialization-related, but one targets quantized models and the other Pipeline objects." }, { "left": "issue:36010", "right": "issue:42757", "accept": false, - "reason": "Both are import errors, but they involve different missing symbols from different modules." + "reason": "GenerationMixin import failure and is_offline_mode import failure are unrelated ImportErrors." }, { - "left": "issue:43299", - "right": "issue:43867", + "left": "issue:45276", + "right": "issue:46077", "accept": false, - "reason": "Both are load-model regressions, but one is tied to base_model_prefix and the other to sorted state_dict behavior." + "reason": "Both are Gemma4-related, but one is resize_token_embeddings propagation and the other is bidirectional-attention mask construction; different bugs." }, { - "left": "issue:45276", - "right": "issue:46077", + "left": "issue:43299", + "right": "issue:43867", "accept": false, - "reason": "Both are Gemma4-related, but one is embedding resizing and the other is attention mask construction." + "reason": "One is Qwen3VL MoE loading breakage and the other is a state_dict sorting load error; not the same underlying loading bug." }, { "left": "issue:44625", "right": "issue:46365", "accept": false, - "reason": "A Qwen3.5 config propagation bug is unrelated to a Trainer import conflict with datasets." + "reason": "Qwen3.5 num_labels propagation and Trainer import conflict with datasets are unrelated." }, { "left": "issue:44483", "right": "issue:45290", "accept": false, - "reason": "One is a chat-completions API acceptance issue; the other is a tokenizer/chat-template crash on tool-call messages." + "reason": "Both involve chat/template flows, but one is endpoint request validation and the other is tool-call assistant message handling in tokenize=True." }, { "left": "issue:43611", "right": "issue:44655", "accept": false, - "reason": "Model loading with base_model_prefix and saving Pipeline objects are different serialization paths and failures." + "reason": "base_model_prefix loading regression and Pipeline save_pretrained failure are different serialization/load paths." }, { "left": "issue:36683", "right": "issue:43475", "accept": false, - "reason": "Different missing attributes in different model families; not the same underlying config bug." + "reason": "Gemma3 missing vocab_size and SAM 3 Video missing fpn_position_embeddings are both attribute errors, but on different models and fields." }, { "left": "issue:43352", "right": "issue:43937", "accept": false, - "reason": "Flash Attention support for Nemotron and invalid GenerationConfig for GLM-5 are unrelated issues." + "reason": "Nemotron FlashAttention2 support and GLM-5 GenerationConfig validity are unrelated model-specific issues." }, { "left": "issue:44355", "right": "issue:46877", "accept": false, - "reason": "Compiled Python execution errors and auto_docstring import-time stdout are unrelated runtime/tooling problems." + "reason": "Compiled Python file errors and auto_docstring printing diagnostics at import time are separate import/runtime concerns." }, { "left": "issue:44655", "right": "issue:46050", "accept": false, - "reason": "Saving Pipeline objects and saving/loading a sinq-quantized model are distinct serialization failures." + "reason": "Pipeline save_pretrained and sinq quantized model round-trip save/load are both serialization issues, but not the same object or code path." }, { "left": "issue:44625", "right": "issue:46077", "accept": false, - "reason": "Different Gemma/Qwen config bugs with no shared code path." + "reason": "Qwen3.5 num_labels propagation and Gemma4 bidirectional attention masks are unrelated." }, { "left": "issue:36683", "right": "issue:44077", "accept": false, - "reason": "A missing Gemma3 config field is unrelated to patchtsmixer's post_init validation." + "reason": "Gemma3 missing vocab_size and patchtsmixer post_init allowance are unrelated config/model-definition issues." }, { "left": "issue:43976", "right": "issue:44938", "accept": false, - "reason": "Both are version/platform compatibility issues, but they target different Python versions and failure symptoms." + "reason": "Both are Python-compatibility reports, but one is about 5.1.0 on Python 3.9+ and the other is loading failure on Python 3.14; different compatibility problems." }, { "left": "issue:43867", "right": "issue:45216", "accept": false, - "reason": "State_dict sorting and Qwen3.5 checkpoint saving are different model-loading/saving regressions." + "reason": "A sorted state_dict load error and a broken Qwen3.5 save_pretrained checkpoint are both save/load regressions, but they are distinct failures." }, { "left": "issue:46077", "right": "issue:46082", "accept": false, - "reason": "Both are config issues, but for different models and different attributes/validation logic." + "reason": "Gemma4 causal-attention mask behavior and LlamaConfig head_dim validation are unrelated model/config bugs." }, { "left": "issue:43618", "right": "issue:45216", "accept": false, - "reason": "A CLIP attentions regression has nothing to do with Qwen3.5 save_pretrained checkpoint corruption." + "reason": "CLIPOutput attentions missing and Qwen3.5 checkpoint corruption after save_pretrained are unrelated regressions." }, { "left": "issue:36683", "right": "issue:43525", "accept": false, - "reason": "Different missing config attributes in different model families." + "reason": "Gemma3 missing vocab_size and Llama4 missing pad_token_id are similar in shape but refer to different model configs and separate fixes." }, { "left": "issue:36683", "right": "issue:44625", "accept": false, - "reason": "Both are config propagation bugs, but they concern different models and fields." + "reason": "Gemma3 vocab_size missing and Qwen3.5 num_labels propagation are unrelated attribute/config issues." }, { "left": "issue:29870", "right": "issue:36296", "accept": false, - "reason": "Bart generate() masking bug and tensor-parallel training bug are unrelated." + "reason": "BART generate decoder_attention_mask UnboundLocalError and tensor-parallel training bug are different execution paths and failures." }, { "left": "issue:43064", "right": "issue:44492", "accept": false, - "reason": "Optimizer/state issues under FSDP are unrelated to a typo in cache strategy docs/labels." + "reason": "FSDP2 optimizer-state corruption and a cache-strategies typo have nothing in common beyond both being in Transformers." } ] }, @@ -17573,7 +17622,7 @@ { "label": "issue", "component_index": 1, - "component_count": 25, + "component_count": 26, "review_unit_index": 57, "review_unit_count": 61, "status": "reviewed", @@ -17634,12 +17683,12 @@ ], "soft_pairs": [ "issue:43010|issue:45278", - "issue:44075|issue:46365", "issue:43992|issue:45310", + "issue:44075|issue:46365", "issue:44908|issue:46851", "issue:44393|issue:47000", - "issue:44938|issue:45071", "issue:29870|issue:33453", + "issue:44938|issue:45071", "issue:44355|issue:44855", "issue:43867|issue:46202", "issue:36683|issue:43064", @@ -17663,14 +17712,14 @@ "issue:44734|issue:46617", "issue:43723|issue:44556" ], - "prepared_review_unit_hash": "44caf915fdb234aa36d831815b7d1bc287b6147487340a6a5b2e83ee6b415939", + "prepared_review_unit_hash": "55710c92ba50fa9b0aa88057d6b85f56a7e9dd73bdf95825bf5eb6c7047a5aa4", "component_budget": { "node_count": 505, "item_count": 505, "soft_pair_count": 1854, - "serialized_chars": 962354, - "estimated_input_tokens": 240589, - "estimated_eval_tokens": 481434 + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 }, "budget": { "node_count": 48, @@ -17685,197 +17734,200 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This set is mostly a loose collection of unrelated Transformers issues. The only plausible duplicate edge is the import-time slowdown report versus the import-* removal/performance issue; the rest differ by model, code path, or failure mode.", - "confidence": 0.63, - "canonical_issue_reason": "Issue 44246 is the clearest user-facing symptom report for the only plausible duplicate area here: intermittent slowness when importing transformers.", + "summary": "This cluster is mostly a grab bag of unrelated Transformers regressions: model-loading/import failures, version-compatibility breakages, training bugs, tokenizer/processor issues, and a few CI/test flakiness reports. None of the soft-edge pairs look like the same concrete bug or change under strict duplicate triage.", + "confidence": 0.9, + "canonical_issue_reason": "issue:45278 is the broadest and most representative report of the cluster\u2019s dominant theme: post-upgrade import/load regressions in v5.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 44246 is the best representative anchor for this cluster because it describes the broad import-time regression directly and can absorb the related fix-oriented performance issue.", + "best_issue_reason": "issue:45278 is the best cluster anchor because it captures the shared high-level regression class without being overly model-specific.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43010", "right": "issue:45278", "accept": false, - "reason": "Different code paths: cache/layer update no_grad behavior versus broad import errors after upgrade." + "reason": "Different bug class: cache/layer update semantics vs broad import regressions after upgrade." }, { - "left": "issue:44075", - "right": "issue:46365", + "left": "issue:43992", + "right": "issue:45310", "accept": false, - "reason": "Unrelated failures: SGD optimizer args not used versus importing Trainer breaking with datasets installed." + "reason": "Different models and different from_pretrained failures; no shared concrete code path." }, { - "left": "issue:43992", - "right": "issue:45310", + "left": "issue:44075", + "right": "issue:46365", "accept": false, - "reason": "Both are loading errors, but they concern different models and likely different missing-key paths." + "reason": "Optimizer argument handling is unrelated to a Trainer import/dependency issue." }, { "left": "issue:44908", "right": "issue:46851", "accept": false, - "reason": "Scheduler kwargs bug versus Qwen3.5 variable-length position_ids failure; no shared underlying bug." + "reason": "Scheduler kwargs bug vs Qwen3.5 variable-length position_ids issue; unrelated." }, { "left": "issue:44393", "right": "issue:47000", "accept": false, - "reason": "Both are vision-related, but one is Qwen3-VL bbox output hallucination and the other is RTDetr loss infeasibility." + "reason": "Different tasks and failures: VLM bbox output vs RTDetr loss infeasibility." }, { - "left": "issue:44938", - "right": "issue:45071", + "left": "issue:29870", + "right": "issue:33453", "accept": false, - "reason": "Python 3.14 load failure and PretrainedConfig type-checking regression are different issues." + "reason": "BART generate decoder mask bug vs tokenizer-loading regression; distinct code paths." }, { - "left": "issue:29870", - "right": "issue:33453", + "left": "issue:44938", + "right": "issue:45071", "accept": false, - "reason": "Bart generate decoder_attention_mask error is unrelated to tokenizer-loading regression." + "reason": "Python-version load failure vs PretrainedConfig type-checking break; not the same defect." }, { "left": "issue:44355", "right": "issue:44855", "accept": false, - "reason": "Both mention import/runtime failures, but one is about compiled Python files and the other is a DebertaV2Model torch.jit parsing issue." + "reason": "Both are import-related, but one is compiled-file behavior and the other is a torch.jit parsing failure on 3.13." }, { "left": "issue:43867", "right": "issue:46202", "accept": false, - "reason": "State-dict sorting load error and Gemma pipeline roundtrip issue are different failure modes." + "reason": "State_dict sorting load error vs Gemma pipeline roundtrip issue; unrelated." }, { "left": "issue:36683", "right": "issue:43064", "accept": false, - "reason": "Gemma3Config missing vocab_size is unrelated to the FSDP2/PEFT optimizer-state bug." + "reason": "Gemma3 config missing vocab_size is unrelated to FSDP2/PEFT optimizer-state corruption." }, { "left": "issue:43867", "right": "issue:45375", "accept": false, - "reason": "Both involve model loading, but one is state-dict ordering and the other is a missing config field on Qwen3_5MoeVisionConfig." + "reason": "A load error and a config-field serialization issue are different problems." }, { "left": "issue:42890", "right": "issue:45561", "accept": false, - "reason": "Both are test flakiness, but one is missing set_seed coverage and the other is an xdist race on captured_info.txt." + "reason": "Seed-related flaky tests vs xdist file race; different test failures." }, { "left": "issue:41669", "right": "issue:44246", - "accept": true, - "reason": "These describe the same underlying performance problem: import transformers is slow, with 41669 pointing at import-* in models as the likely cause." + "accept": false, + "reason": "Import-star performance regression vs intermittent slow import; too broad to be the same bug." + }, + { + "left": "issue:43010", + "right": "issue:45850", + "accept": false, + "reason": "Specific no_grad/update bug vs vague \"many models break\" report; not the same issue." }, { "left": "issue:36296", "right": "issue:43575", "accept": false, - "reason": "Tensor parallel training bug versus OOM when loading a specific large Qwen2 model under tp; not the same bug." + "reason": "Tensor-parallel training bug vs TP OOM on one model; different concrete failures." }, { "left": "issue:43618", "right": "issue:44077", "accept": false, - "reason": "CLIPOutput attentions regression and patchtsmixer post_init validation are unrelated." + "reason": "CLIP attentions assignment bug vs patchtsmixer post_init policy change; unrelated." }, { "left": "issue:43010", "right": "issue:43976", "accept": false, - "reason": "Cache/layer no_grad decoration and Python 3.9+ load incompatibility are distinct issues." + "reason": "Cache update bug vs Python compatibility/load failure; unrelated." }, { "left": "issue:43867", "right": "issue:45357", "accept": false, - "reason": "Load-time state_dict sorting bug and save_pretrained visual-encoder key regression are different directions of the pipeline." + "reason": "Model load error vs save_pretrained key mismatch; opposite directions and different causes." }, { "left": "issue:43992", "right": "issue:45478", "accept": false, - "reason": "Different models and missing-key symptoms; one is UMT5Encoder, the other Qwen3.5 MoE loading." + "reason": "Both are from_pretrained complaints, but they involve different model families and different failures." }, { "left": "issue:28282", "right": "issue:33453", "accept": false, - "reason": "PyTorch missing ImportError for AutoModel is unrelated to tokenizer-loading regression." + "reason": "Missing PyTorch installation error vs tokenizer-loading regression; unrelated." }, { "left": "issue:43618", "right": "issue:46285", "accept": false, - "reason": "Both are CLIP-related, but one is attentions assignment and the other is structural divergence in text models." + "reason": "Both CLIP-related, but one is missing attentions assignment and the other is class-structure inconsistency." }, { "left": "issue:44360", "right": "issue:45440", "accept": false, - "reason": "DSA indexer ReLU discussion is unrelated to DeepseekV3MoE divergence from remote implementation." + "reason": "DSA indexer/ReLU discussion is unrelated to DeepseekV3MoE implementation divergence." }, { "left": "issue:42890", "right": "issue:44964", "accept": false, - "reason": "Test seeding flakiness is unrelated to loading Phi-4-multimodal-instruct on latest transformers." + "reason": "SAM-HQ test flakiness vs Phi-4 multimodal load failure; unrelated." }, { "left": "issue:42890", "right": "issue:44112", "accept": false, - "reason": "Both are CI/test reports, but they involve different tests and failure mechanisms." + "reason": "Seed-related test instability vs GraniteSpeech device-override CI issue; different bugs." }, { "left": "issue:42890", "right": "issue:45600", "accept": false, - "reason": "Seed-related test flakiness and auto_mappings references to removed configs are different CI failures." + "reason": "Seed/test flakiness vs auto_mappings referencing removed configs; unrelated." }, { "left": "issue:43502", "right": "issue:44062", "accept": false, - "reason": "local_files_only still triggering API calls is unrelated to AddedToken receiving duplicate special kwargs." + "reason": "local_files_only network leakage vs AddedToken keyword collision; different code paths." }, { "left": "issue:42898", "right": "issue:46664", "accept": false, - "reason": "Tokenizer cleanup behavior change and SmolVLM num_frames inconsistency are unrelated regressions." + "reason": "Tokenization cleanup behavior in v5 vs video processor num_frames semantics; unrelated." }, { "left": "issue:44734", "right": "issue:46617", "accept": false, - "reason": "KV-cache continuation tensor indexing and trust_remote_code cache FileNotFoundError are different server/loading bugs." + "reason": "KV-cache continuation crash vs cache loading FileNotFoundError; different runtime paths." }, { "left": "issue:43723", "right": "issue:44556", "accept": false, - "reason": "Both are v5 loading regressions, but one is tokenizer loading and the other is checkpoint reload compatibility; not the same bug." + "reason": "Tokenizer loading regression vs checkpoint reload compatibility issue; not the same underlying bug." } ] }, - "evaluator_result": { - "accept": true, - "feedback": "The summary is well grounded in the packet and stays conservative. The only accepted soft edge (41669 \u2194 44246) is supported by both titles being import-performance related, and the rejections avoid overclaiming shared bugs across different models/code paths." - }, - "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 }, { "label": "issue", "component_index": 1, - "component_count": 25, + "component_count": 26, "review_unit_index": 58, "review_unit_count": 61, "status": "reviewed", @@ -17953,8 +18005,8 @@ "issue:46664|issue:46726", "issue:43827|issue:44961", "issue:30333|issue:36296", - "issue:43299|issue:44556", "issue:29870|issue:31795", + "issue:43299|issue:44556", "issue:33453|issue:34689", "issue:44355|issue:45071", "issue:43576|issue:45071", @@ -17966,28 +18018,28 @@ "issue:44485|issue:44961", "issue:44351|issue:46617", "issue:36296|issue:43901", - "issue:43352|issue:44961", "issue:43576|issue:44938", + "issue:43352|issue:44961", "issue:43475|issue:46617", - "issue:46143|issue:46144", "issue:44222|issue:44230", + "issue:46143|issue:46144", "issue:41628|issue:42548", "issue:44315|issue:46209" ], - "prepared_review_unit_hash": "293539a426c627c7db47f088abd64132e470464f30ed1c0e380007066a1764bf", + "prepared_review_unit_hash": "b495925f766b53c42935fe42d6be0e0875eda9dd3b3724fe396a4ba236666774", "component_budget": { "node_count": 505, "item_count": 505, "soft_pair_count": 1854, - "serialized_chars": 962354, - "estimated_input_tokens": 240589, - "estimated_eval_tokens": 481434 + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 }, "budget": { "node_count": 47, "item_count": 47, "soft_pair_count": 39, - "serialized_chars": 38290, + "serialized_chars": 38291, "estimated_input_tokens": 9573, "estimated_eval_tokens": 19402 }, @@ -17996,246 +18048,246 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "These items are mostly unrelated issue reports that only share broad keywords like loading, import, v5, or fp8. I do not see any soft pair that clearly describes the same underlying bug/change, so I reject all candidate duplicates. If one issue has to serve as the cluster anchor, #43576 is the clearest high-signal user-facing regression.", - "confidence": 0.91, - "canonical_issue_reason": "#43576 is the most broadly actionable, user-facing regression in the set and has the strongest signal for being a representative anchor; the rest are largely unrelated issues rather than duplicates.", + "summary": "This cluster is mostly a loose collection of unrelated Transformers issues, with a weaker central theme around v5 loading/import/regression breakages. None of the soft pairs look like true duplicates or the same concrete fix path, so all should stay separate.", + "confidence": 0.84, + "canonical_issue_reason": "issue:43576 is the most central and broadly representative item: a v5 regression affecting the environment/CLI surface, which aligns with several other load/import/version-breakage reports in the cluster.", "canonical_pr_reason": null, - "best_issue_reason": "#43576 is the best standalone issue to keep because it is concrete, reproducible, and broadly relevant, even though the surrounding items do not form a true duplicate cluster.", + "best_issue_reason": "issue:43576 is the best single representative issue because it captures the broadest recurring theme in the cluster (v5 regression / breakage) without being overly narrow to one model or one symbol import.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44222", "right": "issue:45125", "accept": false, - "reason": "Both mention MoE/FP8/tensor parallelism, but one is about save_pretrained and the other about missing _tp_plan for loading/parallelism; different bugs." + "reason": "Both involve Qwen/MoE/FP8-related problems, but one is about save_pretrained behavior and the other about tensor-parallel planning/loading; not the same bug." }, { "left": "issue:43824", "right": "issue:46617", "accept": false, - "reason": "One is a missing Qwen export import error; the other is a cache FileNotFoundError for trust_remote_code. Different root causes." + "reason": "One is an import error for a specific model class, the other is a FileNotFoundError loading a trust_remote_code model from cache; different failure modes and code paths." }, { "left": "issue:43502", "right": "issue:44485", "accept": false, - "reason": "local_files_only network leakage and GLM-5 RoPE implementation are unrelated code paths." + "reason": "Local-files-only network leakage and GLM-5 RoPE implementation are unrelated issues." }, { "left": "issue:43874", "right": "issue:46617", "accept": false, - "reason": "Missing get_number_of_image_patches in GLM46V is unrelated to cache file lookup failures." + "reason": "Missing image-patch helper in a processor and cache FileNotFoundError are separate problems." }, { "left": "issue:43010", "right": "issue:43065", "accept": false, - "reason": "Cache/layer no_grad decoration and a dummy Conv2d in Sam3PixelDecoder are different fixes in different components." + "reason": "Cache update no_grad decoration and a dummy Conv2d in Sam3PixelDecoder are unrelated implementation bugs." }, { "left": "issue:44387", "right": "issue:44556", "accept": false, - "reason": "int4 quantization OOM and checkpoint reload failures after upgrade are separate regression classes." + "reason": "Int4 quantization memory growth and reloading checkpoints after upgrade are different regression types." }, { "left": "issue:43576", "right": "issue:46132", "accept": false, - "reason": "Broken env command and AttentionInterface.register behavior are unrelated." + "reason": "A broken env command in v5 and changed behavior of AttentionInterface.register are not the same concrete bug." }, { "left": "issue:45125", "right": "issue:45478", "accept": false, - "reason": "Missing _tp_plan for tensor parallelism and Qwen3.5 MoE from_pretrained errors are model-family-adjacent but not the same concrete bug." + "reason": "Both mention Qwen3.5 MoE, but one is missing _tp_plan for tensor parallelism while the other is a generic from_pretrained error; insufficient evidence of one root cause." }, { "left": "issue:30333", "right": "issue:33453", "accept": false, - "reason": "MLflow reporting on failed jobs and tokenizer loading regression are unrelated." + "reason": "MLFlow reporting on failed jobs and tokenizer loading regression are unrelated." }, { "left": "issue:44355", "right": "issue:44749", "accept": false, - "reason": "Compiled Python file errors and a preprocessing slowdown after upgrade are different problems." + "reason": "Compiled Python file errors and a tokenizer performance regression do not describe the same bug." }, { "left": "issue:42548", "right": "issue:44351", "accept": false, - "reason": "Both are import errors, but they involve different missing symbols and likely different export regressions." + "reason": "Both are import errors, but for different symbols and likely different missing exports." }, { "left": "issue:45125", "right": "issue:45375", "accept": false, - "reason": "Missing _tp_plan and missing deepstack_visual_indexes are distinct config/model issues." + "reason": "Missing tensor-parallel plan and missing config field silently dropped by strict validation are different issues." }, { "left": "issue:30990", "right": "issue:31795", "accept": false, - "reason": "Sentence Transformers loading hangs and documentation confusion about forward arguments are not the same bug." + "reason": "Loading gets stuck and confusing forward-docs wording are not the same underlying problem." }, { "left": "issue:43502", "right": "issue:44222", "accept": false, - "reason": "Network access despite local_files_only and FP8 save_pretrained are unrelated." + "reason": "Network access despite local_files_only and FP8 moe save_pretrained behavior are unrelated." }, { "left": "issue:43502", "right": "issue:44961", "accept": false, - "reason": "A real bug about local files only does not match the unrelated 'racoon' issue." + "reason": "API requests with local_files_only and the placeholder title 'racoon' have no relation." }, { "left": "issue:43010", "right": "issue:43232", "accept": false, - "reason": "Both touch generation/cache behavior, but one is about no_grad decoration and the other about sync_gpus/_update_model_kwargs_for_generation; different code-path failures." + "reason": "Both touch generation internals, but one is about no_grad on cache updates and the other about sync_gpus/model kwargs ordering; not the same concrete fix." }, { "left": "issue:46664", "right": "issue:46726", "accept": false, - "reason": "Video num_frames semantics and missing vision backend handling in AutoProcessor are different issues." + "reason": "A video num_frames semantic mismatch and an AutoProcessor vision-backend import error are different bugs." }, { "left": "issue:43827", "right": "issue:44961", "accept": false, - "reason": "Documentation still referencing pipeline() and 'racoon' are unrelated." + "reason": "Docs still referencing pipeline after removals is unrelated to the placeholder issue." }, { "left": "issue:30333", "right": "issue:36296", "accept": false, - "reason": "MLflow failure reporting and tensor parallel training bugs are unrelated." + "reason": "MLFlow failure reporting and tensor parallel training bug are unrelated." }, { - "left": "issue:43299", - "right": "issue:44556", + "left": "issue:29870", + "right": "issue:31795", "accept": false, - "reason": "Both are version-related loading regressions, but they concern different models and different failure modes." + "reason": "BART generate() UnboundLocalError and docs confusion around input_ids/past_key_values do not share the same code-path bug." }, { - "left": "issue:29870", - "right": "issue:31795", + "left": "issue:43299", + "right": "issue:44556", "accept": false, - "reason": "BART generate UnboundLocalError and forward-argument documentation confusion are not the same underlying bug." + "reason": "Qwen3VL MoE loading regression and general checkpoint reload failure after upgrade are related only at a high level, not the same underlying defect." }, { "left": "issue:33453", "right": "issue:34689", "accept": false, - "reason": "Tokenizer loading regression and Llama 3.2 vision model loading breakage are separate loader issues." + "reason": "Tokenizer loading regression and Llama 3.2 Vision model loading breakage are separate model-loading issues." }, { "left": "issue:44355", "right": "issue:45071", "accept": false, - "reason": "Compiled Python file errors and PretrainedConfig type checking breakage are unrelated." + "reason": "Compiled Python-file failures and PretrainedConfig type-checking breakage are unrelated." }, { "left": "issue:43576", "right": "issue:45071", "accept": false, - "reason": "Broken env command and type checking regressions are different subsystems and symptoms." + "reason": "A broken env command and PretrainedConfig type-checking issues are different regression surfaces." }, { "left": "issue:42757", "right": "issue:43576", "accept": false, - "reason": "Missing is_offline_mode import and the env command regression are unrelated." + "reason": "Missing huggingface_hub import and a broken transformers env command are not the same bug." }, { "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 not applied with model_init and a missing ReLU in the DSA indexer are unrelated." }, { "left": "issue:43576", "right": "issue:43824", "accept": false, - "reason": "Env command breakage and a Qwen model import error do not share the same bug." + "reason": "The env command regression and a specific model-class import error are different failures." }, { "left": "issue:31795", "right": "issue:38617", "accept": false, - "reason": "Docs confusion and an ImportError for layer_type_validation are not the same issue." + "reason": "Docs confusion in model.forward and missing layer_type_validation import are unrelated." }, { "left": "issue:44961", "right": "issue:45468", "accept": false, - "reason": "The placeholder racoon report and a Gemma-4 audio positional encoding bug are unrelated." + "reason": "Placeholder issue 'racoon' is unrelated to the Gemma-4 positional encoding bug." }, { "left": "issue:44485", "right": "issue:44961", "accept": false, - "reason": "GLM-5 RoPE implementation and the racoon placeholder issue are not connected." + "reason": "GLM-5 RoPE discussion and the placeholder issue are unrelated." }, { "left": "issue:44351", "right": "issue:46617", "accept": false, - "reason": "Missing HybridCache import and cache FileNotFoundError are both load-related but clearly different failures." + "reason": "Importing HybridCache and FileNotFoundError when loading from cache are different root causes." }, { "left": "issue:36296", "right": "issue:43901", "accept": false, - "reason": "Tensor parallel training bug and text-classification docs mentioning return_all_scores are unrelated." + "reason": "Tensor parallel training bug and TextClassificationPipeline docs mismatch are unrelated." }, { - "left": "issue:43352", - "right": "issue:44961", + "left": "issue:43576", + "right": "issue:44938", "accept": false, - "reason": "Flash Attention 2 support for NemotronH and the racoon issue are unrelated." + "reason": "Broken env command in v5 and Python 3.14 load failure are different version-related regressions, but not the same concrete issue." }, { - "left": "issue:43576", - "right": "issue:44938", + "left": "issue:43352", + "right": "issue:44961", "accept": false, - "reason": "Broken env command in v5 and Python 3.14 load failure are different regressions." + "reason": "Flash Attention 2.0 support for Nemotron and the placeholder issue are unrelated." }, { "left": "issue:43475", "right": "issue:46617", "accept": false, - "reason": "Sam3VisionEncoderOutput missing attribute and cache FileNotFoundError are unrelated." + "reason": "SAM 3 Video attribute error and trust_remote_code cache FileNotFoundError are different code paths." }, { - "left": "issue:46143", - "right": "issue:46144", + "left": "issue:44222", + "right": "issue:44230", "accept": false, - "reason": "Both are RoFormer-related, but one is kwargs propagation and the other is attention-interface usage; not the same concrete bug." + "reason": "Both mention FP8/MoE, but one is about save_pretrained and the other about inference support; not mergeable into one fix." }, { - "left": "issue:44222", - "right": "issue:44230", + "left": "issue:46143", + "right": "issue:46144", "accept": false, - "reason": "FP8 save_pretrained and FP8 inference support are different phases of the workflow and different bugs." + "reason": "Missing kwargs propagation and attention interface usage are closely related conceptually, but the titles describe different implementation gaps and not one concrete duplicate bug." }, { "left": "issue:41628", "right": "issue:42548", "accept": false, - "reason": "AutoImageProcessor and PreTrainedModel import errors involve different missing exports." + "reason": "Both are import errors, but for different top-level symbols and likely separate export regressions." }, { "left": "issue:44315", "right": "issue:46209", "accept": false, - "reason": "Liger Kernel not applied with model_init and FP8 inference failure on B200 are unrelated." + "reason": "Liger kernel application and FP8 inference on B200 are unrelated." } ] }, @@ -18249,7 +18301,7 @@ { "label": "issue", "component_index": 1, - "component_count": 25, + "component_count": 26, "review_unit_index": 59, "review_unit_count": 61, "status": "reviewed", @@ -18314,12 +18366,12 @@ "issue:44485|issue:46664", "issue:44273|issue:45542", "issue:44623|issue:45230", - "issue:44273|issue:45003", "issue:44038|issue:45125", + "issue:44273|issue:45003", "issue:44360|issue:46664", "issue:44393|issue:45071", - "issue:44393|issue:45753", "issue:41669|issue:43673", + "issue:44393|issue:45753", "issue:43576|issue:44556", "issue:43525|issue:45850", "issue:43572|issue:45850", @@ -18338,9 +18390,9 @@ "issue:44908|issue:45071", "issue:44295|issue:45230", "issue:44297|issue:45230", + "issue:41669|issue:44351", "issue:44492|issue:46328", "issue:44704|issue:44908", - "issue:41669|issue:44351", "issue:41669|issue:45341", "issue:43976|issue:45850", "issue:43976|issue:46218", @@ -18350,274 +18402,274 @@ "issue:44315|issue:44829", "issue:43519|issue:46655" ], - "prepared_review_unit_hash": "ba96c85a95fe82630f3d7146a39788b6afebc4e9a6025602155a7001cb47ceab", + "prepared_review_unit_hash": "6ec0fa1ff31965d04d3dafdc5db88468eb4ca475c3676af91b2c4a58bcc906d2", "component_budget": { "node_count": 505, "item_count": 505, "soft_pair_count": 1854, - "serialized_chars": 962354, - "estimated_input_tokens": 240589, - "estimated_eval_tokens": 481434 + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 40, - "serialized_chars": 39087, - "estimated_input_tokens": 9772, - "estimated_eval_tokens": 19800 + "serialized_chars": 39091, + "estimated_input_tokens": 9773, + "estimated_eval_tokens": 19802 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly a false-positive similarity set: the issues span unrelated regressions, model-specific bugs, import/load problems, and documentation/API changes. I found no soft pair that looks like a true duplicate or one mergeable underlying fix.", - "confidence": 0.86, - "canonical_issue_reason": "issue:45278 is the broadest concrete regression report here: it captures multiple import failures after upgrading from 4.57.0 to 5.5.0, which makes it the best umbrella issue among otherwise scattered v5 breakage reports.", + "summary": "This cluster is highly heterogeneous: it mixes documentation issues, model-specific bugs, loading/regression reports, and version-upgrade complaints. The proposed soft pairs are mostly false positives rather than true duplicates of the same underlying bug or change.", + "confidence": 0.89, + "canonical_issue_reason": "Issue 45850 is the closest thing to a cluster anchor because it is the broadest regression complaint ('Since 5.0 version it breaks too many models'). Even so, the set is too diverse for a strong true-canonical duplicate.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45278 is the most representative and triage-friendly issue in the set because it is concrete, version-scoped, and broad enough to stand in for several upgrade/import regressions without being overly vague.", + "best_issue_reason": "Issue 45850 is the best single representative of the cluster because it captures the widest v5 regression theme and has the most umbrella-like framing, but it should not be treated as a real duplicate target for the others.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:45362", "right": "issue:45850", "accept": false, - "reason": "Qwen3.5 chat crash is a specific model/runtime bug; the other is a generic complaint about v5 breaking many models." + "reason": "Different concrete problems: Qwen3.5 chat crash vs a broad v5 breakage complaint." }, { "left": "issue:43673", "right": "issue:44492", "accept": false, - "reason": "One is a cache-missing regression during chunked prefill; the other is a typo in cache strategy docs/labels." + "reason": "Chunked-prefill cache regression and a cache-strategy typo are unrelated." }, { "left": "issue:44485", "right": "issue:46664", "accept": false, - "reason": "GLM-5 RoPE implementation and SmolVLM video frame semantics are unrelated bugs in different code paths." + "reason": "GLM-5 RoPE implementation and SmolVLM video-frame padding are unrelated." }, { "left": "issue:44273", "right": "issue:45542", "accept": false, - "reason": "Lazy loading/import behavior and a TensorFlow backend installation error are different failure modes." + "reason": "Lazy loading bug vs TensorFlow/tensorboard backend error; different code paths." }, { "left": "issue:44623", "right": "issue:45230", "accept": false, - "reason": "A missing processor save file bug is not the same as an unscoped generic bug report." + "reason": "Specific processor.save_pretrained file bug vs an unspecified generic bug report." }, { - "left": "issue:44273", - "right": "issue:45003", + "left": "issue:44038", + "right": "issue:45125", "accept": false, - "reason": "Both touch import mechanics, but one is lazy loading not working and the other is unsafe sys.modules access; not the same defect." + "reason": "Qwen3-VL-Moe fp8 issue vs missing tensor-parallel plan; different functionality." }, { - "left": "issue:44038", - "right": "issue:45125", + "left": "issue:44273", + "right": "issue:45003", "accept": false, - "reason": "Different Qwen3-MoE issues: FP8/inference support versus missing tensor-parallel plan." + "reason": "Both involve loading/imports, but one is lazy loading and the other is unsafe sys.modules access." }, { "left": "issue:44360", "right": "issue:46664", "accept": false, - "reason": "DSA indexer/ReLU discussion is unrelated to video processor frame handling." + "reason": "DSA indexer/ReLU discussion is unrelated to SmolVLM video processing." }, { "left": "issue:44393", "right": "issue:45071", "accept": false, - "reason": "Qwen3-VL 2D box output errors are unrelated to PretrainedConfig type checking breakage." + "reason": "Qwen3-VL bbox output bug and PretrainedConfig type checking are unrelated." }, { - "left": "issue:44393", - "right": "issue:45753", + "left": "issue:41669", + "right": "issue:43673", "accept": false, - "reason": "Same model family, but one is 2D bbox hallucination/error and the other is an infinite loop on a specific image; different root problems." + "reason": "Import-* performance issue is not the same as the chunked-prefill cache regression." }, { - "left": "issue:41669", - "right": "issue:43673", + "left": "issue:44393", + "right": "issue:45753", "accept": false, - "reason": "Import-* slowdown is a lazy import/performance issue, not the same as missing cache in chunked prefill." + "reason": "Both are Qwen3-VL-family issues, but one is bbox output and the other is an infinite loop on a specific image." }, { "left": "issue:43576", "right": "issue:44556", "accept": false, - "reason": "Broken env CLI and checkpoint reload incompatibility are both v5 regressions but not the same code-path bug." + "reason": "Broken env command and checkpoint reload incompatibility are different v5 regressions." }, { "left": "issue:43525", "right": "issue:45850", "accept": false, - "reason": "Missing pad_token_id in one config is a specific API break, not the same as a generic 'v5 breaks too many models' complaint." + "reason": "Missing pad_token_id in Llama4Config is a model-specific config bug, not the broad v5 regression." }, { "left": "issue:43572", "right": "issue:45850", "accept": false, - "reason": "Different config attribute regression; too broad to treat as the same underlying issue." + "reason": "StableLm pad_token_idx regression is specific; it is not the same issue as the broad breakage report." }, { "left": "issue:44556", "right": "issue:45092", "accept": false, - "reason": "Both relate to loading old checkpoints after v5, but the affected models and failure causes differ." + "reason": "Both are upgrade/load compatibility complaints, but they affect different checkpoint-loading paths and failure modes." }, { "left": "issue:44492", "right": "issue:45850", "accept": false, - "reason": "A cache-strategy typo is not the same as a generic version-breakage report." + "reason": "Cache-strategy typo is unrelated to the general v5 breakage complaint." }, { "left": "issue:45278", "right": "issue:45850", "accept": false, - "reason": "Both are broad v5 regressions, but one is specifically about import errors after upgrade while the other is a vague general complaint." + "reason": "Broad import errors after upgrade is related in theme, but not the same concrete underlying bug." }, { "left": "issue:42371", "right": "issue:44492", "accept": false, - "reason": "TF32 API settings and cache strategy naming are unrelated." + "reason": "TF32 API/settings guidance and cache-strategy typo are unrelated." }, { "left": "issue:41669", "right": "issue:46218", "accept": false, - "reason": "Import-time slowdown from removing import * is not the same as failing to import a specific Qwen3_5 module." + "reason": "Import-* slowdown and missing Qwen3_5 module import are different import problems." }, { "left": "issue:44230", "right": "issue:46095", "accept": false, - "reason": "FP8 inference support for Qwen3-VL/MoE is different from a weight-conversion script question." + "reason": "FP8 inference support request and weight-conversion question are unrelated." }, { "left": "issue:43408", "right": "issue:44315", "accept": false, - "reason": "Model-type mismatch warning and Liger Kernel application during model_init are unrelated." + "reason": "Model-type mismatch warning and Liger kernel application are different code paths." }, { "left": "issue:44246", "right": "issue:44273", "accept": false, - "reason": "Import slowness and lazy-loading correctness are related broadly, but not the same concrete bug." + "reason": "Import slowdowns and lazy-loading malfunction overlap loosely, but not the same underlying bug." }, { "left": "issue:42890", "right": "issue:44908", "accept": false, - "reason": "Missing set_seed in tests is unrelated to inverse_sqrt scheduler kwargs handling." + "reason": "Test flakiness from missing seeds is unrelated to inverse_sqrt scheduler kwargs handling." }, { "left": "issue:43576", "right": "issue:44908", "accept": false, - "reason": "Broken env CLI and scheduler kwargs loss are different subsystems and defects." + "reason": "Env-command breakage and scheduler kwargs regression are unrelated." }, { "left": "issue:43352", "right": "issue:44393", "accept": false, - "reason": "Flash Attention 2 support error in Nemotron is unrelated to Qwen3-VL bounding-box generation issues." + "reason": "Flash Attention support limitation and Qwen3-VL bbox hallucination are different model issues." }, { "left": "issue:43976", "right": "issue:44393", "accept": false, - "reason": "Python version compatibility and Qwen3-VL output bugs are not the same problem." + "reason": "Python version compatibility and Qwen3-VL output errors are unrelated." }, { "left": "issue:44908", "right": "issue:45071", "accept": false, - "reason": "Scheduler kwargs handling and PretrainedConfig type checking are separate API issues." + "reason": "Scheduler kwargs bug and PretrainedConfig type-checking breakage are unrelated." }, { "left": "issue:44295", "right": "issue:45230", "accept": false, - "reason": "Buffer access error for position_ids is unrelated to the generic bug report." + "reason": "position_ids buffer access bug is unrelated to a generic bug report." }, { "left": "issue:44297", "right": "issue:45230", "accept": false, - "reason": "tokenizer_class mismatch in save_pretrained is a distinct tokenizer serialization bug." + "reason": "Tokenizer save_pretrained metadata mismatch is unrelated to a generic bug report." + }, + { + "left": "issue:41669", + "right": "issue:44351", + "accept": false, + "reason": "Import-* slowdown and HybridCache import failure are different issues." }, { "left": "issue:44492", "right": "issue:46328", "accept": false, - "reason": "Cache-strategy typo and SequenceFeatureExtractor padding slowdown are unrelated." + "reason": "Cache-strategy typo and SequenceFeatureExtractor.pad conversion inefficiency are unrelated." }, { "left": "issue:44704", "right": "issue:44908", "accept": false, - "reason": "AutoProcessor kwargs forwarding and inverse_sqrt scheduler kwargs are different code paths." - }, - { - "left": "issue:41669", - "right": "issue:44351", - "accept": false, - "reason": "Import * removal performance regression is not the same as HybridCache import failure." + "reason": "Both mention kwargs, but they affect different APIs and different behaviors." }, { "left": "issue:41669", "right": "issue:45341", "accept": false, - "reason": "Import-time slowdown and a testing_utils bug are unrelated." + "reason": "Import-* performance issue and testing_utils bug are unrelated." }, { "left": "issue:43976", "right": "issue:45850", "accept": false, - "reason": "Python compatibility breakage is not the same as a generic 'many models break' complaint." + "reason": "Python 3.9/3.10 compatibility issue is not the same as broad v5 model breakage." }, { "left": "issue:43976", "right": "issue:46218", "accept": false, - "reason": "Python version support and a missing Qwen3_5 module are unrelated." + "reason": "Python-version compatibility and missing Qwen3_5 import are unrelated." }, { "left": "issue:44077", "right": "issue:44861", "accept": false, - "reason": "Optional post_init enforcement and _get_tied_weight_keys list handling are different issues." + "reason": "Optional post_init policy change and tied-weights AttributeError are unrelated." }, { "left": "issue:31795", "right": "issue:33453", "accept": false, - "reason": "Documentation confusion about forward arguments is not a tokenizer loading regression." + "reason": "Documentation confusion in model.forward and tokenizer loading regression are unrelated." }, { "left": "issue:43576", "right": "issue:44273", "accept": false, - "reason": "Broken env command and lazy loading problems are both import/tooling-adjacent but not the same bug." + "reason": "Broken env command and lazy loading malfunction are both infrastructure-ish but not the same bug." }, { "left": "issue:44315", "right": "issue:44829", "accept": false, - "reason": "Liger Kernel application during model_init and flash_attention_3 training degeneration are different defects." + "reason": "Liger kernel application and flash_attention_3 training degeneration are unrelated." }, { "left": "issue:43519", "right": "issue:46655", "accept": false, - "reason": "Qwen3VL timestamp calculation and SmolVLM video padding calculation are separate video-processor bugs." + "reason": "Qwen3VL timestamp calculation and SmolVLM video padding are separate model-specific bugs." } ] }, @@ -18631,7 +18683,7 @@ { "label": "issue", "component_index": 1, - "component_count": 25, + "component_count": 26, "review_unit_index": 60, "review_unit_count": 61, "status": "reviewed", @@ -18703,8 +18755,8 @@ "issue:31795|issue:34689", "issue:43704|issue:44485", "issue:44485|issue:45468", - "issue:46018|issue:46172", "issue:42371|issue:43704", + "issue:46018|issue:46172", "issue:43502|issue:43519", "issue:45850|issue:46218", "issue:43519|issue:44623", @@ -18735,292 +18787,292 @@ "issue:43298|issue:44230", "issue:43717|issue:46144" ], - "prepared_review_unit_hash": "1f67e308582488ad022eb5fbf9a6debb032ab0a49c93134b178682adc0a4eba1", + "prepared_review_unit_hash": "4d429559e184a1b64544459b7f57992fdbd5d6ff8d7cd18d8ed6cdcea2f6aaa1", "component_budget": { "node_count": 505, "item_count": 505, "soft_pair_count": 1854, - "serialized_chars": 962354, - "estimated_input_tokens": 240589, - "estimated_eval_tokens": 481434 + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 43, - "serialized_chars": 39629, - "estimated_input_tokens": 9908, - "estimated_eval_tokens": 20072 + "serialized_chars": 39626, + "estimated_input_tokens": 9907, + "estimated_eval_tokens": 20070 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "These issues are mostly unrelated Transformers bugs, feature requests, and docs regressions. Several look semantically similar on the surface, but none are the same concrete underlying change, so I rejected all soft edges. The broad v5 regression report is the closest cluster anchor.", - "confidence": 0.87, - "canonical_issue_reason": "issue:45850 is the broadest umbrella-style report and is the closest thing to a cluster anchor among these mostly unrelated issues.", + "summary": "This cluster is mostly heterogeneous: the soft pairs share broad vocabulary like model bugs, loading, or tokenizer changes, but they do not look like the same underlying issue. I would not merge any of the proposed soft pairs.", + "confidence": 0.92, + "canonical_issue_reason": "No true duplicate hub is visible here; the items span unrelated bugs and feature requests. If a representative anchor is required, issue 45468 is the most concrete model-specific bug report among the set, but it should not be treated as a duplicate parent.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45850 is the best representative issue overall, but it is still too generic to serve as a precise duplicate target.", + "best_issue_reason": "Issue 45468 is the best single representative only because it is a specific, technically described bug with multiple nearby-but-not-duplicate neighbors; it is not a good canonical duplicate target for the whole cluster.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44360", "right": "issue:45468", "accept": false, - "reason": "Different model/component bugs; no shared code path." + "reason": "Different models and different failure modes; only superficial NLP overlap around indexing/encoding." }, { "left": "issue:44273", "right": "issue:44908", "accept": false, - "reason": "Lazy loading vs scheduler kwargs; unrelated subsystems." + "reason": "Lazy loading regression vs scheduler argument handling; unrelated code paths." }, { "left": "issue:44655", "right": "issue:44861", "accept": false, - "reason": "save_pretrained pipeline issue vs tied-weight key crash." + "reason": "Saving Pipeline objects is unrelated to tied-weight key handling." }, { "left": "issue:45230", "right": "issue:45310", "accept": false, - "reason": "Generic bug report vs specific Qwen3.5 MoE from_pretrained failure." + "reason": "A generic bug report vs a specific Qwen3.5 MoE from_pretrained error; not the same issue." }, { "left": "issue:45468", "right": "issue:46894", "accept": false, - "reason": "Different models and failure modes; not the same bug." + "reason": "Different models and different bugs; one is a positional encoding issue, the other is a training-loss shift problem." }, { "left": "issue:43576", "right": "issue:46218", "accept": false, - "reason": "CLI env command breakage vs model import error." + "reason": "Broken env command vs import failure for a specific model module; unrelated." }, { "left": "issue:43519", "right": "issue:44485", "accept": false, - "reason": "Processor timestamp bug vs RoPE implementation bug." + "reason": "Timestamp calculation in a processor vs RoPE implementation in GLM-5; different components." }, { "left": "issue:43519", "right": "issue:44908", "accept": false, - "reason": "Processor bug vs learning-rate scheduler bug." + "reason": "Processor timestamp math has nothing to do with scheduler kwargs propagation." }, { "left": "issue:41669", "right": "issue:43704", "accept": false, - "reason": "Import-star cleanup vs VRAM leak; different changes." + "reason": "Import star cleanup/performance issue is unrelated to a VRAM leak in Qwen3ForCausalLM." }, { "left": "issue:31795", "right": "issue:34689", "accept": false, - "reason": "Documentation confusion vs model loading regression." + "reason": "Documentation confusion vs a model loading breakage in Llama 3.2 vision; not the same bug." }, { "left": "issue:43704", "right": "issue:44485", "accept": false, - "reason": "VRAM leak vs RoPE implementation; unrelated." + "reason": "VRAM leak in one model and RoPE implementation in another are distinct." }, { "left": "issue:44485", "right": "issue:45468", "accept": false, - "reason": "Different model-specific implementation bugs." + "reason": "Both are model internals, but they concern different architectures and different fixes." }, { - "left": "issue:46018", - "right": "issue:46172", + "left": "issue:42371", + "right": "issue:43704", "accept": false, - "reason": "Both deepseek-v4, but distinct issues and code paths." + "reason": "TF32 configuration guidance is unrelated to a model VRAM leak." }, { - "left": "issue:42371", - "right": "issue:43704", + "left": "issue:46018", + "right": "issue:46172", "accept": false, - "reason": "TF32 settings change vs dataloader-thread memory leak." + "reason": "Shared-expert gating bug vs a tensor transpose/casting issue in DeepSeek-V4; not the same concrete defect." }, { "left": "issue:43502", "right": "issue:43519", "accept": false, - "reason": "Offline loading/network leak vs timestamp computation." + "reason": "Local-files-only network requests and timestamp computation are unrelated." }, { "left": "issue:45850", "right": "issue:46218", "accept": false, - "reason": "Broad v5 regressions vs a specific model import failure." + "reason": "Broad complaint about 5.0 breakages vs a specific import error for one model." }, { "left": "issue:43519", "right": "issue:44623", "accept": false, - "reason": "Processor timestamp bug vs missing files on save_pretrained." + "reason": "Processor timestamp logic and missing save_pretrained files are different problems." }, { "left": "issue:44263", "right": "issue:45468", "accept": false, - "reason": "Indexer split behavior vs audio positional encoding bug." + "reason": "A torch.split/indexing issue is not the same as Gemma-4 audio positional encoding." }, { "left": "issue:43606", "right": "issue:45468", "accept": false, - "reason": "CPU offload device mismatch vs Gemma-4 audio encoding." + "reason": "CPU offload device mismatch vs positional encoding bug; unrelated." }, { "left": "issue:43298", "right": "issue:46095", "accept": false, - "reason": "Continuous batching support vs weight conversion question." + "reason": "Continuous batching audio support and DeepSeek weight conversion are unrelated feature requests." }, { "left": "issue:43519", "right": "issue:43704", "accept": false, - "reason": "Processor timestamp bug vs VRAM leak." + "reason": "Processor timestamp bug and VRAM leak are different code paths." }, { "left": "issue:44351", "right": "issue:46218", "accept": false, - "reason": "Different missing import symbols and causes." + "reason": "Different import errors for different symbols/models." }, { "left": "issue:43352", "right": "issue:46218", "accept": false, - "reason": "Flash-attention support gap vs import failure." + "reason": "Flash Attention 2.0 support request for NemotronH vs Qwen module import failure; not the same issue." }, { "left": "issue:42913", "right": "issue:43066", "accept": false, - "reason": "Related tokenizer-v5 symptoms, but not clearly the same concrete bug." + "reason": "Both involve tokenizer/v5 changes, but one is a general behavior regression and the other is a specific decoder-type bug." }, { "left": "issue:28348", "right": "issue:45250", "accept": false, - "reason": "Specific GPT-2 FA2 support request vs generic Flash Attention topic." + "reason": "GPT2-specific FA2 support request is not clearly the same as a broader Flash Attention 2.0 topic." }, { "left": "issue:44908", "right": "issue:44998", "accept": false, - "reason": "Scheduler bug vs unrelated placeholder issue." + "reason": "Scheduler kwargs bug vs unrelated placeholder issue." }, { "left": "issue:44492", "right": "issue:44998", "accept": false, - "reason": "Cache typo vs unrelated placeholder issue." + "reason": "Cache strategy typo is unrelated to the placeholder issue." }, { "left": "issue:43618", "right": "issue:44998", "accept": false, - "reason": "CLIPOutput attention regression vs unrelated placeholder issue." + "reason": "CLIPOutput attentions regression is unrelated to the placeholder issue." }, { "left": "issue:44998", "right": "issue:45850", "accept": false, - "reason": "Unrelated placeholder issue vs broad v5 regression report." + "reason": "Placeholder issue has no technical overlap with broad 5.0 breakage complaints." }, { "left": "issue:44938", "right": "issue:44998", "accept": false, - "reason": "Python 3.14 load failure vs unrelated placeholder issue." + "reason": "Python 3.14 load failure is unrelated to the placeholder issue." }, { "left": "issue:44861", "right": "issue:45290", "accept": false, - "reason": "Tied-weights crash vs chat-template tool-call crash." + "reason": "Tied-weight key handling and chat template tool-call crash are different bugs." }, { "left": "issue:31515", "right": "issue:38617", "accept": false, - "reason": "Slow checkpoint loading vs import error; different root causes." + "reason": "Slow from_pretrained loading vs an ImportError; different failure modes." }, { "left": "issue:43572", "right": "issue:44861", "accept": false, - "reason": "Missing config field vs tied-weights crash." + "reason": "Missing pad_token_idx after v5 update is unrelated to tied-weight key crashes." }, { "left": "issue:43525", "right": "issue:44861", "accept": false, - "reason": "Llama4 config attribute error vs tied-weights crash." + "reason": "Llama4Config pad_token_id attribute error is unrelated to tied-weight key handling." }, { "left": "issue:44230", "right": "issue:44623", "accept": false, - "reason": "FP8 inference support request vs missing save_pretrained files." + "reason": "FP8 support request and processor.save_pretrained missing files are unrelated." }, { "left": "issue:28348", "right": "issue:43352", "accept": false, - "reason": "Both Flash Attention 2.0, but different models and support gaps." + "reason": "Both mention Flash Attention 2.0, but they are different model-specific support requests and not the same fix." }, { "left": "issue:29870", "right": "issue:31515", "accept": false, - "reason": "Generate-time decoder mask bug vs slow checkpoint loading." + "reason": "BART generate() decoder mask bug is unrelated to slow checkpoint loading." }, { "left": "issue:30333", "right": "issue:31515", "accept": false, - "reason": "MLflow reporting on failed jobs vs checkpoint loading speed." + "reason": "MLflow finalization bug and slow from_pretrained loading are unrelated." }, { "left": "issue:31515", "right": "issue:36296", "accept": false, - "reason": "Loading performance issue vs tensor-parallel training bug." + "reason": "Slow loading vs tensor-parallel training bug; no shared code path." }, { "left": "issue:31515", "right": "issue:31795", "accept": false, - "reason": "Checkpoint load slowdown vs documentation confusion." + "reason": "Performance/loading issue vs documentation confusion." }, { "left": "issue:44230", "right": "issue:45310", "accept": false, - "reason": "FP8 support request vs from_pretrained regression." + "reason": "FP8 inference support request is not the same as a Qwen3.5 MoE from_pretrained error." }, { "left": "issue:43298", "right": "issue:44230", "accept": false, - "reason": "Audio continuous batching support vs FP8 inference support." + "reason": "Audio continuous batching support and FP8 support are different feature requests." }, { "left": "issue:43717", "right": "issue:46144", "accept": false, - "reason": "Init-weights behavior vs attention-interface usage; unrelated code paths." + "reason": "Mamba init_weights behavior and RoFormer attention interface usage are unrelated implementation issues." } ] }, @@ -19034,7 +19086,7 @@ { "label": "issue", "component_index": 1, - "component_count": 25, + "component_count": 26, "review_unit_index": 61, "review_unit_count": 61, "status": "reviewed", @@ -19066,100 +19118,100 @@ "issue:37934|issue:43668", "issue:42936|issue:43668" ], - "prepared_review_unit_hash": "342b8a3cc0b71f45cd9dd0d12bbf6debcd6cff2ca65117eea7543f1b29e87298", + "prepared_review_unit_hash": "b9df463b070c48627378376d8cd559d177435c8eb9d04716716eb2abf127f25d", "component_budget": { "node_count": 505, "item_count": 505, "soft_pair_count": 1854, - "serialized_chars": 962354, - "estimated_input_tokens": 240589, - "estimated_eval_tokens": 481434 + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 }, "budget": { "node_count": 8, "item_count": 8, "soft_pair_count": 11, - "serialized_chars": 7778, - "estimated_input_tokens": 1945, - "estimated_eval_tokens": 4146 + "serialized_chars": 7775, + "estimated_input_tokens": 1944, + "estimated_eval_tokens": 4144 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is a loose collection of unrelated closed issues across different models and subsystems. The soft-similarity matches are superficial; none describe the same underlying bug or fix.", - "confidence": 0.97, - "canonical_issue_reason": "Issue 43240 is the most substantive and broadly bug-shaped report in the set, with multiple inbound references and a concrete runtime behavior problem. Even so, it is not truly representative of the others beyond generic wording.", + "summary": "The items are not one duplicate set; they span unrelated bugs across different models and components. The soft pairings only show weak textual similarity and do not describe the same concrete defect or change.", + "confidence": 0.95, + "canonical_issue_reason": "No single canonical issue: the cluster contains unrelated reports (norm config, tokenizer warnings, loss kwargs, typo, gating, attention interface, etc.), so there is no shared underlying bug to canonicalize.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 43240 is the best standalone issue to anchor the cluster because it describes an actionable code-path defect rather than a typo, config type hint, or model-specific question.", + "best_issue_reason": "issue:46144 is the most concrete, code-path-oriented report, but it is only the strongest standalone issue here, not a representative duplicate target for the whole cluster.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:42936", "right": "issue:46018", "accept": false, - "reason": "Different models and different bugs: Mask2Former ignore_value handling vs DeepSeek-V4 shared expert gating. No shared code path or concrete duplicate evidence." + "reason": "Different models and failures: Mask2Former ignore_value usage vs DeepSeek-V4 shared expert gating." }, { "left": "issue:43668", "right": "issue:44031", "accept": false, - "reason": "A config type-hint mistake is unrelated to tokenizer regex warnings. Different subsystems and different fix types." + "reason": "One is a type-hint issue in ModernBERTConfig; the other is a tokenizer regex warning regression." }, { "left": "issue:42936", "right": "issue:43240", "accept": false, - "reason": "Both mention loss-related behavior, but one is Mask2Former ignore_value usage and the other is kwargs propagation into loss calculation. Not the same bug." + "reason": "Unrelated bug classes: missing use of ignore_value vs kwargs not forwarded to a loss function." }, { "left": "issue:43240", "right": "issue:46144", "accept": false, - "reason": "Loss-function kwargs plumbing and RoFormer attention interface usage are unrelated code paths with no clear duplicate overlap." + "reason": "Different code paths and symptoms: loss-kwargs forwarding vs RoFormer attention interface usage." }, { "left": "issue:42936", "right": "issue:44031", "accept": false, - "reason": "Mask2Former loss ignore_value handling is unrelated to a tokenizer regex warning regression." + "reason": "Mask2Former loss parameter handling is unrelated to tokenizer regex warnings." }, { "left": "issue:42936", "right": "issue:46144", "accept": false, - "reason": "Different model components and failures: segmentation loss handling vs RoFormer attention implementation details." + "reason": "Different subsystems and fixes: Mask2Former loss logic vs RoFormer attention implementation." }, { "left": "issue:42936", "right": "issue:43813", "accept": false, - "reason": "Mask2Former loss behavior and a typo in a quantization-operation name are not the same underlying issue." + "reason": "A model-loss behavior bug is not the same as a typo in a quantization attribute name." }, { "left": "issue:43668", "right": "issue:43813", "accept": false, - "reason": "A ModernBERT config type hint bug and a quantization typo are distinct, isolated correctness issues." + "reason": "Both are small correctness issues, but in different configs/code paths with no shared underlying defect." }, { "left": "issue:43240", "right": "issue:43668", "accept": false, - "reason": "Passing kwargs to a loss function is a runtime behavior bug; incorrect type hints in config are a separate API annotation issue." + "reason": "Loss-function kwargs forwarding is unrelated to an incorrect type hint in a config class." }, { "left": "issue:37934", "right": "issue:43668", "accept": false, - "reason": "Llama4TextL2Norm normalization semantics and ModernBERT norm_eps typing are unrelated model-specific concerns." + "reason": "Different component and defect type: Llama4 norm behavior vs ModernBERT config typing." }, { "left": "issue:42936", "right": "issue:43668", "accept": false, - "reason": "Mask2Former ignore_value usage and ModernBERT config typing are different bugs in different modules." + "reason": "Mask2Former ignore_value handling and ModernBERT norm_eps typing are unrelated." } ] }, @@ -19173,7 +19225,7 @@ { "label": "issue", "component_index": 2, - "component_count": 25, + "component_count": 26, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", @@ -19192,52 +19244,52 @@ "issue:10519|issue:11005", "issue:11005|issue:11307" ], - "prepared_review_unit_hash": "b8e6a5d5263485ccabfe804a29233aa02e0ec57502b15fc78bdd8e8175696431", + "prepared_review_unit_hash": "de2bf2c8d2aa545ab07c34409fc85918ec612da013e8cead28b60088e3941ad7", "component_budget": { "node_count": 3, "item_count": 3, "soft_pair_count": 3, - "serialized_chars": 2876, - "estimated_input_tokens": 719, - "estimated_eval_tokens": 1694 + "serialized_chars": 2877, + "estimated_input_tokens": 720, + "estimated_eval_tokens": 1696 }, "budget": { "node_count": 3, "item_count": 3, "soft_pair_count": 3, - "serialized_chars": 2876, - "estimated_input_tokens": 719, - "estimated_eval_tokens": 1694 + "serialized_chars": 2877, + "estimated_input_tokens": 720, + "estimated_eval_tokens": 1696 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These three issues are not duplicates of one another; they only share weak lexical similarity. One is about tokenizer truncation direction, one asks for ReduceLROnPlateau-like scheduler behavior, and one asks for Wav2Vec2 word boundary timestamps.", - "confidence": 0.96, - "canonical_issue_reason": "Issue 10519 is the most concrete and self-contained request, but the cluster is heterogeneous, so there is no strong canonical duplicate here.", + "summary": "These three issues are not duplicates of one another; they cover unrelated Transformer truncation, learning-rate scheduling, and Wav2Vec2 word offsets.", + "confidence": 0.97, + "canonical_issue_reason": "No single canonical issue: the items discuss unrelated feature requests with no shared bug/change path.", "canonical_pr_reason": null, - "best_issue_reason": "10519 is the clearest standalone feature request; the other two discuss unrelated functionality in optimization scheduling and speech timestamping.", + "best_issue_reason": "None of the issues is a good cluster representative because the set is heterogeneous rather than duplicate-focused.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:10519", "right": "issue:11307", "accept": false, - "reason": "Different feature areas: truncation strategy for tokenization vs word-level time offsets in Wav2Vec2." + "reason": "Different features: tokenizer truncation behavior vs Wav2Vec2 timestamp/offset extraction." }, { "left": "issue:10519", "right": "issue:11005", "accept": false, - "reason": "Unrelated problems: tokenizer truncation behavior vs learning-rate scheduling functionality." + "reason": "Unrelated requests: truncation direction option vs ReduceLROnPlateau-like scheduler behavior." }, { "left": "issue:11005", "right": "issue:11307", "accept": false, - "reason": "No shared bug or change path; one is about scheduler behavior, the other about speech model timestamp extraction." + "reason": "No shared underlying bug or code path; scheduler functionality and Wav2Vec2 word offsets are distinct." } ] }, @@ -19251,7 +19303,7 @@ { "label": "issue", "component_index": 3, - "component_count": 25, + "component_count": 26, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", @@ -19269,7 +19321,7 @@ "issue:13244|issue:16618", "issue:15980|issue:16618" ], - "prepared_review_unit_hash": "10b60d719a3361ec5674af017e9d799dedb325f03342667b7cd6feebbacd8462", + "prepared_review_unit_hash": "27141eaf5ea5079e1610f4190ed7c645bcf48eb2388521a69f10f668d6cd4836", "component_budget": { "node_count": 3, "item_count": 3, @@ -19291,24 +19343,24 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "All three issues appear to describe different problems: Tapas tokenization mismatch, a bad auth/download error for private models, and a general inability to load bert-base-uncased. No duplicate pair is strong enough to merge.", + "summary": "These three closed issues do not appear to describe the same underlying bug: one is about Tapas tokenization mismatch, one about an auth/download error message for private models, and one about failing to load bert-base-uncased. They are at most loosely related to model loading, so I would not cluster them as duplicates.", "confidence": 0.94, - "canonical_issue_reason": "issue #16618 is the broadest and most central model-loading report in this cluster, so it is the best representative even though it is not a duplicate of the others.", + "canonical_issue_reason": "Issue #16618 is the broadest and most central user-facing failure report in this set, with the highest discussion activity; it is the best single representative if one is needed for the cluster, though it is not a duplicate of the others.", "canonical_pr_reason": null, - "best_issue_reason": "#16618 is the most general and user-visible failure case among the three, making it the best single issue to anchor the cluster.", + "best_issue_reason": "Issue #16618 is the most generally framed and the most active of the three, so it is the best issue to surface as the representative of this loose cluster.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:13244", "right": "issue:16618", "accept": false, - "reason": "Different subsystem and symptom: Tapas tokenization behavior vs. model loading for bert-base-uncased. No evidence of the same underlying bug." + "reason": "Different bug classes: tokenization mismatch versus model loading failure. They do not look like the same underlying issue." }, { "left": "issue:15980", "right": "issue:16618", "accept": false, - "reason": "#15980 is specifically about the error message when downloading a private model without login, while #16618 is about failing to load bert-base-uncased. The failure modes and contexts differ." + "reason": "One is an authentication/error-message problem when downloading a private model, the other is a generic inability to load a public model. Not the same bug." } ] }, @@ -19322,7 +19374,7 @@ { "label": "issue", "component_index": 4, - "component_count": 25, + "component_count": 26, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", @@ -19338,7 +19390,7 @@ "soft_pairs": [ "issue:26350|issue:33260" ], - "prepared_review_unit_hash": "92a071a806af500be2cde6e1c59cd7c2d96345e6f7001d1b15a7ff9277c7c365", + "prepared_review_unit_hash": "3640334fad19d25ec040c1fa22cda792d1d49cac3e477b57b7a2c38091873776", "component_budget": { "node_count": 2, "item_count": 2, @@ -19360,18 +19412,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are two separate enhancement requests: one for Flash Attention 2 support across more architectures, and one for GGUF support across more architectures. They share a high-level theme but target different features and code paths, so they should not be deduplicated.", - "confidence": 0.98, - "canonical_issue_reason": "Issue #26350 is the better canonical issue only because it is older and has more discussion activity; it can serve as the representative issue for the cluster, but it is not a duplicate of #33260.", + "summary": "These are two different feature-tracking issues in the same general area, not duplicates: one requests Flash Attention 2 support for more architectures, the other requests GGUF support for more architectures.", + "confidence": 0.96, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "Issue #26350 is the strongest representative issue due to age and higher community activity, but it addresses a different request than #33260.", + "best_issue_reason": "issue:26350 is the stronger representative issue because it is the older, more active tracker for a concrete support expansion request, but it is not a duplicate of issue:33260.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:26350", "right": "issue:33260", "accept": false, - "reason": "Both are broad feature requests about supporting more architectures, but one is for Flash Attention 2 and the other for GGUF. They involve different underlying functionality and are not the same bug/change." + "reason": "Different underlying work: Flash Attention 2 support vs GGUF support. They share a broad theme of architecture support, but not the same bug or change." } ] }, @@ -19385,7 +19437,7 @@ { "label": "issue", "component_index": 5, - "component_count": 25, + "component_count": 26, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", @@ -19436,12 +19488,12 @@ "issue:46222|issue:46931", "issue:46226|issue:46931" ], - "prepared_review_unit_hash": "816c28f1a012c75d0e2783f516ae3d9b441c3362d26b13451b5d6ef9cdf15a16", + "prepared_review_unit_hash": "5226c857bf02b4751de7e0a075a0f6c7dfbe76a3d868ff7073b61e298167dd10", "component_budget": { "node_count": 9, "item_count": 9, "soft_pair_count": 29, - "serialized_chars": 15983, + "serialized_chars": 15982, "estimated_input_tokens": 3996, "estimated_eval_tokens": 8248 }, @@ -19449,7 +19501,7 @@ "node_count": 9, "item_count": 9, "soft_pair_count": 29, - "serialized_chars": 15983, + "serialized_chars": 15982, "estimated_input_tokens": 3996, "estimated_eval_tokens": 8248 }, @@ -19458,195 +19510,195 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Most items are separate i18n translation requests, so the high textual similarity is mostly from a shared issue template rather than true duplicates. The only clear duplicate pair is the two identical Thai-titled issues; the generic placeholder-title issues also look like duplicates of each other.", - "confidence": 0.84, - "canonical_issue_reason": "Issue 43834 is the clearest representative of the duplicated placeholder-template group: it is the earliest generic '[i18n-] Translating docs to ' issue and matches the same near-identical template text seen in the other generic entries.", + "summary": "This cluster is mostly heterogeneous translation-request issues, but it does contain two distinct duplicate-like subgroups: the placeholder-title trio (43834/44365/44370) and the exact Thai-title pair (46222/46226). Those are the only clear duplicate sets; the rest are just broad i18n similarity across different languages/topics.", + "confidence": 0.82, + "canonical_issue_reason": "issue:43834 is the earliest and best anchor for the placeholder-title duplicate subgroup, but the cluster also contains a separate exact-duplicate pair at issue:46222/issue:46226, so no single issue fully dominates the whole cluster.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 43834 is the best cluster representative because its title is the cleanest generic template form and it appears to be the anchor for the repeated placeholder duplicates.", + "best_issue_reason": "issue:43834 is the strongest single representative because it sits in the more connected placeholder-title subgroup, while still acknowledging that issue:46222/issue:46226 is a separate exact-duplicate subgroup rather than part of the same one.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:27088", "right": "issue:44370", "accept": false, - "reason": "Different concrete language request vs placeholder template; similar wording only." + "reason": "Different language request versus a generic placeholder-titled issue; same broad theme, not the same underlying report." }, { "left": "issue:27088", "right": "issue:44365", "accept": false, - "reason": "Different concrete language request vs placeholder template; not the same issue." + "reason": "Different language request versus a generic placeholder-titled issue; same broad theme, not the same underlying report." }, { "left": "issue:27088", "right": "issue:43834", "accept": false, - "reason": "Turkish translation request is distinct from the generic placeholder entry." + "reason": "Turkish translation request versus a generic placeholder template; only broad i18n similarity." }, { "left": "issue:27088", "right": "issue:38435", "accept": false, - "reason": "Different target language; these are separate translation requests." + "reason": "Turkish and Romanian translation requests are different concrete tasks, despite shared category." }, { "left": "issue:27088", "right": "issue:28936", "accept": false, - "reason": "Different target language; same subsystem but not the same underlying issue." + "reason": "Turkish and Spanish translation requests are different concrete tasks, despite shared category." }, { "left": "issue:28936", "right": "issue:43834", "accept": false, - "reason": "Spanish translation request is distinct from the generic placeholder entry." + "reason": "Spanish request versus a generic placeholder-titled issue; not enough evidence they are the same issue." }, { "left": "issue:28936", "right": "issue:44365", "accept": false, - "reason": "Different concrete translation request vs placeholder template." + "reason": "Spanish request versus a generic placeholder-titled issue; not enough evidence they are the same issue." }, { "left": "issue:28936", "right": "issue:38435", "accept": false, - "reason": "Spanish and Romanian translation requests are separate items." + "reason": "Spanish and Romanian translation requests are different concrete tasks, even if related by topic." }, { "left": "issue:28936", "right": "issue:44370", "accept": false, - "reason": "Different concrete translation request vs placeholder template." + "reason": "Spanish request versus a generic placeholder-titled issue; not enough evidence they are the same issue." }, { "left": "issue:38435", "right": "issue:44370", "accept": false, - "reason": "Romanian translation request vs generic placeholder; not the same underlying issue." + "reason": "Romanian request versus a generic placeholder-titled issue; similar theme but not the same report." }, { "left": "issue:38435", "right": "issue:44365", "accept": false, - "reason": "Different concrete translation request vs placeholder template." + "reason": "Romanian request versus a generic placeholder-titled issue; similar theme but not the same report." }, { "left": "issue:38435", "right": "issue:43834", "accept": false, - "reason": "Romanian request is distinct from the generic placeholder entry." + "reason": "Romanian request versus a generic placeholder template; only broad i18n similarity." }, { "left": "issue:38435", "right": "issue:46222", "accept": false, - "reason": "Different titles and different translation request content; not clearly the same issue." + "reason": "Different Thai-titled issue and no concrete evidence it is the same underlying translation task." }, { "left": "issue:38435", "right": "issue:46226", "accept": false, - "reason": "Different titles and different translation request content; not clearly the same issue." + "reason": "Different Thai-titled issue and no concrete evidence it is the same underlying translation task." }, { "left": "issue:43834", "right": "issue:44370", "accept": true, - "reason": "Both are the same generic placeholder translation issue and appear duplicate-by-template." + "reason": "Identical placeholder title and near-identical body size; this looks like a repost of the same issue." }, { "left": "issue:43834", "right": "issue:44365", "accept": true, - "reason": "Same generic placeholder title/body pattern; looks like a duplicate copy." + "reason": "Identical placeholder title and near-identical body size; this looks like a repost of the same issue." }, { "left": "issue:43834", "right": "issue:46222", "accept": false, - "reason": "Placeholder template issue vs a different Thai-titled issue; not clearly the same artifact." + "reason": "Placeholder-template issue versus a Thai-titled issue; no evidence they describe the same report." }, { "left": "issue:43834", "right": "issue:46226", "accept": false, - "reason": "Placeholder template issue vs a different Thai-titled issue; not clearly the same artifact." + "reason": "Placeholder-template issue versus a Thai-titled issue; no evidence they describe the same report." }, { "left": "issue:43834", "right": "issue:46931", "accept": false, - "reason": "Different titles; similarity is too superficial to treat as a duplicate." + "reason": "Different title and only broad i18n similarity; not the same concrete issue." }, { "left": "issue:44365", "right": "issue:44370", "accept": true, - "reason": "Same generic placeholder translation issue; the two entries look interchangeable." + "reason": "Identical placeholder title and near-identical body size; this looks like a repost of the same issue." }, { "left": "issue:44365", "right": "issue:46222", "accept": false, - "reason": "Different title/content pattern; not enough evidence of the same issue." + "reason": "Generic placeholder title versus a Thai-titled issue; insufficient evidence of duplication." }, { "left": "issue:44365", "right": "issue:46226", "accept": false, - "reason": "Different title/content pattern; not enough evidence of the same issue." + "reason": "Generic placeholder title versus a Thai-titled issue; insufficient evidence of duplication." }, { "left": "issue:44365", "right": "issue:46931", "accept": false, - "reason": "Different titles and likely different reports; do not dedupe." + "reason": "Different title and only broad i18n similarity; not the same concrete issue." }, { "left": "issue:44370", "right": "issue:46222", "accept": false, - "reason": "Generic placeholder issue vs Thai-titled issue; not clearly the same underlying report." + "reason": "Generic placeholder title versus a Thai-titled issue; insufficient evidence of duplication." }, { "left": "issue:44370", "right": "issue:46226", "accept": false, - "reason": "Generic placeholder issue vs Thai-titled issue; not clearly the same underlying report." + "reason": "Generic placeholder title versus a Thai-titled issue; insufficient evidence of duplication." }, { "left": "issue:44370", "right": "issue:46931", "accept": false, - "reason": "Different titles and insufficient evidence of true duplication." + "reason": "Different title and only broad i18n similarity; not the same concrete issue." }, { "left": "issue:46222", "right": "issue:46226", "accept": true, - "reason": "Identical Thai title and matching metadata strongly indicate a duplicate." + "reason": "Exact same Thai title and identical body size; this is a clear duplicate issue report." }, { "left": "issue:46222", "right": "issue:46931", "accept": false, - "reason": "Thai-titled issues differ; similarity alone is not enough to merge." + "reason": "Both are Thai-titled, but the titles differ and there is no concrete evidence they are the same issue." }, { "left": "issue:46226", "right": "issue:46931", "accept": false, - "reason": "Thai-titled issues differ; similarity alone is not enough to merge." + "reason": "Both are Thai-titled, but the titles differ and there is no concrete evidence they are the same issue." } ] }, "evaluator_result": { - "accept": true, - "feedback": "The judgments are conservative and grounded in the packet: the identical Thai issues (46222/46226) are a clear duplicate pair, and the three placeholder-template issues (43834/44365/44370) share identical titles with near-identical bodies, supporting duplicate-by-template calls. The negative verdicts on the concrete language-request issues versus the template/Thai items are also appropriately cautious." + "accept": false, + "feedback": "The cluster summary is mostly grounded, but the canonical/best-representative reasoning overstates the evidence. The packet shows at least two duplicate-like subgroups: the placeholder-title trio (43834/44365/44370) and the exact Thai-title pair (46222/46226). So 43834 is not the only clearly repeated subgroup, and the rationale should not frame it that way. The soft-edge verdicts are generally conservative, but the representative selection needs to acknowledge the second duplicate-like subgroup and avoid implying a uniquely dominant one." }, "evaluator_used": true, - "retried": false, + "retried": true, "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null @@ -19654,7 +19706,7 @@ { "label": "issue", "component_index": 6, - "component_count": 25, + "component_count": 26, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", @@ -19736,238 +19788,238 @@ "issue:42503|issue:46653", "issue:45027|issue:46653" ], - "prepared_review_unit_hash": "30ebf7f2cc603557c8427f053e36393afa01847dbe8071a93fce517c8c6a2c44", + "prepared_review_unit_hash": "55bc652708a47394caa8c9888092d92df6274d7c56e446cc23b9cafc1710b164", "component_budget": { "node_count": 21, "item_count": 21, "soft_pair_count": 48, - "serialized_chars": 28611, - "estimated_input_tokens": 7153, - "estimated_eval_tokens": 14562 + "serialized_chars": 28618, + "estimated_input_tokens": 7155, + "estimated_eval_tokens": 14566 }, "budget": { "node_count": 21, "item_count": 21, "soft_pair_count": 48, - "serialized_chars": 28611, - "estimated_input_tokens": 7153, - "estimated_eval_tokens": 14562 + "serialized_chars": 28618, + "estimated_input_tokens": 7155, + "estimated_eval_tokens": 14566 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "This cluster is heterogeneous: the items are mostly distinct model/support requests that share only broad wording like \"Add model\" or \"support\". I found no true duplicate pairs, so all soft edges should be rejected.", - "confidence": 0.91, - "canonical_issue_reason": "No single issue is a real duplicate umbrella here; if a representative must be chosen, issue 36879 is the most generic and central model-add request, but only as a loose cluster stand-in.", + "summary": "These are mostly unrelated model-add/support requests that only look similar at the title level. I found no pair that clearly describes the same underlying bug or change, so all soft edges are rejected.", + "confidence": 0.96, + "canonical_issue_reason": "issue:46133 is the richest self-contained model-add request here and the best anchor for this mixed cluster.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 36879 is the best single representative because it is broad, clearly about adding a model, and sits near the center of the soft-similarity set\u2014though this still does not make the group duplicate-worthy.", + "best_issue_reason": "issue:46133 has the clearest scope and the strongest discussion/context, making it the best representative issue in this set.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:36718", "right": "issue:47004", "accept": false, - "reason": "Both are keypoint matching-related, but RoMa matcher and LoMa detector+matcher are different models/change requests, not the same concrete addition." + "reason": "Related model names, but RoMa matcher and LoMa detector+matcher are different requests and likely different code paths." }, { "left": "issue:36718", "right": "issue:41115", "accept": false, - "reason": "Keypoint matcher request vs MiniCPM3 architecture request; same repo theme only, not the same bug or feature." + "reason": "One is a keypoint matcher request, the other is a MiniCPM3 architecture request; not the same change." }, { "left": "issue:36817", "right": "issue:41211", "accept": false, - "reason": "EuroBert config addition and DEIMv2 support are unrelated model requests." + "reason": "EuroBert config support and DEIMv2 model support are separate model additions." }, { "left": "issue:36817", "right": "issue:43671", "accept": false, - "reason": "EuroBert config vs Qwen3-TTS support; different model families and modalities." + "reason": "BERT config work and Qwen3-TTS audio support are unrelated underlying changes." }, { "left": "issue:36817", "right": "issue:36879", "accept": false, - "reason": "EuroBert config and RF-DETR model are different additions with no shared concrete code path." + "reason": "EuroBert config and RF-DETR model support are different model families and implementation paths." }, { "left": "issue:36817", "right": "issue:42954", "accept": false, - "reason": "EuroBert config and XiaoMi MiMo are distinct model requests." + "reason": "EuroBert config and XiaoMi MiMo are distinct model requests with no shared code-path bug." }, { "left": "issue:36879", "right": "issue:41211", "accept": false, - "reason": "RF-DETR and DEIMv2 are separate vision models; same broad area, not the same change." + "reason": "RF-DETR and DEIMv2 are different detection models; same subsystem, different concrete change." }, { "left": "issue:36879", "right": "issue:43909", "accept": false, - "reason": "RF-DETR vision model request vs LFM2.5 Audio 1.5B audio model request; different modality and implementation." + "reason": "RF-DETR object detection and LFM2.5 Audio are unrelated model integrations." }, { "left": "issue:36879", "right": "issue:44121", "accept": false, - "reason": "RF-DETR model request and OpenAI weight-sparse transformer request are unrelated architectures." + "reason": "RF-DETR model support and weight-sparse transformer support are different architectural changes." }, { "left": "issue:36879", "right": "issue:43837", "accept": false, - "reason": "RF-DETR vision model and Qwen3-ASR support are different feature requests." + "reason": "RF-DETR vision model support and Qwen3-ASR audio support are unrelated requests." }, { "left": "issue:36879", "right": "issue:43671", "accept": false, - "reason": "RF-DETR vs Qwen3-TTS support; unrelated model families and code paths." + "reason": "RF-DETR and Qwen3-TTS target different model families and tasks." }, { "left": "issue:37015", "right": "issue:43837", "accept": false, - "reason": "NeoBERT addition and Qwen3-ASR support are different model requests." + "reason": "NeoBERT and Qwen3-ASR are different model additions, not one shared bug/change." }, { "left": "issue:37015", "right": "issue:44121", "accept": false, - "reason": "NeoBERT and weight-sparse transformer are distinct language-model changes." + "reason": "NeoBERT support and circuit-sparsity transformer support do not describe the same code path." }, { "left": "issue:37015", "right": "issue:43909", "accept": false, - "reason": "NeoBERT request vs LFM2.5 Audio 1.5B; different model types and modalities." + "reason": "NeoBERT and LFM2.5 Audio are unrelated model requests." }, { "left": "issue:36879", "right": "issue:37015", "accept": false, - "reason": "RF-DETR and NeoBERT are unrelated model additions despite similar title phrasing." + "reason": "RF-DETR and NeoBERT are separate model integrations with different implementation surfaces." }, { "left": "issue:37015", "right": "issue:42503", "accept": false, - "reason": "NeoBERT and ModernVBERT are different BERT-family models, not the same concrete feature." + "reason": "NeoBERT and ModernVBERT are distinct model additions, even if both are BERT-like." }, { "left": "issue:41115", "right": "issue:43837", "accept": false, - "reason": "MiniCPM3 architecture and Qwen3-ASR support are unrelated model requests." + "reason": "MiniCPM3 architecture work and Qwen3-ASR support are different model families/tasks." }, { "left": "issue:41115", "right": "issue:44121", "accept": false, - "reason": "MiniCPM3 architecture request vs weight-sparse transformer request; different implementation targets." + "reason": "MiniCPM3 architecture support and weight-sparse transformer support are not the same underlying change." }, { "left": "issue:41115", "right": "issue:42503", "accept": false, - "reason": "MiniCPM3 and ModernVBERT are different model additions." + "reason": "MiniCPM3 and ModernVBERT are separate architecture/model requests." }, { "left": "issue:41115", "right": "issue:45027", "accept": false, - "reason": "MiniCPM3 architecture and Voxtral TTS support are separate tasks with different code paths." + "reason": "MiniCPM3 architecture support and Voxtral TTS support are unrelated." }, { "left": "issue:41211", "right": "issue:43671", "accept": false, - "reason": "DEIMv2 and Qwen3-TTS are unrelated; shared \"add support\" wording is not enough." + "reason": "DEIMv2 and Qwen3-TTS are different model support requests." }, { "left": "issue:42503", "right": "issue:43837", "accept": false, - "reason": "ModernVBERT models and Qwen3-ASR support are different model families and modalities." + "reason": "ModernVBERT and Qwen3-ASR are different model families and tasks." }, { "left": "issue:42503", "right": "issue:43671", "accept": false, - "reason": "ModernVBERT and Qwen3-TTS are unrelated requests." + "reason": "ModernVBERT models and Qwen3-TTS support are unrelated additions." }, { "left": "issue:42503", "right": "issue:42954", "accept": false, - "reason": "ModernVBERT models and XiaoMi MiMo are different model additions." + "reason": "ModernVBERT and XiaoMi MiMo are separate model requests." }, { "left": "issue:42738", "right": "issue:45027", "accept": false, - "reason": "BERT-like models with RoPE and Voxtral TTS support are not the same underlying change." + "reason": "BERT-like models with RoPE and Voxtral TTS are not the same concrete change." }, { "left": "issue:42503", "right": "issue:42738", "accept": false, - "reason": "ModernVBERT models vs generic BERT-like models with RoPE; related area, but not duplicate concrete work." + "reason": "ModernVBERT models and generic BERT-like RoPE support are different requests." }, { "left": "issue:42738", "right": "issue:42954", "accept": false, - "reason": "BERT-like models with RoPE and XiaoMi MiMo are different requests." + "reason": "Generic BERT-like RoPE work and XiaoMi MiMo support are unrelated." }, { "left": "issue:42738", "right": "issue:46653", "accept": false, - "reason": "BERT-like models with RoPE and Mirko Privitera are unrelated." + "reason": "BERT-like RoPE support and an unrelated person-named issue are clearly not the same change." }, { "left": "issue:42954", "right": "issue:44205", "accept": false, - "reason": "XiaoMi MiMo and SAM3-LiteText are different model support requests." + "reason": "XiaoMi MiMo and SAM3-LiteText are different model additions." }, { "left": "issue:42954", "right": "issue:43671", "accept": false, - "reason": "MiMo and Qwen3-TTS are different models and tasks." + "reason": "XiaoMi MiMo and Qwen3-TTS support are unrelated model requests." }, { "left": "issue:42954", "right": "issue:45177", "accept": false, - "reason": "MiMo and DeepSeek-OCR2 are distinct model requests." + "reason": "XiaoMi MiMo and DeepSeek-OCR2 are separate model additions." }, { "left": "issue:42954", "right": "issue:46653", "accept": false, - "reason": "XiaoMi MiMo and Mirko Privitera are not the same issue." + "reason": "XiaoMi MiMo and this unrelated issue are not the same underlying change." }, { "left": "issue:43671", "right": "issue:43837", "accept": false, - "reason": "Qwen3-TTS and Qwen3-ASR share a family name but target different modalities and code paths." + "reason": "Qwen3-TTS and Qwen3-ASR share a vendor and modality, but they are different tasks and should not be merged as duplicates." }, { "left": "issue:43671", "right": "issue:45027", "accept": false, - "reason": "Qwen3-TTS support and Voxtral TTS support are different model integrations." + "reason": "Qwen3-TTS support and Voxtral-4B-TTS support are different model integrations." }, { "left": "issue:43671", @@ -19979,79 +20031,79 @@ "left": "issue:43671", "right": "issue:44121", "accept": false, - "reason": "Qwen3-TTS support and weight-sparse transformer support are unrelated." + "reason": "Qwen3-TTS support and weight-sparse transformer support are different code paths." }, { "left": "issue:43837", "right": "issue:43909", "accept": false, - "reason": "Qwen3-ASR support and LFM2.5 Audio 1.5B are both audio-related, but they are different models and separate features." + "reason": "Qwen3-ASR and LFM2.5 Audio are both audio-related, but they are distinct models and tasks." }, { "left": "issue:43837", "right": "issue:45322", "accept": false, - "reason": "Qwen3-ASR support and EUPE are unrelated model requests." + "reason": "Qwen3-ASR and EUPE are different model families and problem types." }, { "left": "issue:43837", "right": "issue:44121", "accept": false, - "reason": "Qwen3-ASR support and weight-sparse transformer support are different changes." + "reason": "ASR support and weight-sparse transformer support are unrelated." }, { "left": "issue:43909", "right": "issue:44205", "accept": false, - "reason": "LFM2.5 Audio 1.5B and SAM3-LiteText are different model additions." + "reason": "LFM2.5 Audio and SAM3-LiteText are different model additions with no shared implementation." }, { "left": "issue:43909", "right": "issue:46645", "accept": false, - "reason": "Audio model support and Granite SWA+sink option are not the same code-path problem." + "reason": "LFM2.5 Audio and Granite SWA+sink are unrelated changes." }, { "left": "issue:44121", "right": "issue:45322", "accept": false, - "reason": "Weight-sparse transformer request and EUPE are different model architectures." + "reason": "Weight-sparse transformer support and EUPE are distinct model requests." }, { "left": "issue:43837", "right": "issue:45027", "accept": false, - "reason": "Qwen3-ASR support and Voxtral TTS support are different audio-model integrations." + "reason": "Qwen3-ASR and Voxtral-4B-TTS are different audio model supports." }, { "left": "issue:37015", "right": "issue:45027", "accept": false, - "reason": "NeoBERT and Voxtral TTS are unrelated model requests." + "reason": "NeoBERT and Voxtral-4B-TTS are unrelated model additions." }, { "left": "issue:43837", "right": "issue:45177", "accept": false, - "reason": "Qwen3-ASR support and DeepSeek-OCR2 are unrelated." + "reason": "Qwen3-ASR and DeepSeek-OCR2 are not the same underlying change." }, { "left": "issue:45322", "right": "issue:46133", "accept": false, - "reason": "EUPE and TIPSv2 are both model proposals, but they are different models with different implementation targets." + "reason": "EUPE and TIPSv2 are both model requests, but they are different architectures from different projects." }, { "left": "issue:42503", "right": "issue:46653", "accept": false, - "reason": "ModernVBERT models and Mirko Privitera are unrelated." + "reason": "ModernVBERT models and this unrelated closed issue are not duplicates." }, { "left": "issue:45027", "right": "issue:46653", "accept": false, - "reason": "Voxtral TTS support and Mirko Privitera are unrelated." + "reason": "Voxtral-4B-TTS support and this unrelated closed issue are clearly different." } ] }, @@ -20065,7 +20117,7 @@ { "label": "issue", "component_index": 7, - "component_count": 25, + "component_count": 26, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", @@ -20081,7 +20133,7 @@ "soft_pairs": [ "issue:38740|issue:46443" ], - "prepared_review_unit_hash": "775839731d3002717140261b21504fd250f89da13e8450edb388bf378c14a5cf", + "prepared_review_unit_hash": "c7cf5521976580ca35c8687c1e41aaaa65477eeb461fda0505f6d79a5e47df3b", "component_budget": { "node_count": 2, "item_count": 2, @@ -20103,18 +20155,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These two issues are unrelated: one is a docs request about adding Pruna as an optimization framework, and the other is a Vision RoPE refactor. They should not be clustered as duplicates.", - "confidence": 0.98, - "canonical_issue_reason": null, + "summary": "The two issues are not duplicates and describe unrelated work: one is a closed docs request to add Pruna as an optimization framework, while the other is an open Vision RoPE refactor. There is no shared code path or concrete change overlap.", + "confidence": 0.97, + "canonical_issue_reason": "Issue 46443 is the better anchor only because it is the active technical change request; issue 38740 is a closed documentation-related request and unrelated to the refactor.", "canonical_pr_reason": null, - "best_issue_reason": null, + "best_issue_reason": "Issue 46443 is the most suitable representative issue in this cluster because it concerns a concrete code refactor, whereas issue 38740 is a separate docs enhancement.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:38740", "right": "issue:46443", "accept": false, - "reason": "Different topics and scopes: docs/framework listing vs. Vision RoPE code refactor; no shared bug or change path." + "reason": "Different subjects entirely: a docs add request for Pruna versus a Vision RoPE refactor. No evidence they track the same underlying bug or change." } ] }, @@ -20128,7 +20180,7 @@ { "label": "issue", "component_index": 8, - "component_count": 25, + "component_count": 26, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", @@ -20144,7 +20196,7 @@ "soft_pairs": [ "issue:41084|issue:43037" ], - "prepared_review_unit_hash": "863c256ec8623f859d33128dcdad0c8f8680af9cd0b5661840458f2e246d359f", + "prepared_review_unit_hash": "7f8f49be9869834a032b79e4f26da0bd7ba06a0fb3fd7dd8136a95034d88883c", "component_budget": { "node_count": 2, "item_count": 2, @@ -20166,18 +20218,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "The two issues do not look like duplicates: one is a generic request about block decoding, while the other is a model-specific request for DeepSeek v3.2 support. There are no shared files, targets, or concrete code-path indicators tying them together.", - "confidence": 0.94, - "canonical_issue_reason": null, + "summary": "The two issues are not duplicates: one is a generic decoder feature request, the other is a specific model support request. There is no clear shared underlying bug or change scope.", + "confidence": 0.95, + "canonical_issue_reason": "No canonical issue: the titles and implied scopes are unrelated, so neither issue can serve as a duplicate target for the other.", "canonical_pr_reason": null, - "best_issue_reason": "Issue #43037 is the more concrete and actionable request; #41084 is too terse and underspecified to serve as a representative canonical issue.", + "best_issue_reason": "Issue #43037 is the better representative only in the sense that it is more specific and actionable, but it is not a duplicate of #41084.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:41084", "right": "issue:43037", "accept": false, - "reason": "These are about different concerns: block decoding versus support for a specific model version. The titles and available metadata do not indicate the same underlying bug or change." + "reason": "Low textual overlap and different problem domains: block decoding behavior vs. DeepSeek v3.2 model support. They do not appear to describe the same underlying bug or change." } ] }, @@ -20191,7 +20243,7 @@ { "label": "issue", "component_index": 9, - "component_count": 25, + "component_count": 26, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", @@ -20207,12 +20259,12 @@ "soft_pairs": [ "issue:41306|issue:41826" ], - "prepared_review_unit_hash": "4a3928abe7d49c9268e11bb8a59b7a012b71cccb0f14746b03a2956c4563c0cd", + "prepared_review_unit_hash": "99e6a57116b83c3b8694d40c44334bddb9cc805f0030b0955761afae0580fbc5", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1583, + "serialized_chars": 1584, "estimated_input_tokens": 396, "estimated_eval_tokens": 1048 }, @@ -20220,7 +20272,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1583, + "serialized_chars": 1584, "estimated_input_tokens": 396, "estimated_eval_tokens": 1048 }, @@ -20229,18 +20281,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are not duplicates: one asks for a `from_pretrained()` flag for TiledFusedLogitsLoss/FLCE, while the other requests integrating TiledMLP for lower memory use. They concern different model components and different kinds of changes.", - "confidence": 0.97, - "canonical_issue_reason": "Issue 41306 is the more specific, actionable request: it targets a concrete `from_pretrained()` API flag for TiledFusedLogitsLoss/FLCE.", + "summary": "These two issues are related only by vague theme (tiled / memory-oriented model changes) but describe different requested capabilities: one is about exposing a TiledFusedLogitsLoss/FLCE flag in `from_pretrained()`, the other about integrating TiledMLP for lower memory use.", + "confidence": 0.93, + "canonical_issue_reason": "Issue 41306 is the more concrete, narrowly scoped request and is easier to use as a representative node for the cluster.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 41306 is the best representative issue in this cluster because it has a clear, narrowly scoped implementation target rather than a broader architecture optimization proposal.", + "best_issue_reason": "Issue 41306 is the better representative because it states a specific API/flag problem, whereas 41826 is a broader architectural feature request.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:41306", "right": "issue:41826", "accept": false, - "reason": "Different underlying requests: one is about exposing a FLCE/TiledFusedLogitsLoss flag in `from_pretrained()`, the other is about adding TiledMLP for memory reduction. They do not describe the same bug or change." + "reason": "Different underlying changes: one concerns enabling a FLCE/TiledFusedLogitsLoss option in `from_pretrained()`, the other concerns adding TiledMLP for memory savings. Similar subsystem, but not the same bug or change." } ] }, @@ -20254,7 +20306,7 @@ { "label": "issue", "component_index": 10, - "component_count": 25, + "component_count": 26, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", @@ -20270,7 +20322,7 @@ "soft_pairs": [ "issue:43598|issue:43600" ], - "prepared_review_unit_hash": "41ed49a78c2e97b0d72a8c7e84876678c7362ce1002cc384dee1039cd06cfe5f", + "prepared_review_unit_hash": "3ca343b2de5a32fb021cfb05036b85c303cc714d56a51b611ebdb1ceccb1be91", "component_budget": { "node_count": 2, "item_count": 2, @@ -20292,18 +20344,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These two closed issues both mention custom trainers, but they describe different Trainer code paths: one is about `remove_unused_column` and input-column pruning, the other is about FLOPS estimation. They are related at a high level but not duplicates.", - "confidence": 0.94, - "canonical_issue_reason": "issue #43598 is the more concrete, self-contained Trainer customization problem and is the clearer representative of the shared theme.", + "summary": "These two issues both mention custom trainer behavior, but they describe different underlying problems: one is about `remove_unused_column` customizability in `Trainer`, the other about FLOPS estimation for custom trainers. They are related at the subsystem level, but not duplicates.", + "confidence": 0.89, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "issue #43598 is the better cluster representative because it focuses on a specific Trainer behavior likely to need a code change, whereas #43600 is a separate FLOPS-accounting concern.", + "best_issue_reason": "Issue 43598 is the broader and more architectural Trainer customization request, so it is the better representative issue of the two, though it is not a duplicate of 43600.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43598", "right": "issue:43600", "accept": false, - "reason": "Both concern custom trainers, but one is about unused-column handling and the other about FLOPS estimation; they are different underlying bugs and would not be merged as the same issue." + "reason": "Both touch custom trainer behavior, but one concerns `remove_unused_column` customization while the other concerns FLOPS estimation logic; they are distinct bugs/feature requests and should not be merged." } ] }, @@ -20317,7 +20369,7 @@ { "label": "issue", "component_index": 11, - "component_count": 25, + "component_count": 26, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", @@ -20333,12 +20385,12 @@ "soft_pairs": [ "issue:43684|issue:44405" ], - "prepared_review_unit_hash": "93b0bc528f0bb2f8a776bc3a763080f3b0de808ec274f9eb2a275804fa8f841f", + "prepared_review_unit_hash": "42f306e0c815d50517ef1bfe7d3389e2fee8e6b3b135e98c022f5ed8803dd467", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1611, + "serialized_chars": 1612, "estimated_input_tokens": 403, "estimated_eval_tokens": 1062 }, @@ -20346,7 +20398,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1611, + "serialized_chars": 1612, "estimated_input_tokens": 403, "estimated_eval_tokens": 1062 }, @@ -20355,18 +20407,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are related to Qwen model support gaps, but they describe different missing integrations: Qwen3-Omni registration for generation APIs versus Qwen3.5 sequence-classification support. They are not duplicates of the same underlying bug/change.", - "confidence": 0.97, - "canonical_issue_reason": "No single canonical issue: 43684 is about AutoModel/AutoModelForConditionalGeneration registration for Qwen3-Omni, while 44405 is about AutoModelForSequenceClassification support for a different Qwen3.5 config.", + "summary": "These are related only at the high level of \"add model support,\" but they target different Qwen variants and different AutoModel heads, so they are not duplicates.", + "confidence": 0.96, + "canonical_issue_reason": "Issue 43684 is the broader and more central request: adding Qwen3-Omni registration for AutoModel and AutoModelForConditionalGeneration.", "canonical_pr_reason": null, - "best_issue_reason": "43684 is the broader representative of the cluster theme (missing Qwen auto-model registration), but it still does not cover the classification-specific request in 44405.", + "best_issue_reason": "43684 is the best representative issue for the cluster because it is the more general model-registration request and closest to the shared theme.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43684", "right": "issue:44405", "accept": false, - "reason": "Different model families and different code paths: one asks for Qwen3-Omni generation registration, the other for Qwen3.5 sequence-classification support. They should be tracked separately." + "reason": "Different underlying work: Qwen3-Omni AutoModel/AutoModelForConditionalGeneration registration vs Qwen3.5 AutoModelForSequenceClassification support. Different model family and different task head, so they should not be merged as one bug/change." } ] }, @@ -20380,7 +20432,7 @@ { "label": "issue", "component_index": 12, - "component_count": 25, + "component_count": 26, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", @@ -20403,84 +20455,87 @@ "issue:43845|issue:43859", "issue:43846|issue:43859" ], - "prepared_review_unit_hash": "4e4f6a0f0d33311cf385fcbb012dddbd8cd7175bf486965f27d94dd5dd235061", + "prepared_review_unit_hash": "afbd5ab0b46be2fe8f7eb47922bed36d63a8ef73170941c5d771a1f514b4b460", "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": 4262, + "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": 4262, + "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 sparse to confidently link any pair as true duplicates. Three issues share the same minimal title \"huggingface \", but there is no body detail, code target, or other concrete evidence tying them to the same underlying problem; \"Fett\" appears unrelated.", - "confidence": 0.37, - "canonical_issue_reason": "Issue #43845 is the best available representative only because it is the earliest of the repeated \"huggingface \" reports and has the same minimal title as the others.", + "summary": "The three issues titled \"huggingface\" look like duplicates of the same report; issue #43835 (\"Fett\") does not match that cluster. I\u2019d pick #43845 as the canonical issue because it is the earliest of the matching set.", + "confidence": 0.79, + "canonical_issue_reason": "#43845 is the earliest issue with the repeated \"huggingface\" title and matches the other two same-title reports closely.", "canonical_pr_reason": null, - "best_issue_reason": "Issue #43845 is the strongest cluster anchor: it matches the repeated title pattern and was created first among the similar issues.", + "best_issue_reason": "#43845 is the best representative of the duplicate cluster because it appears earliest and is textually aligned with #43846 and #43859.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43835", "right": "issue:43845", "accept": false, - "reason": "Titles are different and the available text is too minimal to show the same bug or request." + "reason": "Different titles; \"Fett\" does not look like the same underlying issue as \"huggingface\"." }, { "left": "issue:43835", "right": "issue:43846", "accept": false, - "reason": "Titles are different and there is no body/detail evidence of a shared underlying issue." + "reason": "Different titles; no evidence these report the same bug." }, { "left": "issue:43835", "right": "issue:43859", "accept": false, - "reason": "Titles are different and the sparse metadata does not indicate the same problem." + "reason": "Different titles; insufficient evidence of a shared underlying issue." }, { "left": "issue:43845", "right": "issue:43846", - "accept": false, - "reason": "The titles match, but both issues are nearly empty; that is not enough to confirm they describe the same underlying bug." + "accept": true, + "reason": "Same title, same topic, and created seconds apart; these look like duplicate reports." }, { "left": "issue:43845", "right": "issue:43859", - "accept": false, - "reason": "Same generic title, but no concrete description, target, or discussion to prove a true duplicate." + "accept": true, + "reason": "Same title and clearly overlapping report topic; likely a duplicate of the same issue." }, { "left": "issue:43846", "right": "issue:43859", - "accept": false, - "reason": "Same generic title, but the issues contain too little information to safely merge as duplicates." + "accept": true, + "reason": "Same title and same apparent report topic; likely duplicates of one issue." } ] }, - "evaluator_result": null, - "evaluator_used": false, + "evaluator_result": { + "accept": true, + "feedback": "Grounded in the packet: #43845 is the earliest of the repeated \"huggingface\" issues, #43835 has a different title, and the soft-similarity evidence supports grouping the three \"huggingface\" reports together. The verdicts stay conservative enough based on the available metadata." + }, + "evaluator_used": true, "retried": false, - "accepted_nontrivial_soft_edge": false, + "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null }, { "label": "issue", "component_index": 13, - "component_count": 25, + "component_count": 26, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", @@ -20507,88 +20562,88 @@ "issue:44572|issue:44937", "issue:44573|issue:44937" ], - "prepared_review_unit_hash": "b18f5028f503eb68989e7bf49171d32d73e6a13d3a7071e2c1c96563cd3c5163", + "prepared_review_unit_hash": "b1a02bd293f341689386bc585664ef2b25e4463c7b570020867009cba924bcdf", "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": 5782, + "estimated_input_tokens": 1446, + "estimated_eval_tokens": 3148 }, "budget": { "node_count": 5, "item_count": 5, "soft_pair_count": 9, - "serialized_chars": 5791, - "estimated_input_tokens": 1448, - "estimated_eval_tokens": 3152 + "serialized_chars": 5782, + "estimated_input_tokens": 1446, + "estimated_eval_tokens": 3148 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "This cluster is dominated by placeholder/spam issues with no substantive content. The three \"\" items are textually similar, but there is no actual bug/change description to justify duplicate merging; the other two issues are clearly unrelated.", - "confidence": 0.91, - "canonical_issue_reason": "No strong canonical issue: the items have either placeholder spam titles or unrelated generic titles, with no descriptions, repro steps, or references to a concrete problem.", + "summary": "All items are low-information spam/promotional issues, not a real technical duplicate set. The three \"\" issues are the closest match, but the remaining two are different kinds of promotional/support-style noise, so no soft-edge pair is safe to merge.", + "confidence": 0.92, + "canonical_issue_reason": "Issue 44371 is the best representative of the repeated generic spam placeholder issues: it is the earliest of the identical \"\" entries and matches the strongest duplicate-like pattern in the set.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 44572 is the most representative of the placeholder-spam subset, but it is still not a good canonical issue because none of the issues describe a real, actionable bug.", + "best_issue_reason": "44371 is the strongest issue canonical candidate because it is one of the exact \"\" duplicates and is earliest among them; the others are either the same placeholder or clearly different spam-like content.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44371", "right": "issue:44572", "accept": false, - "reason": "Both are placeholder \"\" issues, but there is no substantive bug/change content to confirm a true duplicate relationship." + "reason": "Same generic spam title, but no concrete bug/change is described, so there is no identifiable underlying issue to safely deduplicate." }, { "left": "issue:44371", "right": "issue:44573", "accept": false, - "reason": "Both are placeholder \"\" issues, but the lack of any actual issue description makes duplicate triage too weak to accept." + "reason": "These are both placeholder spam reports, but the record does not show a specific shared bug or change; too little substance to merge." }, { "left": "issue:44371", "right": "issue:44418", "accept": false, - "reason": "Different titles and no shared concrete problem; one is a spam placeholder and the other is a generic documentation suggestion." + "reason": "Both are noisy/low-signal, but one is a spam placeholder and the other is a documentation-suggestion title; not the same underlying issue." }, { "left": "issue:44371", "right": "issue:44937", "accept": false, - "reason": "Unrelated titles and no supporting details; a spam placeholder does not match \"Check out \\\"Google LLC\\\"\" as the same issue." + "reason": "A generic spam placeholder and a promotional-looking title are not evidence of the same concrete problem." }, { "left": "issue:44418", "right": "issue:44572", "accept": false, - "reason": "Generic documentation suggestion versus a spam placeholder; no concrete shared bug or change." + "reason": "Documentation-suggestion wording does not match the spam placeholder issue, and there is no shared technical bug description." }, { "left": "issue:44418", "right": "issue:44573", "accept": false, - "reason": "Generic documentation suggestion versus a spam placeholder; no concrete shared bug or change." + "reason": "Different low-information topics: doc enhancement suggestion versus spam placeholder; no same underlying issue is visible." }, { "left": "issue:44572", "right": "issue:44573", "accept": false, - "reason": "Textually identical placeholder spam titles, but there is still no actionable bug/change context to justify merging as a true duplicate." + "reason": "Both are identical spam placeholders, but identical spam text alone is not enough to prove the same underlying bug/change." }, { "left": "issue:44572", "right": "issue:44937", "accept": false, - "reason": "Unrelated titles and no shared concrete problem; placeholder spam does not match the Google LLC topic." + "reason": "Two different spam/promotional titles with no concrete technical problem to unify." }, { "left": "issue:44573", "right": "issue:44937", "accept": false, - "reason": "Unrelated titles and no shared concrete problem; placeholder spam does not match the Google LLC topic." + "reason": "The placeholder spam issue and the promotional title do not indicate the same issue or change request." } ] }, @@ -20602,7 +20657,7 @@ { "label": "issue", "component_index": 14, - "component_count": 25, + "component_count": 26, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", @@ -20618,7 +20673,7 @@ "soft_pairs": [ "issue:45145|issue:45151" ], - "prepared_review_unit_hash": "b2f7d7ae3919b87b5134f75df20b003d503a60596e985519f1750056d24838c9", + "prepared_review_unit_hash": "51e97648d845cc5f4f8fa4438ba73ea194c35aad1717551596fbe9a9aa246707", "component_budget": { "node_count": 2, "item_count": 2, @@ -20640,24 +20695,24 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "The two closed issues share an identical title and moderately similar bodies, which makes them plausible duplicate reports, but the packet does not prove they are the same underlying article/proposal or near-identical in content. The safest read is a tentative duplicate relationship, not a strong one. No PRs are present.", - "confidence": 0.69, - "canonical_issue_reason": "#45145 is the slightly better canonical anchor because it is earlier and has an inbound reference, so it is the more connected of the two. That said, the evidence only supports a cautious duplicate/related-issue inference, not a firm identity claim.", + "summary": "The two issues are highly similar and likely related, but the evidence is only soft similarity; there is not enough to treat them as definite duplicates. Issue #45145 is the better canonical representative because it was created first and already has an inbound reference.", + "confidence": 0.86, + "canonical_issue_reason": "Issue #45145 is the better canonical choice because it was created first and already has an inbound reference, making it the more established representative of this similar issue pair.", "canonical_pr_reason": null, - "best_issue_reason": "#45145 is the best issue representative for triage because it is earlier and has the only inbound reference, making it the more natural anchor if these are later-found duplicates.", + "best_issue_reason": "Issue #45145 is the strongest issue candidate overall due to earlier creation and existing inbound linkage.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:45145", "right": "issue:45151", "accept": false, - "reason": "Identical titles and soft body similarity are suggestive, but the packet does not establish that the reports describe the same underlying bug/change or that the bodies are effectively the same. This is not strong enough to merge as a duplicate pair." + "reason": "The titles match and the bodies appear similar, but the packet only provides soft similarity with no concrete duplicate evidence, so this is not safe to accept as a definitive duplicate." } ] }, "evaluator_result": { "accept": false, - "feedback": "The packet supports that the titles are identical and the bodies are very similar in length, with only soft-similarity evidence. It does not establish that the bodies are \u2018near-identical\u2019 or that they are the same \u2018energy-optimization article/proposal.\u2019 The duplicate/anchor inference should be stated more conservatively, and the high confidence is not grounded in the packet." + "feedback": "The packet supports that the two issues are very similar and that issue 45145 is the more established canonical choice (earlier creation, one inbound reference). However, the soft-edge verdict is too strong: the packet only provides soft similarity and the deterministic accept is false, so labeling them as definite duplicates is not conservative enough. The summary also overstates the evidence by adding \"spam issue,\" which is not grounded in the packet." }, "evaluator_used": true, "retried": true, @@ -20668,7 +20723,7 @@ { "label": "issue", "component_index": 15, - "component_count": 25, + "component_count": 26, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", @@ -20722,7 +20777,7 @@ { "label": "issue", "component_index": 16, - "component_count": 25, + "component_count": 26, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", @@ -20738,7 +20793,7 @@ "soft_pairs": [ "issue:45762|issue:45763" ], - "prepared_review_unit_hash": "f9033b400de9a37c367d3d509d843ec692da36d4b336d3318b3fd1bd9aca76ec", + "prepared_review_unit_hash": "21e2144ac963148ca74453492f013b33a858db6c330e5d3a0d3048092b13d955", "component_budget": { "node_count": 2, "item_count": 2, @@ -20760,18 +20815,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "The two closed issues reference different named individuals with different dates and no shared technical context, so they do not appear to describe the same underlying problem.", - "confidence": 0.99, - "canonical_issue_reason": "No canonical issue: the titles point to different people/dates and there is no overlapping bug or feature context.", + "summary": "Two closed issues contain only different personal-name/date titles and no body or linkage. There is no evidence they describe the same bug or change, so they should not be clustered as duplicates.", + "confidence": 0.96, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "Issue 45762 is only marginally more generic, but neither issue is a good representative because they appear unrelated.", + "best_issue_reason": null, "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:45762", "right": "issue:45763", "accept": false, - "reason": "Different names and dates; no evidence they refer to the same underlying issue." + "reason": "Only superficial surname overlap; the titles refer to different people and dates, with no shared context, body, or links indicating the same underlying issue." } ] }, @@ -20785,7 +20840,7 @@ { "label": "issue", "component_index": 17, - "component_count": 25, + "component_count": 26, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", @@ -20801,7 +20856,7 @@ "soft_pairs": [ "issue:45764|issue:45765" ], - "prepared_review_unit_hash": "b4e1335ef91d68726cd693dab6e65b2db05c15ebf57e5ff6189e87d9cee1e380", + "prepared_review_unit_hash": "5ebe17517ac735f7d673dc8f9d172084def9c92196a45e32aa3c201a9c4aa506", "component_budget": { "node_count": 2, "item_count": 2, @@ -20823,18 +20878,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "The two closed issues appear to be unrelated name/date entries, not duplicates of the same underlying bug or request. I would not merge them.", - "confidence": 0.97, - "canonical_issue_reason": "Issue #45764 is as suitable a representative as #45765; neither contains any substantive report content, but #45764 is the earlier-created item.", + "summary": "These two issues do not describe the same underlying bug or change; they are separate empty-name/date entries with only superficial title similarity.", + "confidence": 0.98, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "#45764 is the better global issue representative only because it is earlier and equally content-free; there is no meaningful technical signal to prefer either one.", + "best_issue_reason": null, "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:45764", "right": "issue:45765", "accept": false, - "reason": "The titles identify different people and different dates; there is no indication they refer to the same bug, request, or code path." + "reason": "The titles are just different personal-name/date strings with no shared problem statement, code path, or explicit linkage; this looks like unrelated content rather than duplicates." } ] }, @@ -20848,7 +20903,7 @@ { "label": "issue", "component_index": 18, - "component_count": 25, + "component_count": 26, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", @@ -20864,12 +20919,12 @@ "soft_pairs": [ "issue:45810|issue:46295" ], - "prepared_review_unit_hash": "83e7a758bf78f65b31469a545c7b4f87d20b4f35fe91fb7e87b80e3f795f4be9", + "prepared_review_unit_hash": "d223086690dc75f3c6119e8e5c8fbfd0584952a77e1c6cc4b051aa10adb271bc", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1542, + "serialized_chars": 1541, "estimated_input_tokens": 386, "estimated_eval_tokens": 1028 }, @@ -20877,7 +20932,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1542, + "serialized_chars": 1541, "estimated_input_tokens": 386, "estimated_eval_tokens": 1028 }, @@ -20886,18 +20941,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are not duplicates: one requests a new Qwen3.5 token-classification/value-model head, while the other asks for FP8 training support. They touch different code paths and goals.", - "confidence": 0.96, - "canonical_issue_reason": "No canonical issue for the cluster: the two issues describe separate feature requests with different implementation areas.", + "summary": "The two issues are not duplicates: one requests a new `Qwen3_5ForTokenClassification` model class for value-model use, while the other asks for FP8 training support. They concern different features and code paths.", + "confidence": 0.95, + "canonical_issue_reason": "Issue #45810 is the more specific and concrete implementation request; #46295 is broader and less tied to a single model/API addition.", "canonical_pr_reason": null, - "best_issue_reason": "issue #45810 is the more concrete, narrowly scoped request; #46295 is orthogonal and not a duplicate.", + "best_issue_reason": "Issue #45810 is the best representative of the cluster only because it is the most concrete feature request, but the two issues are not actually the same underlying problem.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:45810", "right": "issue:46295", "accept": false, - "reason": "Different features: model-head/token-classification support versus FP8 training support. They do not appear to describe the same underlying bug or change." + "reason": "Different underlying requests: adding a token-classification model class versus enabling FP8 training support. They do not describe the same bug or change path." } ] }, @@ -20911,7 +20966,7 @@ { "label": "issue", "component_index": 19, - "component_count": 25, + "component_count": 26, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", @@ -20937,12 +20992,12 @@ "issue:45871|issue:45873", "issue:45872|issue:45873" ], - "prepared_review_unit_hash": "faea4ab0435b03bc6ab62b6b2ddccd9d1f48777eba970e822b919c0af75fd455", + "prepared_review_unit_hash": "3e57b6dad31555b89633a64e43b9ef6b6c87560ad5c30f0edf70895acd013ece", "component_budget": { "node_count": 5, "item_count": 5, "soft_pair_count": 8, - "serialized_chars": 5722, + "serialized_chars": 5721, "estimated_input_tokens": 1431, "estimated_eval_tokens": 3118 }, @@ -20950,7 +21005,7 @@ "node_count": 5, "item_count": 5, "soft_pair_count": 8, - "serialized_chars": 5722, + "serialized_chars": 5721, "estimated_input_tokens": 1431, "estimated_eval_tokens": 3118 }, @@ -20959,66 +21014,66 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "The cluster appears to contain two strong duplicate issue pairs (45870/45871 and 45872/45873) plus one unrelated dataset request (45869). The duplicate pairs are highly similar by title and body length, but the evidence shown is still indirect, so they should be treated as strong duplicate candidates rather than asserted as proven exact duplicates.", - "confidence": 0.84, - "canonical_issue_reason": "45870 is the cleanest canonical issue choice because it is the earliest member of one of the strongest near-duplicate pairs (with 45871) and is a reasonable representative for that repeated request.", + "summary": "This cluster splits into two near-identical duplicate issue pairs (45870/45871 and 45872/45873) plus a separate, thematically related financial dataset promo issue (45869).", + "confidence": 0.96, + "canonical_issue_reason": "No single canonical issue covers the full cluster cleanly, because it contains two separate near-identical duplicate pairs and a separate financial dataset promo issue. If forced to choose one anchor, 45870 is the best representative of the prediction-example pair.", "canonical_pr_reason": null, - "best_issue_reason": "45870 is the best overall issue representative because it anchors a strong duplicate pair and is the earliest item among the near-identical issues; the cluster is mixed, so no single issue cleanly represents all five items.", + "best_issue_reason": "45870 is the best single issue representative only in a narrow sense: it anchors one near-identical duplicate pair and has the most visible activity among the example issues, but it does not represent the scraping-service pair or the dataset promo issue.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:45869", "right": "issue:45870", "accept": false, - "reason": "Different requests: a financial dataset bundle versus a DataForge time-series example; not the same underlying issue." + "reason": "Related by financial/DataForge theme, but the dataset bundle promo and the prediction example are different concrete issues." }, { "left": "issue:45869", "right": "issue:45871", "accept": false, - "reason": "Different requests: a dataset offer versus an example/tutorial issue; not a duplicate." + "reason": "The dataset bundle promo and the prediction example are thematically related, but not the same underlying issue." }, { "left": "issue:45870", "right": "issue:45871", "accept": true, - "reason": "Same title and same body length with very high soft similarity; strong duplicate candidate for the same issue." + "reason": "Same title and highly similar body length; this looks like a near-identical duplicate pair for the same prediction-example issue." }, { "left": "issue:45870", "right": "issue:45872", "accept": false, - "reason": "Related finance/DataForge theme, but the concrete requests differ: example/tutorial versus time-series plus scraping service." + "reason": "Both are about financial/DataForge topics, but one is a prediction example and the other is a scraping-service request; different concrete issues." }, { "left": "issue:45870", "right": "issue:45873", "accept": false, - "reason": "Related theme, but this is a different concrete request than the example/tutorial issue." + "reason": "Broad theme overlaps, but the prediction-example issue and the scraping-service issue are different concrete requests." }, { "left": "issue:45871", "right": "issue:45872", "accept": false, - "reason": "Similar domain wording, but the concrete requests do not match." + "reason": "Different reported topics: prediction example versus scraping service. Not the same underlying issue." }, { "left": "issue:45871", "right": "issue:45873", "accept": false, - "reason": "Similar domain wording, but the concrete requests do not match." + "reason": "Only broad thematic overlap; the issues describe different concrete requests." }, { "left": "issue:45872", "right": "issue:45873", "accept": true, - "reason": "Same title and same body length with very high soft similarity; strong duplicate candidate for the same issue." + "reason": "Same title and highly similar body length; this looks like a near-identical duplicate pair for the scraping-service issue." } ] }, "evaluator_result": { "accept": false, - "feedback": "The cluster-level conclusion is broadly plausible, but the response overstates the evidence by calling 45870\u219445871 and 45872\u219445873 \"exact duplicate\" pairs. In the packet we only have identical titles/body lengths plus soft similarity; no body text is shown and deterministic_accept is false for all pairs. A more conservative wording would be \"very likely duplicate\" or \"strong duplicate candidate.\" The canonical choice of 45870 as earliest in its pair is grounded, but the exact-duplicate assertion is not sufficiently supported." + "feedback": "The clustering read is mostly grounded, but a couple of phrasings are a bit too strong for the packet. In particular, 45869 is better described as a separate, thematically related issue rather than \"unrelated,\" and the two matching pairs are supported as very strong duplicates, but \"exact duplicates\" is slightly stronger than the visible evidence unless the underlying body text is known. Please keep the summary more conservative: two near-identical duplicate pairs plus a separate financial dataset promo issue." }, "evaluator_used": true, "retried": true, @@ -21029,7 +21084,7 @@ { "label": "issue", "component_index": 20, - "component_count": 25, + "component_count": 26, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", @@ -21045,7 +21100,7 @@ "soft_pairs": [ "issue:46439|issue:46441" ], - "prepared_review_unit_hash": "f8fdf1eba8cadf6cc6c8460e46b38912c106c8e70b015bd03a3abfd86a7011d7", + "prepared_review_unit_hash": "400a7d13ed5b4c7b91f1c73d9b69b35d6ead1bef662a6f3ed2e1a01c27104f9a", "component_budget": { "node_count": 2, "item_count": 2, @@ -21067,24 +21122,27 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Both issues involve `cache_implementation=\"static\"` and `linear_attention`, but they describe different failure modes: one reports shape corruption/runtime errors during `early_initialization`, the other a `KeyError` in Qwen3.5. They are related, but not clearly the same underlying bug.", - "confidence": 0.74, - "canonical_issue_reason": "Issue 46439 is the better canonical issue because it describes a more general/static-cache failure mode and names the likely offending path (`early_initialization`) rather than a model-specific symptom.", + "summary": "The two issues are related around static cache and linear-attention crashes, but the available evidence is too thin to treat them as the same underlying bug. #46439 is broader; #46441 reads like a specific Qwen3.5 KeyError symptom that is not proven to share the same root cause.", + "confidence": 0.87, + "canonical_issue_reason": "No canonical duplicate should be selected from the available evidence: #46439 is the broader report, but the titles alone do not establish that #46441 is the same bug rather than a related static-cache failure mode.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 46439 is the best issue to anchor the cluster because it is broader and more likely to represent the root cause, whereas 46441 is a narrower downstream crash report.", + "best_issue_reason": "#46439 is the best representative issue because it describes the more general mechanism and failure class, whereas #46441 is narrower and product-specific.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:46439", "right": "issue:46441", "accept": false, - "reason": "Related subsystem and cache mode, but the reported errors differ (`RuntimeError` from wrong shapes vs `KeyError: 'linear_attention'`) and the reports do not clearly point to the same concrete bug or code path." + "reason": "Reject: the titles suggest a soft relationship, but they do not demonstrate the same concrete bug. #46439 mentions early initialization corrupting linear-attention shapes; #46441 reports a Qwen3.5 static-cache KeyError. Related subsystem, but not enough to merge as one issue." } ] }, - "evaluator_result": null, - "evaluator_used": false, - "retried": false, + "evaluator_result": { + "accept": false, + "feedback": "Reject: the packet only supports a soft-similarity relationship between the two issues, and the analyst overstates the evidence by asserting a shared root cause (\u201csame static-cache/linear-attention initialization bug\u201d) that is not demonstrated here. Based on the titles alone, 46439 and 46441 are related, but the specific mechanism and equivalence are not sufficiently grounded to accept a same-bug merge verdict." + }, + "evaluator_used": true, + "retried": true, "accepted_nontrivial_soft_edge": false, "error_kind": null, "error_message": null @@ -21092,7 +21150,7 @@ { "label": "issue", "component_index": 21, - "component_count": 25, + "component_count": 26, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", @@ -21108,12 +21166,12 @@ "soft_pairs": [ "issue:46551|issue:46552" ], - "prepared_review_unit_hash": "e2dcbfb319d00cb8072601846e0d2d2023201b03ea793dc3c997e78609e5bb58", + "prepared_review_unit_hash": "96ca8d09c18e12d6f371ae516eeed79225ed573f1a5f1072ba83da63f9c00ea0", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1706, + "serialized_chars": 1707, "estimated_input_tokens": 427, "estimated_eval_tokens": 1110 }, @@ -21121,7 +21179,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1706, + "serialized_chars": 1707, "estimated_input_tokens": 427, "estimated_eval_tokens": 1110 }, @@ -21130,24 +21188,24 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "The two issues are effectively identical duplicates describing the same Wav2Vec2CTCTokenizer v5 word-delimiter special-token regression. The open issue is the better canonical target.", + "summary": "These two issues are duplicates of the same Wav2Vec2CTCTokenizer v5 special-token leak bug. The open issue is the better tracker.", "confidence": 0.99, - "canonical_issue_reason": "Issue #46552 is the open duplicate with the same title/problem statement, so it is the best place to retain active tracking.", + "canonical_issue_reason": "Issue #46552 is open and already has inbound references, so it is the better canonical tracker for the duplicate report.", "canonical_pr_reason": null, - "best_issue_reason": "#46552 is the strongest issue to keep because it is still open and appears to be the same underlying bug as #46551.", + "best_issue_reason": "#46552 is the best issue because it is the active open report for the exact same bug and is more likely to collect follow-up discussion and fixes.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:46551", "right": "issue:46552", "accept": true, - "reason": "Same title and same described tokenizer bug; these are clearly duplicate reports of the same issue." + "reason": "Same title, same described behavior, and no distinguishing details; these are the same underlying bug report." } ] }, "evaluator_result": { "accept": true, - "feedback": "The conclusion is grounded in the packet: the two issues have identical titles, identical body length, and a soft-similarity signal of 1.0, so treating them as duplicate reports of the same bug is reasonable. Choosing the open issue (#46552) as the canonical tracker is conservative and appropriate." + "feedback": "Grounded by the packet: the two issues have identical titles, identical body length, and a soft_similarity score of 1.0, while #46552 is open and has inbound references. The canonical-tracker choice is conservative and well supported." }, "evaluator_used": true, "retried": false, @@ -21158,7 +21216,7 @@ { "label": "issue", "component_index": 22, - "component_count": 25, + "component_count": 26, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", @@ -21174,12 +21232,12 @@ "soft_pairs": [ "issue:46657|issue:46692" ], - "prepared_review_unit_hash": "a165ce8b069c11d9661be7956f1cc81ce1264a082305265857b8f11e638727ee", + "prepared_review_unit_hash": "c7d353fd89f60220e1203e8b8367a34ac164f4cfef4a6d6bd6e0a3ffa6989d8b", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1569, + "serialized_chars": 1570, "estimated_input_tokens": 393, "estimated_eval_tokens": 1042 }, @@ -21187,7 +21245,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1569, + "serialized_chars": 1570, "estimated_input_tokens": 393, "estimated_eval_tokens": 1042 }, @@ -21196,18 +21254,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These two issues are only loosely related by being training-time failures in object detection models, but they describe different symptoms and likely different root causes: a tensor shape mismatch vs. an out-of-memory failure after several epochs.", - "confidence": 0.91, - "canonical_issue_reason": "Issue #46657 is the cleaner canonical issue because it states a concrete failure mode (\u201cSize of tensors mismatch\u201d) that is more directly actionable and bug-shaped than the broader OOM report.", + "summary": "These issues describe different failure modes in training object detection models: one is a tensor size mismatch error, the other is an out-of-memory condition after several epochs on RT-DETR v2. They are related to training, but not the same underlying bug.", + "confidence": 0.96, + "canonical_issue_reason": "Issue #46657 is the better canonical issue only in the sense that it is a concrete, clearly defined bug report with a specific runtime error; however, it does not cover the OOM-at-epoch problem in #46692.", "canonical_pr_reason": null, - "best_issue_reason": "Issue #46657 is the best representative for this cluster since it captures a specific training-time defect; #46692 appears to be a separate resource/memory problem rather than the same bug.", + "best_issue_reason": "#46657 is the stronger representative for this cluster because it is more specific and actionable, with a distinct error signature. #46692 appears to be a separate training stability/memory issue rather than a duplicate.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:46657", "right": "issue:46692", "accept": false, - "reason": "Do not merge: the reported failures are different (tensor size mismatch vs. OOM after some epochs), suggesting different code paths and likely different fixes." + "reason": "The titles point to different concrete failures: tensor shape mismatch versus out-of-memory during later epochs. They may both occur in object detection training, but they do not look like the same underlying bug and would not be merged as duplicates." } ] }, @@ -21221,7 +21279,7 @@ { "label": "issue", "component_index": 23, - "component_count": 25, + "component_count": 26, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", @@ -21237,7 +21295,7 @@ "soft_pairs": [ "issue:46822|issue:46823" ], - "prepared_review_unit_hash": "9a637433eb0f46e7593fb55570cc4d80cdc8f613f99019a65e9d89522957733c", + "prepared_review_unit_hash": "2f358da6bf9d1810e867b5dfe70628e52b705d4c844e6a30d95af20ad97e0920", "component_budget": { "node_count": 2, "item_count": 2, @@ -21259,18 +21317,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These two issues are about different loading/dispatch failures: one is a ZeRO-3 `zero.Init` partitioning/OOM problem in a composite multimodal submodule, the other is a 4-bit `device_map='auto'` / `max_memory` CPU offload misdispatch causing a `ValueError`. They do not look like the same underlying bug.", - "confidence": 0.97, - "canonical_issue_reason": "Issue 46822 is the more specific, technically grounded report and better captures the concrete ZeRO-3 composite-submodule OOM failure.", + "summary": "The two issues are both model-loading failures, but they describe different code paths and different root causes: one is a ZeRO-3/zero.Init partitioning bug for a composite VL submodule, the other is a 4-bit auto device-mapping bug that wrongly dispatches to CPU despite fitting memory. Not a duplicate pair.", + "confidence": 0.95, + "canonical_issue_reason": "Issue 46822 is the better canonical issue because it has more discussion/activity and is the more established report, while still being a distinct ZeRO-3 loading bug.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 46822 is the better representative because it names a clear code path (`ZeRO-3 zero.Init`) and a distinct failure mode (partitioning omission leading to OOM).", + "best_issue_reason": "Issue 46822 is the strongest representative of its own problem, but it should not be used to cover issue 46823 because the failures involve different mechanisms (ZeRO partitioning vs. 4-bit device_map dispatch).", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:46822", "right": "issue:46823", "accept": false, - "reason": "Different mechanisms and symptoms: ZeRO-3 partitioning/OOM vs 4-bit auto device-mapping with CPU dispatch error. Shared theme is model loading, but not the same bug." + "reason": "Different underlying bugs: 46822 is about ZeRO-3 zero.Init failing to partition a composite submodule and causing OOM; 46823 is about 4-bit + device_map='auto' incorrectly sending modules to CPU and raising ValueError. Same broad theme, but not the same fixable issue." } ] }, @@ -21284,7 +21342,7 @@ { "label": "issue", "component_index": 24, - "component_count": 25, + "component_count": 26, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", @@ -21318,124 +21376,124 @@ "issue:46969|issue:46998", "issue:46985|issue:46998" ], - "prepared_review_unit_hash": "783ff3839f163949d7f43e94822a1760b6d6713158496308efd4e5dd43fe3c9b", + "prepared_review_unit_hash": "6aa11fa20649608b9d4c07c422eefd75b02b10fa1896b05fbf3fa8ebe51e1d5e", "component_budget": { "node_count": 6, "item_count": 6, "soft_pair_count": 15, - "serialized_chars": 8795, - "estimated_input_tokens": 2199, - "estimated_eval_tokens": 4654 + "serialized_chars": 8797, + "estimated_input_tokens": 2200, + "estimated_eval_tokens": 4656 }, "budget": { "node_count": 6, "item_count": 6, "soft_pair_count": 15, - "serialized_chars": 8795, - "estimated_input_tokens": 2199, - "estimated_eval_tokens": 4654 + "serialized_chars": 8797, + "estimated_input_tokens": 2200, + "estimated_eval_tokens": 4656 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are daily date-stamped triage issues using the same template for serge integration failures, but they appear to be separate tracker instances rather than duplicate reports of one bug.", - "confidence": 0.93, - "canonical_issue_reason": "issue:46906 is the earliest item and has the strongest reference/discussion signal, so it is the best representative if one issue must anchor the series.", + "summary": "These are a series of similarly titled daily integration-failure triage issues for different dates, not a single duplicated report. They should remain separate unless later evidence links them to the same underlying failure.", + "confidence": 0.95, + "canonical_issue_reason": "issue:46906 is the earliest item and has the strongest thread activity (highest inbound references and discussion), so it is the best anchor for the cluster if one is needed.", "canonical_pr_reason": null, - "best_issue_reason": "issue:46906 is the most central/day-zero triage issue in the set, with the highest inbound references and discussion activity.", + "best_issue_reason": "issue:46906 is the most central and most visible of the daily triage issues, with the highest inbound reference count and the earliest timestamp.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:46906", "right": "issue:46940", "accept": false, - "reason": "Same triage template and subsystem, but the titles are date-specific daily tracker issues; this looks like a new daily report rather than the same underlying issue." + "reason": "Same triage template, but different date-specific integration reports; no evidence they describe the same failure instance." }, { "left": "issue:46906", "right": "issue:46948", "accept": false, - "reason": "High textual similarity is driven by the shared triage format, but the date-stamped issue indicates a separate day\u2019s integration failure triage." + "reason": "These appear to be separate daily failure triage posts rather than one underlying bug report." }, { "left": "issue:46906", "right": "issue:46969", "accept": false, - "reason": "These are recurring daily triage issues, not the same concrete bug or change; the date progression suggests distinct tracking entries." + "reason": "Different dates and standalone triage threads; not enough evidence of a shared root cause." }, { "left": "issue:46906", "right": "issue:46985", "accept": false, - "reason": "Same repeated template, but this is a later daily triage issue rather than a duplicate of the earlier one." + "reason": "Likely another daily integration-failure triage entry, not a duplicate of the earlier report." }, { "left": "issue:46906", "right": "issue:46998", "accept": false, - "reason": "The issue series is date-based and recurring; similarity alone is not enough to treat these as duplicates." + "reason": "Same naming pattern but distinct date-based triage issue; no concrete overlap showing one bug." }, { "left": "issue:46940", "right": "issue:46948", "accept": false, - "reason": "Both are daily serge triage trackers, but they correspond to different dates and likely different failure snapshots." + "reason": "Adjacent daily triage issues, but no direct evidence they track the same incident." }, { "left": "issue:46940", "right": "issue:46969", "accept": false, - "reason": "Near-identical wording reflects the same issue template, not the same underlying bug report." + "reason": "Different day reports with no explicit shared target or linked failure details." }, { "left": "issue:46940", "right": "issue:46985", "accept": false, - "reason": "Different day, different tracker instance; no evidence they are duplicates of one concrete problem." + "reason": "Separate date-stamped integration triage items; not safe to merge as duplicates." }, { "left": "issue:46940", "right": "issue:46998", "accept": false, - "reason": "These are sequential triage reports, so merging would collapse distinct daily tracking entries." + "reason": "No sign these are the same underlying bug; they look like repeated daily triage posts." }, { "left": "issue:46948", "right": "issue:46969", "accept": false, - "reason": "The shared title format is misleading here; the date indicates separate daily triage issues." + "reason": "Similar title pattern only; likely different daily integration failures." }, { "left": "issue:46948", "right": "issue:46985", "accept": false, - "reason": "Not the same underlying bug, just the same recurring integration-failure triage pattern." + "reason": "Distinct date-based triage issues with no concrete shared root cause evidence." }, { "left": "issue:46948", "right": "issue:46998", "accept": false, - "reason": "The issues are temporally distinct daily reports, so they should not be deduplicated." + "reason": "Not enough overlap beyond the template title to treat as duplicates." }, { "left": "issue:46969", "right": "issue:46985", "accept": false, - "reason": "Same tracker series, but separate dates and likely separate investigation cycles." + "reason": "Two separate daily triage entries; no explicit linkage to one bug or change." }, { "left": "issue:46969", "right": "issue:46998", "accept": false, - "reason": "Recurring triage issue titles are not enough to establish duplicate status." + "reason": "Likely independent failure triage reports for different dates." }, { "left": "issue:46985", "right": "issue:46998", "accept": false, - "reason": "These are adjacent daily triage issues, but still distinct issue instances rather than one bug report." + "reason": "These are consecutive daily triage issues, but the available evidence does not show the same underlying failure." } ] }, @@ -21449,7 +21507,7 @@ { "label": "issue", "component_index": 25, - "component_count": 25, + "component_count": 26, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", @@ -21465,7 +21523,7 @@ "soft_pairs": [ "issue:46983|issue:46984" ], - "prepared_review_unit_hash": "c8b93f23c2f49028ac529bdc3482cc70c6a877742600035ee66683b0bd480033", + "prepared_review_unit_hash": "49508a13e178a95af52c8da3150cb18b76164d0fbd682947fe181e079c0c3267", "component_budget": { "node_count": 2, "item_count": 2, @@ -21487,27 +21545,90 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These two issues are highly similar and share the exact title, so they are a strong duplicate/near-duplicate candidate. However, since the bodies are not shown, I can only treat them as likely related rather than definitively the same request. The open issue is the better canonical tracker for now.", - "confidence": 0.83, - "canonical_issue_reason": "Issue #46984 is open and newer, so it is the better place to track the request, but the duplicate call should remain tentative because the bodies are not visible.", + "summary": "The two issues appear to be duplicates of the same feature request: adding RWKV7 model support. The open issue is the better canonical tracker.", + "confidence": 0.98, + "canonical_issue_reason": "Issue 46984 is open and can serve as the active tracker for the same RWKV7 support request; 46983 is closed and appears to cover the same topic.", "canonical_pr_reason": null, - "best_issue_reason": "Issue #46984 is the strongest single issue candidate because it is open, newer, and has the same title as #46983.", + "best_issue_reason": "46984 is the best issue because it is open, current, and has the exact same title/topic as 46983.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:46983", "right": "issue:46984", - "accept": false, - "reason": "The titles are identical and similarity is high, but the body content is unavailable, so there is not enough evidence to confirm they are the same underlying request." + "accept": true, + "reason": "Same title and same feature request (RWKV7 model support); the only meaningful difference is that one is closed and the other is open, so they look like duplicate issues." } ] }, "evaluator_result": { - "accept": false, - "feedback": "The packet only supports a cautious duplicate/near-duplicate judgment: both issues share the same title and have high similarity, but the bodies are not shown, so claiming they are the same underlying request overstates the evidence. The canonical/open-issue choice is plausible, but the soft-edge verdict and summary should be more conservative." + "accept": true, + "feedback": "Grounded and conservative overall: the two issues have identical titles, no conflicting signals, and the open one is a reasonable canonical tracker. The duplicate/soft-match verdict is supported by the packet and the summary appropriately uses \"appear.\"" }, "evaluator_used": true, - "retried": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 26, + "component_count": 26, + "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-47007-2", + "nodes": [ + "issue:47007", + "issue:47015" + ], + "soft_pairs": [ + "issue:47007|issue:47015" + ], + "prepared_review_unit_hash": "d2dd4c0bac7afba9ce71e74ec8d070b73ae69dac5ac58ef72233d9ce5fe905a1", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1622, + "estimated_input_tokens": 406, + "estimated_eval_tokens": 1068 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1622, + "estimated_input_tokens": 406, + "estimated_eval_tokens": 1068 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Both items are about AI-driven vulnerability discovery, but they read like separate writeups/ideas rather than the same concrete issue. I would not deduplicate them.", + "confidence": 0.57, + "canonical_issue_reason": "Issue 47015 is the broader, more general framing of the topic and is the better cluster representative by title alone.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 47015 is the most central umbrella for the cluster because it directly names the AI-based vulnerability discovery system without narrowing to a specific lesson/article angle.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:47007", + "right": "issue:47015", + "accept": false, + "reason": "They share a broad cybersecurity/LLM theme, but there is no sign they describe the same underlying bug, request, or change. The titles suggest different articles or proposals, so this is not a safe duplicate." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, "accepted_nontrivial_soft_edge": false, "error_kind": null, "error_message": null @@ -21523,7 +21644,7 @@ "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-40546-39", + "cluster_id": "cluster-40546-41", "nodes": [ "pull_request:40546", "pull_request:43451", @@ -21560,7 +21681,8 @@ "pull_request:46977", "pull_request:46986", "pull_request:46995", - "pull_request:47006" + "pull_request:47006", + "pull_request:47014" ], "soft_pairs": [ "pull_request:46895|pull_request:46958", @@ -21580,8 +21702,8 @@ "pull_request:46902|pull_request:46965", "pull_request:46907|pull_request:46926", "pull_request:46970|pull_request:46971", - "pull_request:46927|pull_request:46986", "pull_request:46926|pull_request:46928", + "pull_request:46927|pull_request:46986", "pull_request:46951|pull_request:46986", "pull_request:46933|pull_request:46942", "pull_request:46926|pull_request:46941", @@ -21590,11 +21712,11 @@ "pull_request:46928|pull_request:47006", "pull_request:46941|pull_request:46986", "pull_request:46941|pull_request:47006", - "pull_request:46954|pull_request:46961", - "pull_request:46954|pull_request:46977", "pull_request:46907|pull_request:46908", "pull_request:46908|pull_request:46951", + "pull_request:46954|pull_request:46961", "pull_request:46907|pull_request:46941", + "pull_request:46954|pull_request:46977", "pull_request:46926|pull_request:46927", "pull_request:43451|pull_request:46954", "pull_request:46954|pull_request:46968", @@ -21625,349 +21747,469 @@ "pull_request:46707|pull_request:46968", "pull_request:46891|pull_request:46968", "pull_request:43451|pull_request:46836", - "pull_request:40546|pull_request:43451", "pull_request:46912|pull_request:46977", + "pull_request:40546|pull_request:43451", "pull_request:46196|pull_request:46965", "pull_request:46836|pull_request:46961", - "pull_request:46836|pull_request:46977", "pull_request:46836|pull_request:46965", + "pull_request:46836|pull_request:46977", "pull_request:46196|pull_request:46977", - "pull_request:43451|pull_request:46965", + "pull_request:46266|pull_request:47014", "pull_request:43451|pull_request:46658", - "pull_request:46836|pull_request:46995" + "pull_request:43451|pull_request:46965" ], - "prepared_review_unit_hash": "c1f736b6799ab1ddb388da7d04dbcd80c322641e0407e67fd06b9d578c3d9aa9", + "prepared_review_unit_hash": "58babfc487e294cd72377c8403cd337594f131ebf84e5da09e9fd540a4dd6d80", "component_budget": { - "node_count": 39, - "item_count": 39, - "soft_pair_count": 115, - "serialized_chars": 132026, - "estimated_input_tokens": 33007, - "estimated_eval_tokens": 66270 + "node_count": 41, + "item_count": 41, + "soft_pair_count": 126, + "serialized_chars": 142820, + "estimated_input_tokens": 35705, + "estimated_eval_tokens": 71666 }, "budget": { - "node_count": 36, - "item_count": 36, + "node_count": 37, + "item_count": 37, "soft_pair_count": 72, - "serialized_chars": 100675, - "estimated_input_tokens": 25169, - "estimated_eval_tokens": 50594 + "serialized_chars": 105282, + "estimated_input_tokens": 26321, + "estimated_eval_tokens": 52898 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The cluster is mostly unrelated one-off PRs. The only strong duplicate pair is the CohereASR training-loss double-shift fix between #46895 and #46958. #46903 also touches CohereASR, but it bundles an additional PPFormulaNet fix, so it is too broad to treat as the same mergeable PR.", - "confidence": 0.91, + "summary": "Mostly unrelated PRs are clustered together by topic and test-failure triage. The CohereASR pair does refer to the same underlying loss bug, but it is not a mergeable duplicate pair because the two PRs are alternative fixes rather than complementary changes.", + "confidence": 0.88, "canonical_issue_reason": null, - "canonical_pr_reason": "#46895 is the best canonical PR: it is narrowly scoped, explicitly issue-linked, and directly addresses the CohereASR double-shift loss bug. #46958 is a near-equivalent alternative implementation, while #46903 is broader because it adds a separate PPFormulaNet fix.", + "canonical_pr_reason": "pull_request:46895 is the best canonical CohereASR representative: it directly targets issue 46894, stays scoped to the loss-path fix, and is the cleaner issue-linked patch.", "best_issue_reason": null, - "best_pr_reason": "#46895 is the best global representative because it cleanly isolates one concrete bug in one code path and has the clearest problem statement. The rest of the set is mostly unrelated model additions, infra changes, or test-only fixes.", + "best_pr_reason": "pull_request:46895 is the strongest representative PR for the cluster because it is the most direct, narrowly scoped fix for the CohereASR double-shift bug.", "soft_edge_verdicts": [ { "left": "pull_request:46895", "right": "pull_request:46958", - "accept": true, - "reason": "Same underlying CohereASR training-loss double-shift bug in the same forward/loss path; the diffs are alternative ways to apply the same fix." + "accept": false, + "reason": "Same CohereASR bug, but these are alternative loss fixes rather than complementary edits, so they are not mergeable into one PR." }, { "left": "pull_request:46903", "right": "pull_request:46958", "accept": false, - "reason": "Although both touch CohereASR loss handling, #46903 also bundles an extra PPFormulaNet fix, so it is not clearly the same single-PR bugfix and would not be merged as one duplicate change." + "reason": "Same model family, but 46903 also changes PPFormulaNet and is a different fix shape; not the same concrete change." }, { "left": "pull_request:46950", "right": "pull_request:46951", "accept": false, - "reason": "Different problems: Mamba2 OOM/runtime behavior versus PEFT integration loading behavior." + "reason": "Mamba2 OOM behavior vs PEFT loading option; different code paths and fixes." }, { "left": "pull_request:46927", "right": "pull_request:46950", "accept": false, - "reason": "Different model families and failure modes: GLM46V input handling versus Mamba2 OOM/generation behavior." + "reason": "GLM46V dtype casting vs Mamba2 OOM; unrelated bugs." }, { "left": "pull_request:46950", "right": "pull_request:47006", "accept": false, - "reason": "Mamba2 OOM/test fix and Gemma3 generation-cache logic are unrelated code paths." + "reason": "Mamba2 OOM vs Gemma3 generation/cache behavior; unrelated." }, { "left": "pull_request:46927", "right": "pull_request:46951", "accept": false, - "reason": "GLM46V input casting/forward handling and PEFT loading are unrelated." + "reason": "GLM46V dtype fix vs PEFT integration loading; unrelated." }, { "left": "pull_request:46951", "right": "pull_request:46971", "accept": false, - "reason": "PEFT loading logic and a Mamba2 test-only dtype change are not the same bug." + "reason": "PEFT load option vs Mamba2 test dtype argument; different fixes." }, { "left": "pull_request:46949", "right": "pull_request:46971", "accept": false, - "reason": "Whisper/audio cache handling and Mamba2 test dtype changes are different issues." + "reason": "Whisper/audioflamingo cache handling vs Mamba2 OOM test setup; unrelated." }, { "left": "pull_request:46927", "right": "pull_request:47006", "accept": false, - "reason": "Different model families and fixes; no shared concrete bug." + "reason": "GLM46V dtype fix vs Gemma3 generation/cache change; unrelated." }, { "left": "pull_request:46951", "right": "pull_request:47006", "accept": false, - "reason": "PEFT integration loading and Gemma3 generation-config behavior are unrelated." + "reason": "PEFT loading vs Gemma3 generation/cache; unrelated." }, { "left": "pull_request:46971", "right": "pull_request:47006", "accept": false, - "reason": "A Mamba2 test dtype adjustment is not the same as Gemma3 cache-generation logic." + "reason": "Mamba2 test cleanup vs Gemma3 generation/cache; unrelated." }, { "left": "pull_request:46941", "right": "pull_request:46970", "accept": false, - "reason": "Different models and symptoms: DAC tensor mismatch versus MusicGen melody generation mismatch." + "reason": "DAC output-mismatch fix vs musicgen_melody output mismatch; different models/code paths." }, { "left": "pull_request:46908", "right": "pull_request:47006", "accept": false, - "reason": "DeepSeekV3.2 test offload handling is unrelated to Gemma3 generation cache logic." + "reason": "DeepSeekV3.2 offload test fix vs Gemma3 generation/cache; unrelated." }, { "left": "pull_request:46908", "right": "pull_request:46928", "accept": false, - "reason": "DeepSeekV3.2 load/offload test fix and BLIP-2 tokenizer registration are different changes." + "reason": "DeepSeekV3.2 offload test fix vs BLIP2 tokenizer mapping; unrelated." }, { "left": "pull_request:46902", "right": "pull_request:46965", "accept": false, - "reason": "Offloading/save-converted-weights bugfix and DeepSeek V4 use_cache support are different code paths." + "reason": "Offloading/save-converted-weights bugfix vs DeepSeekV4 use_cache support; different runtime paths." }, { "left": "pull_request:46907", "right": "pull_request:46926", "accept": false, - "reason": "EdgeTam processor kwarg cleanup and MiniMax-M3-VL load test fix are unrelated." + "reason": "EdgeTam import_or_config fix vs MiniMax-M3-VL load_error test fix; different bugs." }, { "left": "pull_request:46970", "right": "pull_request:46971", "accept": false, - "reason": "MusicGen generation logic and Mamba2 test dtype changes are unrelated." + "reason": "musicgen_melody model bug vs Mamba2 test-only OOM tweak; different changes." + }, + { + "left": "pull_request:46926", + "right": "pull_request:46928", + "accept": false, + "reason": "MiniMax load error vs BLIP2 tokenizer mapping; unrelated." }, { "left": "pull_request:46927", "right": "pull_request:46986", "accept": false, - "reason": "GLM46V input handling and GLM-OCR output mismatch are separate fixes." + "reason": "GLM46V input dtype fix vs GLM OCR output mismatch; different models and bugs." }, { "left": "pull_request:46951", "right": "pull_request:46986", "accept": false, - "reason": "PEFT loading and GLM-OCR model behavior are unrelated." + "reason": "PEFT loading fix vs GLM OCR model fix; unrelated." }, { "left": "pull_request:46933", "right": "pull_request:46942", "accept": false, - "reason": "Both are dtype-casting fixes, but for different models and concrete code paths." + "reason": "Both are dtype-related, but Gemma4 quantized embedder casting and TimesFM2.5 block casting are separate concrete fixes." }, { "left": "pull_request:46926", "right": "pull_request:46941", "accept": false, - "reason": "MiniMax-M3-VL test load_error fix and DAC output mismatch are unrelated." + "reason": "MiniMax load error vs DAC output mismatch; unrelated." }, { "left": "pull_request:46928", "right": "pull_request:46941", "accept": false, - "reason": "BLIP-2 tokenizer auto-mapping and DAC model output mismatch are not the same bug." + "reason": "BLIP2 tokenizer handling vs DAC output mismatch; unrelated." }, { "left": "pull_request:46907", "right": "pull_request:46928", "accept": false, - "reason": "EdgeTam kwarg handling and BLIP-2 tokenizer registration are unrelated." + "reason": "EdgeTam import_or_config vs BLIP2 tokenizer mapping; unrelated." }, { "left": "pull_request:46928", "right": "pull_request:47006", "accept": false, - "reason": "BLIP-2 tokenizer auto-mapping fix is unrelated to Gemma3 generation behavior." + "reason": "BLIP2 tokenizer mapping vs Gemma3 generation/cache; unrelated." }, { "left": "pull_request:46941", "right": "pull_request:46986", "accept": false, - "reason": "DAC model mismatch and GLM-OCR mismatch are different models and failure modes." + "reason": "DAC output mismatch vs GLM OCR output mismatch; different models and code paths." }, { "left": "pull_request:46941", "right": "pull_request:47006", "accept": false, - "reason": "DAC output mismatch and Gemma3 cache handling are unrelated." + "reason": "DAC fixes vs Gemma3 generation/cache; unrelated." + }, + { + "left": "pull_request:46907", + "right": "pull_request:46908", + "accept": false, + "reason": "EdgeTam import_or_config vs DeepSeekV3.2 offload test; different failures." + }, + { + "left": "pull_request:46908", + "right": "pull_request:46951", + "accept": false, + "reason": "DeepSeekV3.2 offload test fix vs PEFT integration loading; unrelated." }, { "left": "pull_request:46954", "right": "pull_request:46961", "accept": false, - "reason": "GLM-ASR pytest compatibility and Serge branch workflow normalization are unrelated." + "reason": "GLM-ASR pytest-v9 compatibility fix vs Serge branch-normalization workflow; unrelated." + }, + { + "left": "pull_request:46907", + "right": "pull_request:46941", + "accept": false, + "reason": "EdgeTam import_or_config vs DAC output mismatch; unrelated." }, { "left": "pull_request:46954", "right": "pull_request:46977", "accept": false, - "reason": "GLM-ASR pytest compatibility and Mllama arg-name handling are unrelated." + "reason": "GLM-ASR pytest compatibility vs Mllama arg-name deprecation; different changes." }, { "left": "pull_request:43451", "right": "pull_request:46954", "accept": false, - "reason": "Molmo2 model addition and GLM-ASR pytest compatibility fix are different kinds of changes." + "reason": "Molmo2 model addition vs GLM-ASR pytest compatibility; unrelated." }, { "left": "pull_request:46954", "right": "pull_request:46968", "accept": false, - "reason": "GLM-ASR pytest compatibility and docs autolink cleanup are unrelated." + "reason": "GLM-ASR pytest compatibility vs docs autolink cleanup; unrelated." }, { "left": "pull_request:46954", "right": "pull_request:46958", "accept": false, - "reason": "GLM-ASR pytest compatibility is unrelated to CohereASR training-loss correction." - }, - { - "left": "pull_request:46961", - "right": "pull_request:46977", - "accept": false, - "reason": "Workflow normalization and Mllama arg-name handling are unrelated." + "reason": "GLM-ASR pytest compatibility vs CohereASR loss fix; unrelated." }, { - "left": "pull_request:46961", - "right": "pull_request:46974", + "left": "pull_request:46958", + "right": "pull_request:46961", "accept": false, - "reason": "Workflow normalization and MusicGen generate typo fix are unrelated." + "reason": "CohereASR loss bugfix vs workflow automation; unrelated." }, { "left": "pull_request:46958", "right": "pull_request:46977", "accept": false, - "reason": "CohereASR loss fix and Mllama arg-name deprecation are different code paths." + "reason": "CohereASR loss fix vs Mllama arg-name deprecation; different changes." }, { "left": "pull_request:46961", - "right": "pull_request:46968", + "right": "pull_request:46974", "accept": false, - "reason": "Serge branch workflow wiring and docs autolink cleanup are unrelated." + "reason": "Workflow automation vs MusicGen generate typo fix; unrelated." }, { "left": "pull_request:46958", "right": "pull_request:46974", "accept": false, - "reason": "CohereASR training-loss fix and MusicGen generation typo fix are unrelated." + "reason": "CohereASR loss fix vs MusicGen generate typo; unrelated." }, { "left": "pull_request:46974", "right": "pull_request:46977", "accept": false, - "reason": "MusicGen generation typo and Mllama arg-name deprecation are unrelated." + "reason": "MusicGen generate typo fix vs Mllama arg-name deprecation; unrelated." }, { "left": "pull_request:43451", "right": "pull_request:46920", "accept": false, - "reason": "Molmo2 model addition and Voxtral realtime API adaptation are different features." + "reason": "Molmo2 model addition vs Voxtral Realtime API adaptation; unrelated." }, { "left": "pull_request:46920", "right": "pull_request:46968", "accept": false, - "reason": "Voxtral realtime API support and docs autolink cleanup are unrelated." + "reason": "Voxtral Realtime API adaptation vs docs autolink cleanup; unrelated." }, { "left": "pull_request:43451", "right": "pull_request:46958", "accept": false, - "reason": "Molmo2 model addition and CohereASR loss bugfix are unrelated." + "reason": "Molmo2 model addition vs CohereASR loss fix; unrelated." + }, + { + "left": "pull_request:46971", + "right": "pull_request:46974", + "accept": false, + "reason": "Mamba2 test OOM fix vs MusicGen generate typo; unrelated." + }, + { + "left": "pull_request:46968", + "right": "pull_request:46974", + "accept": false, + "reason": "Docs autolink cleanup vs MusicGen generate typo fix; unrelated." + }, + { + "left": "pull_request:46920", + "right": "pull_request:46961", + "accept": false, + "reason": "Voxtral Realtime API adaptation vs Serge workflow automation; unrelated." + }, + { + "left": "pull_request:46961", + "right": "pull_request:46968", + "accept": false, + "reason": "Branch-normalization workflow vs docs autolink cleanup; unrelated." + }, + { + "left": "pull_request:46958", + "right": "pull_request:46968", + "accept": false, + "reason": "CohereASR loss fix vs docs autolink cleanup; unrelated." + }, + { + "left": "pull_request:43451", + "right": "pull_request:46707", + "accept": false, + "reason": "Molmo2 model addition vs native FSDP2 support/migration; unrelated." + }, + { + "left": "pull_request:46266", + "right": "pull_request:46707", + "accept": false, + "reason": "DINOv2 modernization/conversion work vs native FSDP2 support; unrelated." + }, + { + "left": "pull_request:46196", + "right": "pull_request:46995", + "accept": false, + "reason": "Wav2Vec2 SDPA dispatch vs flash-linear-attention MLU availability; different acceleration backends." }, { "left": "pull_request:46707", "right": "pull_request:46961", "accept": false, - "reason": "Native FSDP2 integration and Serge branch normalization are unrelated infrastructure changes." + "reason": "FSDP2 support/migration vs Serge workflow normalization; unrelated." + }, + { + "left": "pull_request:46891", + "right": "pull_request:46961", + "accept": false, + "reason": "PEFT integration CI vs Serge workflow normalization; different CI changes." + }, + { + "left": "pull_request:43451", + "right": "pull_request:46891", + "accept": false, + "reason": "Molmo2 model addition vs PEFT integration CI; unrelated." + }, + { + "left": "pull_request:46707", + "right": "pull_request:46977", + "accept": false, + "reason": "FSDP2 support/migration vs Mllama arg-name modernization; unrelated." + }, + { + "left": "pull_request:46891", + "right": "pull_request:46977", + "accept": false, + "reason": "PEFT CI change vs Mllama arg-name cleanup; unrelated." + }, + { + "left": "pull_request:46707", + "right": "pull_request:46968", + "accept": false, + "reason": "FSDP2 support/migration vs docs autolink cleanup; unrelated." + }, + { + "left": "pull_request:46891", + "right": "pull_request:46968", + "accept": false, + "reason": "PEFT CI change vs docs autolink cleanup; unrelated." + }, + { + "left": "pull_request:43451", + "right": "pull_request:46836", + "accept": false, + "reason": "Molmo2 model addition vs Unlimited OCR model addition; separate feature PRs." + }, + { + "left": "pull_request:46912", + "right": "pull_request:46977", + "accept": false, + "reason": "RoPE test/implementation updates vs Mllama arg-name deprecation; unrelated." + }, + { + "left": "pull_request:40546", + "right": "pull_request:43451", + "accept": false, + "reason": "VibeVoice and Molmo2 are separate model additions." }, { "left": "pull_request:46196", "right": "pull_request:46965", "accept": false, - "reason": "wav2vec2 SDPA dispatch support and DeepSeek V4 cache support are different model/runtime changes." + "reason": "Wav2Vec2 SDPA dispatch vs DeepSeekV4 use_cache support; unrelated." }, { "left": "pull_request:46836", "right": "pull_request:46961", "accept": false, - "reason": "Unlimited OCR model addition and workflow normalization are unrelated." + "reason": "Unlimited OCR model addition vs Serge workflow normalization; unrelated." }, { "left": "pull_request:46836", - "right": "pull_request:46977", + "right": "pull_request:46965", "accept": false, - "reason": "Unlimited OCR model addition and Mllama arg-name cleanup are unrelated." + "reason": "Unlimited OCR model addition vs DeepSeekV4 use_cache support; unrelated." }, { "left": "pull_request:46836", - "right": "pull_request:46965", + "right": "pull_request:46977", "accept": false, - "reason": "Unlimited OCR model addition and DeepSeek V4 cache support are unrelated." + "reason": "Unlimited OCR model addition vs Mllama arg-name cleanup; unrelated." }, { "left": "pull_request:46196", "right": "pull_request:46977", "accept": false, - "reason": "wav2vec2 SDPA dispatch support and Mllama arg-name cleanup are unrelated." + "reason": "Wav2Vec2 SDPA dispatch vs Mllama arg-name cleanup; unrelated." }, { - "left": "pull_request:43451", - "right": "pull_request:46965", + "left": "pull_request:46266", + "right": "pull_request:47014", "accept": false, - "reason": "Molmo2 model addition and DeepSeek V4 cache support are unrelated." + "reason": "DINO modernization/conversion work vs T5 SDPA enablement; unrelated." }, { "left": "pull_request:43451", "right": "pull_request:46658", "accept": false, - "reason": "Molmo2 model addition and Step-3.7-Flash implementation are unrelated." + "reason": "Molmo2 model addition vs Step 3.7 Flash implementation; separate features." }, { - "left": "pull_request:46836", - "right": "pull_request:46995", + "left": "pull_request:43451", + "right": "pull_request:46965", "accept": false, - "reason": "Unlimited OCR model addition and MLU availability check are unrelated." + "reason": "Molmo2 model addition vs DeepSeekV4 use_cache support; unrelated." } ] }, "evaluator_result": { "accept": false, - "feedback": "The CohereASR cluster is plausible, but one soft-edge accept is too broad: 46903 bundles an additional PPFormulaNet fix, so it is not clearly mergeable with 46958 as a single PR for the same bugfix. That makes the duplicate verdict less conservative than the packet supports. The rest of the summary is mostly grounded, but this overstates the evidence." + "feedback": "The cluster summary is mostly grounded, but one soft-edge verdict is too strong: pull_request:46895 and pull_request:46958 are not clearly mergeable into a single PR, since they implement alternative fixes for the same CohereASR loss bug rather than complementary changes. The packet supports that they address the same issue, but not that they should be accepted as a mergeable pair. The confidence/\u2018only clear duplicate pair\u2019 framing is therefore overstated." }, "evaluator_used": true, "retried": true, - "accepted_nontrivial_soft_edge": true, + "accepted_nontrivial_soft_edge": false, "error_kind": null, "error_message": null }, @@ -21982,10 +22224,11 @@ "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-40546-39", + "cluster_id": "cluster-40546-41", "nodes": [ "pull_request:43451", "pull_request:44517", + "pull_request:45694", "pull_request:46196", "pull_request:46658", "pull_request:46707", @@ -22008,28 +22251,38 @@ "pull_request:46971", "pull_request:46977", "pull_request:46986", + "pull_request:46995", "pull_request:47006", - "pull_request:47008" + "pull_request:47008", + "pull_request:47014" ], "soft_pairs": [ + "pull_request:46836|pull_request:46995", "pull_request:46658|pull_request:46961", "pull_request:46658|pull_request:46977", - "pull_request:46658|pull_request:46965", "pull_request:44517|pull_request:46836", "pull_request:43451|pull_request:46912", + "pull_request:45694|pull_request:46707", + "pull_request:44517|pull_request:47014", "pull_request:44517|pull_request:46957", + "pull_request:45694|pull_request:46891", "pull_request:43451|pull_request:46196", + "pull_request:43451|pull_request:47014", "pull_request:43451|pull_request:46957", "pull_request:46196|pull_request:46961", + "pull_request:46961|pull_request:47014", "pull_request:46957|pull_request:46961", "pull_request:46912|pull_request:46961", - "pull_request:44517|pull_request:46891", "pull_request:46707|pull_request:47008", - "pull_request:46957|pull_request:46977", + "pull_request:44517|pull_request:46891", "pull_request:46912|pull_request:47008", - "pull_request:46196|pull_request:46957", + "pull_request:46977|pull_request:47014", + "pull_request:46957|pull_request:46977", + "pull_request:47008|pull_request:47014", + "pull_request:45694|pull_request:47014", "pull_request:46957|pull_request:47008", - "pull_request:46836|pull_request:47008", + "pull_request:45694|pull_request:46957", + "pull_request:45694|pull_request:46836", "pull_request:46933|pull_request:47008", "pull_request:46950|pull_request:46971", "pull_request:46949|pull_request:46950", @@ -22050,412 +22303,337 @@ "pull_request:43451|pull_request:46961", "pull_request:43451|pull_request:46977", "pull_request:46957|pull_request:46965", + "pull_request:46196|pull_request:47014", "pull_request:46658|pull_request:46836", - "pull_request:46891|pull_request:46957", - "pull_request:46836|pull_request:46957", - "pull_request:46912|pull_request:46957", - "pull_request:46836|pull_request:46912" + "pull_request:46891|pull_request:47014", + "pull_request:46836|pull_request:47014", + "pull_request:46658|pull_request:47014", + "pull_request:46957|pull_request:47014", + "pull_request:46912|pull_request:47014", + "pull_request:46912|pull_request:46957" ], - "prepared_review_unit_hash": "58cf6aa06c3e37b3c175bce93d7a02cd70469b8e4edb35355224ec58d509ab65", + "prepared_review_unit_hash": "835858bae33c6e8900ca86540daafbe022ef924145a0879d499146e70dc05f32", "component_budget": { - "node_count": 39, - "item_count": 39, - "soft_pair_count": 115, - "serialized_chars": 132026, - "estimated_input_tokens": 33007, - "estimated_eval_tokens": 66270 + "node_count": 41, + "item_count": 41, + "soft_pair_count": 126, + "serialized_chars": 142820, + "estimated_input_tokens": 35705, + "estimated_eval_tokens": 71666 }, "budget": { - "node_count": 26, - "item_count": 26, - "soft_pair_count": 43, - "serialized_chars": 71679, - "estimated_input_tokens": 17920, - "estimated_eval_tokens": 36096 + "node_count": 29, + "item_count": 29, + "soft_pair_count": 54, + "serialized_chars": 82287, + "estimated_input_tokens": 20572, + "estimated_eval_tokens": 41400 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly heterogeneous: model-addition PRs, docs/CI tweaks, and several unrelated model bug fixes. The only clear duplicate-style pair is the Gemma4 dtype-casting fix (46933 vs 47008); the rest should stay separate.", - "confidence": 0.91, + "summary": "Most items are unrelated PRs spanning different models, docs, CI, or infrastructure. The only clear duplicate-like pair is the Gemma4 quantized embedder dtype-casting fix (46933 and 47008).", + "confidence": 0.97, "canonical_issue_reason": null, - "canonical_pr_reason": "No single canonical PR for the whole cluster; if anchoring the only true duplicate subcluster, 46933 is the better representative because it has the explicit issue target and 47008 is the near-identical follow-up.", + "canonical_pr_reason": "46933 is the better canonical PR: it directly fixes the Gemma4 dtype-casting bug, has the explicit issue target 46932, and is the more established root-fix candidate for the duplicate pair.", "best_issue_reason": null, - "best_pr_reason": "46933 is the best PR to keep as the representative for the one genuine duplicate pair in this cluster. The other PRs are unrelated enough that they should not be merged into the same canonical change.", + "best_pr_reason": "46933 is the strongest overall PR candidate in this set because it is the only true duplicate anchor and the cleanest concrete code-path fix.", "soft_edge_verdicts": [ { - "left": "pull_request:46658", - "right": "pull_request:46961", + "left": "pull_request:46836", + "right": "pull_request:46995", "accept": false, - "reason": "Unrelated: Step-3.7-Flash model scaffolding vs a workflow to normalize Serge branches." + "reason": "Different changes: Unlimited OCR model addition versus MLU support for flash linear attention. Only a broad transformer-model overlap." }, { "left": "pull_request:46658", - "right": "pull_request:46977", + "right": "pull_request:46961", "accept": false, - "reason": "Unrelated model additions/fixups: Step-3.7-Flash vs Mllama arg-name cleanup." + "reason": "Step-3.7 Flash model build/conversion work is unrelated to a Serge branch-normalization workflow." }, { "left": "pull_request:46658", - "right": "pull_request:46965", + "right": "pull_request:46977", "accept": false, - "reason": "Different features: Step-3.7-Flash implementation vs a DeepSeek V4 cache behavior fix." + "reason": "New Step-3.7 Flash model tooling and an Mllama argument cleanup are unrelated." }, { "left": "pull_request:44517", "right": "pull_request:46836", "accept": false, - "reason": "Both add models, but for different models (Qwen3-TTS vs Unlimited OCR), so not duplicates." + "reason": "Two separate model additions (Qwen3-TTS vs Unlimited OCR); shared auto-registration plumbing is not enough." }, { "left": "pull_request:43451", "right": "pull_request:46912", "accept": false, - "reason": "Molmo2 model addition is unrelated to RoPE test coverage changes." + "reason": "Molmo2 model addition versus RoPE test/implementation updates; not the same bug or change." }, { - "left": "pull_request:44517", - "right": "pull_request:46957", + "left": "pull_request:45694", + "right": "pull_request:46707", "accept": false, - "reason": "Qwen3-TTS model addition is unrelated to finegrained FP8 generation/kernel changes." + "reason": "TrainingArguments batch-size fix and native FSDP2 migration address different areas." }, { - "left": "pull_request:43451", - "right": "pull_request:46196", + "left": "pull_request:44517", + "right": "pull_request:47014", "accept": false, - "reason": "Molmo2 addition and wav2vec2 SDPA dispatch support are different changes." + "reason": "Qwen3-TTS addition and T5 SDPA enablement are different model families and fixes." }, { - "left": "pull_request:43451", + "left": "pull_request:44517", "right": "pull_request:46957", "accept": false, - "reason": "Molmo2 model addition is unrelated to finegrained FP8 kernel work." - }, - { - "left": "pull_request:46196", - "right": "pull_request:46961", - "accept": false, - "reason": "wav2vec2 SDPA dispatch support has nothing to do with Serge workflow normalization." - }, - { - "left": "pull_request:46957", - "right": "pull_request:46961", - "accept": false, - "reason": "Finegrained FP8 kernel changes are unrelated to branch-normalization workflow plumbing." - }, - { - "left": "pull_request:46912", - "right": "pull_request:46961", - "accept": false, - "reason": "RoPE test improvements are unrelated to the Serge normalization workflow." + "reason": "Model addition versus finegrained FP8 decode optimization; no concrete overlap." }, { - "left": "pull_request:44517", + "left": "pull_request:45694", "right": "pull_request:46891", "accept": false, - "reason": "Qwen3-TTS model addition is unrelated to PEFT CI workflow testing." + "reason": "TrainingArguments batch-size behavior and PEFT CI workflow changes are unrelated." }, { - "left": "pull_request:46707", - "right": "pull_request:47008", - "accept": false, - "reason": "Native FSDP2 migration and Gemma4 dtype-casting fixes are unrelated." - }, - { - "left": "pull_request:46957", - "right": "pull_request:46977", + "left": "pull_request:43451", + "right": "pull_request:46196", "accept": false, - "reason": "Finegrained FP8 kernel work and Mllama arg-name cleanup are separate changes." + "reason": "Molmo2 addition and wav2vec2 SDPA dispatch are unrelated." }, { - "left": "pull_request:46912", - "right": "pull_request:47008", + "left": "pull_request:43451", + "right": "pull_request:47014", "accept": false, - "reason": "RoPE test changes are unrelated to Gemma4 quantized-embedder dtype casting." + "reason": "Molmo2 addition versus T5 SDPA enablement; different models and code paths." }, { - "left": "pull_request:46196", + "left": "pull_request:43451", "right": "pull_request:46957", "accept": false, - "reason": "wav2vec2 SDPA dispatch support is unrelated to finegrained FP8 decoding changes." - }, - { - "left": "pull_request:46957", - "right": "pull_request:47008", - "accept": false, - "reason": "Different problem space: finegrained FP8 kernel selection vs Gemma4 dtype casting." - }, - { - "left": "pull_request:46836", - "right": "pull_request:47008", - "accept": false, - "reason": "Unlimited OCR model addition is unrelated to Gemma4 dtype-casting fixes." - }, - { - "left": "pull_request:46933", - "right": "pull_request:47008", - "accept": true, - "reason": "Same concrete Gemma4 bug: both patch quantized vision/audio embedders to avoid invalid dtype casting, and 47008 is a near-identical follow-up to 46933." - }, - { - "left": "pull_request:46950", - "right": "pull_request:46971", - "accept": false, - "reason": "Related model area, but not the same fix: one changes Mamba2 runtime behavior, the other only adjusts test loading dtype arguments." + "reason": "Model addition versus generation/kernel optimization; not the same underlying change." }, { - "left": "pull_request:46949", - "right": "pull_request:46950", + "left": "pull_request:46196", + "right": "pull_request:46961", "accept": false, - "reason": "Whisper/audioflamingo cache bug fixes vs Mamba2 OOM behavior are unrelated." + "reason": "wav2vec2 SDPA dispatch and Serge workflow normalization are unrelated." }, { - "left": "pull_request:46949", - "right": "pull_request:46951", + "left": "pull_request:46961", + "right": "pull_request:47014", "accept": false, - "reason": "Whisper/audioflamingo test fixes are unrelated to PEFT integration test loading." + "reason": "Workflow normalization and T5 SDPA enablement have nothing in common beyond repository scope." }, { "left": "pull_request:46957", - "right": "pull_request:46965", + "right": "pull_request:46961", "accept": false, - "reason": "Both touch DeepSeek V4, but one is FP8 generation optimization and the other is use_cache=False support; different code paths." + "reason": "Finegrained FP8 decode optimization versus workflow normalization; different code and intent." }, { - "left": "pull_request:46658", - "right": "pull_request:46836", + "left": "pull_request:46912", + "right": "pull_request:46961", "accept": false, - "reason": "Step-3.7-Flash and Unlimited OCR are separate model integrations." + "reason": "RoPE testing/behavior changes and workflow normalization are unrelated." }, { - "left": "pull_request:46891", - "right": "pull_request:46957", + "left": "pull_request:46707", + "right": "pull_request:47008", "accept": false, - "reason": "PEFT CI workflow edits are unrelated to finegrained FP8 model/runtime changes." + "reason": "FSDP2 migration and Gemma4 dtype-casting fix solve different problems." }, { - "left": "pull_request:46836", - "right": "pull_request:46957", + "left": "pull_request:44517", + "right": "pull_request:46891", "accept": false, - "reason": "Unlimited OCR model addition is unrelated to finegrained FP8 kernel changes." + "reason": "Qwen3-TTS model addition and PEFT CI workflow changes are unrelated." }, { "left": "pull_request:46912", - "right": "pull_request:46957", - "accept": false, - "reason": "RoPE test coverage changes are unrelated to finegrained FP8 runtime work." - }, - { - "left": "pull_request:46961", - "right": "pull_request:46977", - "accept": false, - "reason": "Workflow normalization and Mllama arg-name cleanup are unrelated." - }, - { - "left": "pull_request:43451", - "right": "pull_request:46968", + "right": "pull_request:47008", "accept": false, - "reason": "Molmo2 model addition is unrelated to docs autolink fixes." + "reason": "RoPE test sweep versus Gemma4 dtype-casting fix; no shared concrete bug." }, { - "left": "pull_request:43451", - "right": "pull_request:46961", + "left": "pull_request:46977", + "right": "pull_request:47014", "accept": false, - "reason": "Molmo2 model addition is unrelated to Serge branch normalization." + "reason": "Mllama arg-name cleanup and T5 SDPA enablement are unrelated." }, { - "left": "pull_request:43451", + "left": "pull_request:46957", "right": "pull_request:46977", "accept": false, - "reason": "Molmo2 model addition is unrelated to Mllama arg-name cleanup." + "reason": "Finegrained FP8 decode changes versus Mllama argument cleanup; different bugs." }, { - "left": "pull_request:46957", - "right": "pull_request:46965", + "left": "pull_request:47008", + "right": "pull_request:47014", "accept": false, - "reason": "FP8 kernel/generation tuning and DeepSeek V4 cache semantics are different fixes." + "reason": "Gemma4 dtype-casting fix and T5 SDPA enablement are unrelated." }, { - "left": "pull_request:46836", - "right": "pull_request:46957", + "left": "pull_request:45694", + "right": "pull_request:47014", "accept": false, - "reason": "Unlimited OCR model addition is unrelated to finegrained FP8 kernel work." + "reason": "TrainingArguments batch-size fix and T5 SDPA enablement are unrelated." }, { - "left": "pull_request:46912", + "left": "pull_request:46957", "right": "pull_request:47008", "accept": false, - "reason": "RoPE test changes and Gemma4 dtype casting are different problems." - }, - { - "left": "pull_request:46907", - "right": "pull_request:46927", - "accept": false, - "reason": "EdgeTam import_or_config fix and GLM46V other fix are separate model-specific bugs." - }, - { - "left": "pull_request:46927", - "right": "pull_request:46949", - "accept": false, - "reason": "GLM46V output mismatch fixes are unrelated to Whisper/audioflamingo test issues." + "reason": "Finegrained FP8 decode optimization versus Gemma4 dtype-casting fix; different code paths." }, { - "left": "pull_request:46907", - "right": "pull_request:46949", - "accept": false, - "reason": "EdgeTam import/config test failures and Whisper/audioflamingo failures are different bugs." - }, - { - "left": "pull_request:46907", - "right": "pull_request:46970", + "left": "pull_request:45694", + "right": "pull_request:46957", "accept": false, - "reason": "EdgeTam import/config failures are unrelated to MusicGen Melody output mismatch fixes." + "reason": "TrainingArguments batch-size fix and finegrained FP8 decode optimization are unrelated." }, { - "left": "pull_request:46907", - "right": "pull_request:46986", + "left": "pull_request:45694", + "right": "pull_request:46836", "accept": false, - "reason": "EdgeTam import/config failures are unrelated to GLM OCR output mismatch fixes." + "reason": "TrainingArguments batch-size fix and Unlimited OCR model addition are unrelated." }, { - "left": "pull_request:46907", - "right": "pull_request:47006", - "accept": false, - "reason": "EdgeTam test failures and Gemma3 output mismatch fixes are separate model bugs." + "left": "pull_request:46933", + "right": "pull_request:47008", + "accept": true, + "reason": "Same Gemma4 quantized embedder dtype-casting bug, same files, and shared issue target 46932; these could be merged into one PR." }, { - "left": "pull_request:46907", + "left": "pull_request:46950", "right": "pull_request:46971", "accept": false, - "reason": "EdgeTam import/config failures are unrelated to Mamba2 OOM-related test adjustments." + "reason": "Both concern Mamba2 OOM, but one changes model code while the other only adjusts tests; not the same concrete fix." }, { - "left": "pull_request:46957", - "right": "pull_request:46970", + "left": "pull_request:46949", + "right": "pull_request:46950", "accept": false, - "reason": "Finegrained FP8 kernel changes are unrelated to MusicGen Melody generation mismatch fixes." + "reason": "Whisper/AudioFlamingo cache behavior and Mamba2 OOM fixes are unrelated model bugs." }, { - "left": "pull_request:46957", - "right": "pull_request:46986", + "left": "pull_request:46949", + "right": "pull_request:46951", "accept": false, - "reason": "Finegrained FP8 kernel changes are unrelated to GLM OCR output mismatch fixes." + "reason": "Whisper/AudioFlamingo fix versus PEFT loading tweak; unrelated." }, { - "left": "pull_request:46907", - "right": "pull_request:47006", + "left": "pull_request:46927", + "right": "pull_request:46949", "accept": false, - "reason": "EdgeTam failures and Gemma3 output mismatch fixes are unrelated." + "reason": "GLM46V input-id casting fix and Whisper/AudioFlamingo cache fix are different code paths." }, { "left": "pull_request:46907", - "right": "pull_request:46986", + "right": "pull_request:46927", "accept": false, - "reason": "EdgeTam import/config failures and GLM OCR output mismatch fixes are unrelated." + "reason": "Native FSDP2 migration and GLM46V input-id casting fix are unrelated." }, { "left": "pull_request:46907", - "right": "pull_request:46970", + "right": "pull_request:46949", "accept": false, - "reason": "EdgeTam import/config failures and MusicGen Melody output mismatch fixes are unrelated." + "reason": "FSDP2 migration versus Whisper/AudioFlamingo cache fix; different problems." }, { - "left": "pull_request:46961", - "right": "pull_request:46977", + "left": "pull_request:46907", + "right": "pull_request:47006", "accept": false, - "reason": "Serge workflow normalization and Mllama arg-name cleanup are unrelated." + "reason": "FSDP2 migration and Gemma3 generation-config fix do not address the same underlying change." }, { - "left": "pull_request:46957", - "right": "pull_request:47006", + "left": "pull_request:46907", + "right": "pull_request:46970", "accept": false, - "reason": "Finegrained FP8 kernel changes are unrelated to Gemma3 cache-implementation behavior." + "reason": "FSDP2 migration and MusicGen Melody test fixes are unrelated." }, { - "left": "pull_request:46986", - "right": "pull_request:47006", + "left": "pull_request:46907", + "right": "pull_request:46986", "accept": false, - "reason": "GLM OCR output mismatch fixes and Gemma3 cache-implementation behavior changes are unrelated." + "reason": "FSDP2 migration and GLM-OCR output mismatch fix are unrelated." }, { - "left": "pull_request:46907", - "right": "pull_request:46949", + "left": "pull_request:46961", + "right": "pull_request:46977", "accept": false, - "reason": "EdgeTam import/config failures and Whisper/audioflamingo failures are unrelated." + "reason": "Serge workflow normalization and Mllama arg-name cleanup are unrelated." }, { - "left": "pull_request:46927", - "right": "pull_request:46970", + "left": "pull_request:43451", + "right": "pull_request:46968", "accept": false, - "reason": "GLM46V fixes and MusicGen Melody fixes are separate model bugs." + "reason": "Molmo2 model addition and docs autolink cleanup are unrelated." }, { - "left": "pull_request:46912", - "right": "pull_request:46957", + "left": "pull_request:43451", + "right": "pull_request:46961", "accept": false, - "reason": "RoPE tests and finegrained FP8 runtime changes are not the same bug/change." + "reason": "Molmo2 addition and workflow normalization are unrelated." }, { - "left": "pull_request:46907", - "right": "pull_request:47006", + "left": "pull_request:43451", + "right": "pull_request:46977", "accept": false, - "reason": "EdgeTam and Gemma3 fixes target different model behaviors." + "reason": "Molmo2 model addition and Mllama arg-name cleanup are unrelated." }, { "left": "pull_request:46957", - "right": "pull_request:46970", + "right": "pull_request:46965", "accept": false, - "reason": "Finegrained FP8 kernel changes are unrelated to MusicGen Melody generation fixes." + "reason": "Finegrained FP8 decode optimization and DeepSeek V4 use_cache=False fix solve different problems." }, { - "left": "pull_request:46957", - "right": "pull_request:46986", + "left": "pull_request:46196", + "right": "pull_request:47014", "accept": false, - "reason": "Finegrained FP8 kernel changes are unrelated to GLM OCR output mismatch fixes." + "reason": "wav2vec2 SDPA dispatch and T5 SDPA enablement are different model-family changes." }, { - "left": "pull_request:46907", - "right": "pull_request:46971", + "left": "pull_request:46658", + "right": "pull_request:46836", "accept": false, - "reason": "EdgeTam failures and Mamba2 OOM test adjustments are unrelated." + "reason": "Step-3.7 Flash model tooling and Unlimited OCR model addition are unrelated." }, { - "left": "pull_request:46907", - "right": "pull_request:46986", + "left": "pull_request:46891", + "right": "pull_request:47014", "accept": false, - "reason": "EdgeTam failures and GLM OCR failures are unrelated." + "reason": "PEFT CI workflow changes and T5 SDPA enablement are unrelated." }, { "left": "pull_request:46836", - "right": "pull_request:47008", + "right": "pull_request:47014", "accept": false, - "reason": "Unlimited OCR addition is unrelated to Gemma4 dtype-casting fixes." + "reason": "Unlimited OCR model addition and T5 SDPA enablement are unrelated." }, { - "left": "pull_request:46950", - "right": "pull_request:46971", + "left": "pull_request:46658", + "right": "pull_request:47014", "accept": false, - "reason": "Same model family, but one is a runtime/kernel fix and the other is a test-only loading change; not the same concrete bug." + "reason": "Step-3.7 Flash model tooling and T5 SDPA enablement are unrelated." }, { - "left": "pull_request:46907", - "right": "pull_request:46950", + "left": "pull_request:46957", + "right": "pull_request:47014", "accept": false, - "reason": "EdgeTam import/config failures are unrelated to Mamba2 OOM behavior." + "reason": "Finegrained FP8 decode optimization and T5 SDPA enablement are unrelated." }, { - "left": "pull_request:46949", - "right": "pull_request:46971", + "left": "pull_request:46912", + "right": "pull_request:47014", "accept": false, - "reason": "Whisper/audioflamingo failures are unrelated to Mamba2 test OOM fixes." + "reason": "RoPE test sweep and T5 SDPA enablement are unrelated." }, { - "left": "pull_request:46951", - "right": "pull_request:46971", + "left": "pull_request:46912", + "right": "pull_request:46957", "accept": false, - "reason": "PEFT integration fixes are unrelated to Mamba2 OOM-related test changes." + "reason": "RoPE test sweep and finegrained FP8 decode optimization are unrelated." } ] }, @@ -22485,7 +22663,7 @@ "soft_pairs": [ "pull_request:40755|pull_request:41224" ], - "prepared_review_unit_hash": "2900e246b1dbbe28f319685d33495fb2578b8c8c8d604ae8a1dc8b4a065fb6c2", + "prepared_review_unit_hash": "1f05ae78bccab1e6899c054e9cade9dc22e4581f6982d232ab68f10daf58eecb", "component_budget": { "node_count": 2, "item_count": 2, @@ -22507,18 +22685,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are two unrelated feature PRs: one adds TimesFM covariate forecasting, the other adds DINOv3 image classification support. They should not be clustered as duplicates.", - "confidence": 0.97, + "summary": "The two PRs are unrelated feature additions: TimesFM covariate forecasting vs. DINOv3 image classification support. The clustering is a false positive based only on loose similarity.", + "confidence": 0.98, "canonical_issue_reason": null, - "canonical_pr_reason": "pull_request:41224 is the more focused and self-contained feature PR, with a clear model-specific code path and matching docs/tests for DINOv3 image classification.", + "canonical_pr_reason": null, "best_issue_reason": null, - "best_pr_reason": "pull_request:41224 is the strongest standalone PR in the set because it targets a concrete, narrowly scoped capability addition and appears easier to review and merge cleanly.", + "best_pr_reason": "No single PR is a good cluster representative because the PRs target different models, code paths, and user-facing features.", "soft_edge_verdicts": [ { "left": "pull_request:40755", "right": "pull_request:41224", "accept": false, - "reason": "Reject: these PRs address different model families and different problems. TimesFM covariate forecasting is unrelated to DINOv3 image classification, and they would not plausibly be merged into one PR or fix the same code-path issue." + "reason": "Different underlying changes: 40755 adds covariate-based forecasting for TimesFM, while 41224 adds DINOv3ViT image classification support. They do not fix the same bug or implement the same feature." } ] }, @@ -22639,151 +22817,151 @@ "pull_request:43823|pull_request:45213", "pull_request:43251|pull_request:43254" ], - "prepared_review_unit_hash": "0a662ef195869392924a67712893b4fe0487b72a64ba490077f6d83f7dab907e", + "prepared_review_unit_hash": "8ae4c08b65f288f617f07f046eafa22a396d4d4831dc67154054fb3e33eaab9e", "component_budget": { "node_count": 13, "item_count": 13, "soft_pair_count": 18, - "serialized_chars": 34782, - "estimated_input_tokens": 8696, - "estimated_eval_tokens": 17648 + "serialized_chars": 34787, + "estimated_input_tokens": 8697, + "estimated_eval_tokens": 17650 }, "budget": { "node_count": 13, "item_count": 13, "soft_pair_count": 18, - "serialized_chars": 34782, - "estimated_input_tokens": 8696, - "estimated_eval_tokens": 17648 + "serialized_chars": 34787, + "estimated_input_tokens": 8697, + "estimated_eval_tokens": 17650 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Mostly unrelated PRs; the only true duplicate pair is the two loss_utils patches that both forward supported kwargs into fixed_cross_entropy. Everything else targets different models, subsystems, or documentation/tests and should remain separate.", - "confidence": 0.89, + "summary": "This set is mostly unrelated open PRs spanning different models, docs, and utilities. The only true duplicate-looking pair is the two `fixed_cross_entropy` changes (43251/43254), which address the same loss-helper bug and could plausibly be merged into one PR.", + "confidence": 0.9, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 43251 is a strong canonical representative for the duplicate loss_utils fix because it addresses the same fixed_cross_entropy kwargs-forwarding bug as 43254 in the same code path. That said, both PRs explicitly target issue 43240, so the cluster does not support a claim that 43251 is uniquely tied to the shared bug report.", + "canonical_pr_reason": "PR 43251 is the clearest representative of the duplicate loss-helper fix: it has the explicit issue target and directly forwards the new kwargs into `nn.functional.cross_entropy`. PR 43254 is an alternate, near-identical implementation of the same underlying change.", "best_issue_reason": null, - "best_pr_reason": "PR 43251 is the best overall representative of the cluster\u2019s only real duplicate set: it directly fixes fixed_cross_entropy in loss_utils.py and cleanly forwards the supported cross_entropy kwargs. PR 43254 is a very close sibling, but the pair is still the clearest actionable duplicate relationship in the cluster.", + "best_pr_reason": "PR 43251 is the best overall representative of the only duplicate pair in this set; it is the most explicit and complete version of the cross-entropy kwargs fix, while the rest of the PRs are unrelated to that bug.", "soft_edge_verdicts": [ { "left": "pull_request:41855", "right": "pull_request:43823", "accept": false, - "reason": "Tokenizer method additions and a new MobileLLM model are unrelated changes." + "reason": "Tokenizer API additions for Mistral are unrelated to adding the MobileLLM model; different code paths and different changes." }, { "left": "pull_request:41855", "right": "pull_request:43424", "accept": false, - "reason": "Missing tokenizer methods and executorch export tests do not share the same underlying bug or change." + "reason": "Mistral tokenizer methods vs ExecuTorch dynamic-shape export tests; no shared bug or change." }, { "left": "pull_request:42668", "right": "pull_request:43823", "accept": false, - "reason": "Processor auto-loading robustness and MobileLLM model creation are separate concerns." + "reason": "Processor-auto robustness work is unrelated to introducing MobileLLM; broad infra overlap only." }, { "left": "pull_request:43251", "right": "pull_request:43823", "accept": false, - "reason": "Loss-function kwargs forwarding and MobileLLM model addition are unrelated." + "reason": "Loss-helper kwargs fix and MobileLLM model addition are unrelated." }, { "left": "pull_request:43251", "right": "pull_request:44827", "accept": false, - "reason": "A cross_entropy wrapper fix and Mistral4 test/model changes are not the same code-path problem." + "reason": "Cross-entropy kwargs fix and Mistral4 test/model changes are different issues." }, { "left": "pull_request:43424", "right": "pull_request:44827", "accept": false, - "reason": "Dynamic-shape export tests and Mistral4 fixes target different issues." + "reason": "ExecuTorch dynamic-shape export test is unrelated to Mistral4 test/fix work." }, { "left": "pull_request:43424", "right": "pull_request:43823", "accept": false, - "reason": "Executorch exportability tests are unrelated to adding MobileLLM." + "reason": "Exportability test work does not match the MobileLLM model addition." }, { "left": "pull_request:43823", "right": "pull_request:45655", "accept": false, - "reason": "New model implementation and config-resolution ordering fix are different changes." + "reason": "MobileLLM addition and config-resolution-order fix are unrelated." }, { "left": "pull_request:43823", "right": "pull_request:43995", "accept": false, - "reason": "MobileLLM model addition and Falcon output-collection refactoring are unrelated." - }, - { - "left": "pull_request:43823", - "right": "pull_request:45213", - "accept": false, - "reason": "Model addition and the model-creation skill are not the same underlying bug or feature." + "reason": "New model addition vs Falcon output-interface refactor; different model families and code paths." }, { "left": "pull_request:43995", "right": "pull_request:45213", "accept": false, - "reason": "Falcon refactoring and a new-model skill are unrelated." + "reason": "Falcon refactor and the model-creation skill are not the same change; one is model code, the other is tooling/docs." }, { "left": "pull_request:43995", "right": "pull_request:44827", "accept": false, - "reason": "Falcon output-interface refactoring and Mistral4 fixes do not address the same problem." + "reason": "Falcon output interface refactor is unrelated to Mistral4 test fixes." }, { "left": "pull_request:44827", "right": "pull_request:45170", "accept": false, - "reason": "Mistral4 test fixes and a CLIP-like layernorm rename are different model-specific changes." + "reason": "Mistral4 test/model fixes and CLIP-like `layrnorm` renaming are unrelated changes." }, { "left": "pull_request:44827", "right": "pull_request:46267", "accept": false, - "reason": "Mistral4 fixes and serve chat parsing updates are unrelated subsystems." + "reason": "Mistral4 fixes and serve chat-parsing updates address different subsystems and bugs." }, { "left": "pull_request:44827", "right": "pull_request:46048", "accept": false, - "reason": "Mistral4 changes and adding a TDT loss kernel are unrelated." + "reason": "Mistral4 test/model work is unrelated to adding the TDT loss kernel." }, { "left": "pull_request:44827", "right": "pull_request:45655", "accept": false, - "reason": "Mistral4 test/model work and config resolution ordering fix are separate bugs." + "reason": "Mistral4 changes and the `cls.config` resolution order fix are unrelated." }, { "left": "pull_request:44827", "right": "pull_request:45213", "accept": false, - "reason": "Mistral4 fixes and the model-creation skill are not the same change." + "reason": "Model/test fixes for Mistral4 are unrelated to the new-model creation skill." + }, + { + "left": "pull_request:43823", + "right": "pull_request:45213", + "accept": false, + "reason": "MobileLLM model addition is not the same as the model-creation skill/documentation PR." }, { "left": "pull_request:43251", "right": "pull_request:43254", "accept": true, - "reason": "Both PRs patch the same fixed_cross_entropy wrapper in loss_utils.py to forward supported cross_entropy kwargs; they are plausibly mergeable into one PR." + "reason": "Both PRs fix the same `fixed_cross_entropy` code path by adding supported kwargs to `nn.functional.cross_entropy`; they are near-identical implementations of the same underlying bug." } ] }, "evaluator_result": { - "accept": false, - "feedback": "Rejecting because the canonical PR rationale overstates the evidence: PR 43254 also has an explicit issue target (43240), so 43251 is not the only change in the cluster tied to a shared bug report. The rest of the summary is mostly grounded, and the 43251\u219443254 mergeability verdict looks conservative, but the unique-target claim is not supported by the packet." + "accept": true, + "feedback": "The summary is grounded in the packet and stays conservative. The only supported duplicate pair is 43251/43254, which shares the same issue target and the same `fixed_cross_entropy` file. The other pair verdicts correctly reject soft similarity alone as insufficient evidence, and the representative PR choice is reasonable." }, "evaluator_used": true, - "retried": true, + "retried": false, "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null @@ -22807,7 +22985,7 @@ "soft_pairs": [ "pull_request:42975|pull_request:42976" ], - "prepared_review_unit_hash": "1da5dcb044e8c50e52f57a2a8f830505839d3d81eaad7e40ade5819c9fd1cac2", + "prepared_review_unit_hash": "4dad9fc1980a7febe22cd8adea155ec5c51697fcf34f4ec449197a9ba850601b", "component_budget": { "node_count": 2, "item_count": 2, @@ -22829,18 +23007,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Two open PRs both update GitHub Actions in workflow files, but they target different dependency bumps and different motivations. They overlap in filenames because they touch shared workflows, not because they fix the same concrete issue.", - "confidence": 0.94, + "summary": "Two open PRs both update GitHub Actions workflow dependencies, but they target different upgrade sets: #42975 is a broad Node 24 compatibility sweep across many workflows, while #42976 bumps Docker-related actions and a few other workflow actions to latest versions. They overlap in workflow files but not in the concrete change being made.", + "confidence": 0.95, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 42975 is the better representative because it has a clearer, narrower purpose: upgrading workflow actions specifically for Node 24 compatibility. That makes the underlying change more concrete than the broader 'latest versions' update in 42976.", + "canonical_pr_reason": "No clear duplicate canonical PR: #42975 is a broad Actions/node-compatibility bump, while #42976 is a separate action-version refresh centered on docker/build-push-action and related workflows.", "best_issue_reason": null, - "best_pr_reason": "PR 42975 is the best single PR to keep as canonical for this cluster since its intent is explicit and technical, while 42976 is a separate maintenance sweep over different actions. If one PR must anchor the cluster, 42975 is more precise.", + "best_pr_reason": "If one must be treated as the representative maintenance PR, #42975 is the broader workflow-actions upgrade, but it is not a true duplicate of #42976.", "soft_edge_verdicts": [ { "left": "pull_request:42975", "right": "pull_request:42976", "accept": false, - "reason": "Reject: these are both workflow dependency updates, but they change different actions for different reasons (Node 24 compatibility vs latest-version bumps). They do not appear to fix the same concrete bug or change the same code path, and they would not naturally merge into one PR beyond generic maintenance." + "reason": "Different underlying changes: #42975 upgrades many Actions for Node 24 compatibility; #42976 mainly bumps docker/build-push-action and related workflow action versions. Shared workflow filenames alone are not enough to treat them as the same fix." } ] }, @@ -22875,12 +23053,12 @@ "pull_request:43448|pull_request:45490", "pull_request:45490|pull_request:46562" ], - "prepared_review_unit_hash": "06bb08ff9a287e41612541b64cd72817093cd13a6f037373ebfd8d730c3d1e58", + "prepared_review_unit_hash": "fffca1081be18c7713c1072bc9560bae860d57696874fd494b4b760f4509daee", "component_budget": { "node_count": 4, "item_count": 4, "soft_pair_count": 4, - "serialized_chars": 13046, + "serialized_chars": 13047, "estimated_input_tokens": 3262, "estimated_eval_tokens": 6780 }, @@ -22888,7 +23066,7 @@ "node_count": 4, "item_count": 4, "soft_pair_count": 4, - "serialized_chars": 13046, + "serialized_chars": 13047, "estimated_input_tokens": 3262, "estimated_eval_tokens": 6780 }, @@ -22897,36 +23075,36 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "All four pull requests add different models/features (ViT NEPA, Molmo, CTSM, Step3p7 Flash). They share broad integration patterns and some auto/doc files, but they do not look like the same underlying change.", - "confidence": 0.97, + "summary": "All four PRs are distinct model-addition changes (ViT NEPA, Molmo, CTSM, and Step3p7). They only overlap in shared Transformers registration/docs scaffolding, not in the underlying model or code-path change, so none of the soft edges should be merged as duplicates.", + "confidence": 0.98, "canonical_issue_reason": null, - "canonical_pr_reason": "No clear canonical duplicate PR: each PR introduces a distinct model family and code path, so none is a true stand-in for the others.", + "canonical_pr_reason": "No canonical PR: each item introduces a different model and implementation path, so there is no single duplicate representative.", "best_issue_reason": null, - "best_pr_reason": "PR 46562 is the strongest representative only in the sense that it is the most mature model-addition PR with broad integration and review activity, but it is still not a duplicate of the others.", + "best_pr_reason": "No global best PR for duplicate triage; the PRs are unrelated beyond shared boilerplate, so none is a suitable canonical duplicate target.", "soft_edge_verdicts": [ { "left": "pull_request:42978", "right": "pull_request:43448", "accept": false, - "reason": "Different model additions (ViT NEPA vs Molmo); shared auto/docs files are incidental to model registration, not the same bug or change." + "reason": "ViT NEPA and Molmo are different model additions with different filenames, configs, and tests; shared auto/docs files are just boilerplate." }, { "left": "pull_request:42978", "right": "pull_request:45490", "accept": false, - "reason": "ViT NEPA and CTSM are unrelated model integrations with different architectures and code paths; overlap is only in framework plumbing." + "reason": "ViT NEPA is a vision model PR, while CTSM is a time-series model PR; they touch different implementation paths and are not the same change." }, { "left": "pull_request:43448", "right": "pull_request:45490", "accept": false, - "reason": "Molmo and CTSM are distinct model families; similarity is limited to common Transformer repo scaffolding, not a single mergeable change." + "reason": "Molmo and CTSM add unrelated models with different components (multimodal vs time-series); overlap is only in shared repository wiring." }, { "left": "pull_request:45490", "right": "pull_request:46562", "accept": false, - "reason": "CTSM (time-series) and Step3p7 Flash (multimodal) are different models and implementation areas; they should not be merged as duplicates." + "reason": "CTSM and Step3p7 are different model families and code paths; the shared auto/docs edits do not indicate the same underlying bug or feature." } ] }, @@ -22975,7 +23153,7 @@ "pull_request:46539|pull_request:46922", "pull_request:46886|pull_request:46922" ], - "prepared_review_unit_hash": "02fa5c03b91078d67ad8e371d16d721ba72a184d59a8f434a9aee483c268a947", + "prepared_review_unit_hash": "80596f872106e6d76579de6324d81f6daf1d51d77829f586cd8a5c775a326494", "component_budget": { "node_count": 7, "item_count": 7, @@ -22997,108 +23175,108 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "This cluster is mostly unrelated PRs that touch different models/subsystems and appear to be separate fixes. The only clear duplicate pair is the two YOLOS device-mismatch PRs, which both target the same issue and same code path in `modeling_yolos.py`.", - "confidence": 0.91, + "summary": "This cluster is mostly unrelated PRs across different models and subsystems. The only true duplicate/same-change pair is the YOLOS device-mismatch fix in PRs 46886 and 46922; the rest should stay separate.", + "confidence": 0.97, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 46922 is the best canonical PR because it addresses the YOLOS device-mismatch path more completely in `modeling_yolos.py` and is tied to the same tracking issue as the other YOLOS PR.", + "canonical_pr_reason": "PR 46922 is the best canonical representative for the YOLOS fix because it contains the concrete forward-path device move plus the split-module adjustment, and it matches the explicit issue target shared with 46886.", "best_issue_reason": null, - "best_pr_reason": "PR 46922 is the strongest representative of the actual fix: it patches the runtime device placement in the YOLOS forward path and also adds the no-split module guard for `device_map=\"auto\"`.", + "best_pr_reason": "PR 46922 is the strongest standalone representative: it directly fixes the YOLOS device mismatch in the model code and appears to be the more complete version of the same change as 46886.", "soft_edge_verdicts": [ { "left": "pull_request:43291", "right": "pull_request:43665", "accept": false, - "reason": "Different models and different problems: Whisper tokenization/tests versus CLIP/ViT test edits." + "reason": "Both are test-only PRs, but they target different models and different failures; no shared concrete bug or change." }, { "left": "pull_request:43291", "right": "pull_request:44815", "accept": false, - "reason": "Unrelated code paths: Whisper decoding fix versus FP8 dequant/loading changes." + "reason": "Whisper test cleanup vs FP8/dequant loading logic; unrelated code paths." }, { "left": "pull_request:43291", "right": "pull_request:46380", "accept": false, - "reason": "Different bug domains: Whisper tokenization versus generation watermarking/config validation." + "reason": "Whisper test fix vs generation/watermarking config fix; different underlying issues." }, { "left": "pull_request:43291", "right": "pull_request:46922", "accept": false, - "reason": "Whisper tokenizer/tests and YOLOS device-mapping are unrelated fixes." + "reason": "Whisper tokenizer/tests and YOLOS device placement are unrelated bugs." }, { "left": "pull_request:43665", - "right": "pull_request:46539", + "right": "pull_request:44815", "accept": false, - "reason": "Test-only edits for different model families and no shared underlying bug." + "reason": "Test assertions in CLIP/ViT are unrelated to FP8/dequant conversion changes." }, { "left": "pull_request:43665", - "right": "pull_request:44815", + "right": "pull_request:46380", "accept": false, - "reason": "CLIP/ViT test changes are unrelated to FP8 dequantization/loading." + "reason": "Different model test edits and different product bug; not the same change." }, { "left": "pull_request:43665", - "right": "pull_request:46380", + "right": "pull_request:46539", "accept": false, - "reason": "Different subsystems and no shared concrete failure mode." + "reason": "Failing model tests vs CI memory/device-map infrastructure fix; different underlying problem." }, { "left": "pull_request:43665", "right": "pull_request:46922", "accept": false, - "reason": "No overlap beyond being test/model repo changes; underlying bugs differ." + "reason": "Unrelated model test modifications vs YOLOS inference/device-mapping fix." }, { "left": "pull_request:44815", "right": "pull_request:46539", "accept": false, - "reason": "FP8 quantization/loading fix and Phimoe CI/test fix are unrelated." + "reason": "FP8/dequant loading behavior and CI memory limiting are distinct issues." }, { "left": "pull_request:44815", "right": "pull_request:46380", "accept": false, - "reason": "Different functionality and different failure paths." + "reason": "Different subsystems: quantization/dequant vs generation config validation." }, { "left": "pull_request:44815", "right": "pull_request:46922", "accept": false, - "reason": "FP8 dequant logic and YOLOS device placement are not the same bug." + "reason": "FP8 quantization/dequant changes are not the same as YOLOS device mismatch." }, { "left": "pull_request:46380", "right": "pull_request:46539", "accept": false, - "reason": "Generation config validation changes versus CI/runtime test harness changes are unrelated." + "reason": "Generation config watermarking fix and CI Phimoe integration fix are unrelated." }, { "left": "pull_request:46380", "right": "pull_request:46922", "accept": false, - "reason": "Different models and different concrete fixes." + "reason": "Generation validation changes do not match the YOLOS device-mismatch code path." }, { "left": "pull_request:46539", "right": "pull_request:46922", "accept": false, - "reason": "Phimoe CI/test infrastructure changes do not match a YOLOS runtime bug." + "reason": "CI memory/device-map test harness changes are unrelated to YOLOS model placement bugs." }, { "left": "pull_request:46886", "right": "pull_request:46922", "accept": true, - "reason": "Both PRs fix the same YOLOS `device_map=\"auto\"` device-mismatch issue in `src/transformers/models/yolos/modeling_yolos.py`, and could plausibly be merged into one fix." + "reason": "Same YOLOS device-mismatch issue, same model file, and same explicit issue target; 46922 is essentially the more complete version of the same fix." } ] }, "evaluator_result": { "accept": true, - "feedback": "The cluster-level judgment is grounded: the non-YOLOS PRs appear to be unrelated fixes across different subsystems, and the only plausible mergeable duplicate pair is the two YOLOS PRs. One small precision note: PR 46886 does not itself patch the forward-path device mismatch, so saying both YOLOS PRs fix the same code path is a bit strong; it is safer to describe them as the same tracked YOLOS device-mismatch issue in the same file." + "feedback": "The cluster summary is grounded in the packet, and the pairwise rejections are conservative. The YOLOS pair (46886, 46922) is the only clearly mergeable same-bug fix pair, with shared file and issue target; the other PRs address unrelated models or subsystems." }, "evaluator_used": true, "retried": false, @@ -23128,52 +23306,52 @@ "pull_request:43488|pull_request:43532", "pull_request:43532|pull_request:46747" ], - "prepared_review_unit_hash": "0dce6171f925a9ee718ec691816761ba4a152c1b9d452a1901faf5cb7535c889", + "prepared_review_unit_hash": "2015cede188442ba92047922013946fdc638ddefa79014eaa3c975b92333ee76", "component_budget": { "node_count": 3, "item_count": 3, "soft_pair_count": 3, - "serialized_chars": 6348, - "estimated_input_tokens": 1587, - "estimated_eval_tokens": 3430 + "serialized_chars": 6349, + "estimated_input_tokens": 1588, + "estimated_eval_tokens": 3432 }, "budget": { "node_count": 3, "item_count": 3, "soft_pair_count": 3, - "serialized_chars": 6348, - "estimated_input_tokens": 1587, - "estimated_eval_tokens": 3430 + "serialized_chars": 6349, + "estimated_input_tokens": 1588, + "estimated_eval_tokens": 3432 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "All three items are unrelated '[don't merge]' PRs touching different workflow/setup/test-harness paths; they do not look like duplicates of the same underlying fix or change.", - "confidence": 0.97, + "summary": "Three unrelated '[don't merge]' PRs for repo-bot/workflow testing; they have superficial similarity but change different files and behaviors.", + "confidence": 0.95, "canonical_issue_reason": null, - "canonical_pr_reason": null, + "canonical_pr_reason": "No canonical PR: the PRs touch different areas (setup/deps and tests, comment CI workflows, security-gate workflow) and are not the same fix.", "best_issue_reason": null, - "best_pr_reason": null, + "best_pr_reason": "No best PR for deduping: none of these is a plausible merge target for the others, despite the shared bot/test-oriented framing.", "soft_edge_verdicts": [ { "left": "pull_request:43488", "right": "pull_request:46747", "accept": false, - "reason": "Different changes: 43488 edits setup/versioning and a ViT test, while 46747 only changes the PR security-gate workflow trigger. No shared concrete bug or mergeable code-path fix." + "reason": "Different changes: 43488 updates setup/dependencies and a test, while 46747 only changes the PR security-gate workflow. Not the same underlying bug/change." }, { "left": "pull_request:43488", "right": "pull_request:43532", "accept": false, - "reason": "43488 is a repo-bot/setup/version-format experiment; 43532 rewrites several CI/comment workflows plus testing helpers. They are not the same underlying problem." + "reason": "43488 is a setup/dependency/version-bump test PR; 43532 rewires comment/scheduled/slack workflows and test utilities. Shared 'don't merge' framing is too broad." }, { "left": "pull_request:43532", "right": "pull_request:46747", "accept": false, - "reason": "Both touch workflows, but 43532 is about comment/notification CI and test harness behavior, while 46747 changes the security-gate workflow trigger. Too different to be one PR." + "reason": "43532 alters multiple workflow files plus testing code; 46747 only switches the security-gate trigger from pull_request_target to pull_request. Distinct code paths." } ] }, @@ -23203,12 +23381,12 @@ "soft_pairs": [ "pull_request:43612|pull_request:43613" ], - "prepared_review_unit_hash": "6f520d14f01a74e6abf08868b4c0794cb79ce2fff0157b688246fc6ad9b7b4f0", + "prepared_review_unit_hash": "b00106de8ae0d5471d052ae341518a25ed4869cb78898102ae2d754aee51eb9a", "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 }, @@ -23216,7 +23394,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 6082, + "serialized_chars": 6083, "estimated_input_tokens": 1521, "estimated_eval_tokens": 3298 }, @@ -23225,18 +23403,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are two separate feature PRs: one adds Promptable Concept Segmentation, the other adds Promptable Visual Segmentation. They overlap in pipeline/docs scaffolding but target different tasks, model integrations, and test suites, so they should not be deduplicated.", + "summary": "These are two separate feature PRs for different segmentation pipelines; they share scaffolding and docs but do not fix the same concrete problem or implement the same change.", "confidence": 0.97, "canonical_issue_reason": null, - "canonical_pr_reason": "No canonical PR: 43612 and 43613 implement different pipelines for different segmentation tasks, so they are not the same underlying change.", + "canonical_pr_reason": null, "best_issue_reason": null, - "best_pr_reason": "If a single representative PR is needed, 43613 is the broader and more integrated change, touching more model backends and tests. But it should still remain separate from 43612.", + "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "pull_request:43612", "right": "pull_request:43613", "accept": false, - "reason": "Reject: despite shared pipeline/documentation plumbing, the PRs implement different concrete features (Promptable Concept Segmentation vs Promptable Visual Segmentation) with different filenames, model hooks, and test coverage. They are not the same bug/change and would not plausibly be merged into one PR." + "reason": "Different pipeline features: Promptable Concept Segmentation vs Promptable Visual Segmentation. The overlap is mostly template/docs/shared infra, not the same code-path change, so they should not be merged as duplicates." } ] }, @@ -23266,7 +23444,7 @@ "soft_pairs": [ "pull_request:43651|pull_request:43663" ], - "prepared_review_unit_hash": "26d896756b36f165d4ed521f96f6ba052a5f8a25cefc60b4d49b649e2885b350", + "prepared_review_unit_hash": "ab42c884a4ba5ec065ba641bc05b1b96391fb119417b6b46b650a95ec59a84d2", "component_budget": { "node_count": 2, "item_count": 2, @@ -23288,18 +23466,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These two PRs both modify `Trainer` for custom trainer extensibility, but they address different code paths: one changes gradient-accumulation loss scaling, the other refactors input-column filtering based on model signatures. They are not duplicates.", - "confidence": 0.95, + "summary": "Both items are open PRs in `src/transformers/trainer.py`, but they address different trainer behaviors: PR 43651 changes gradient-accumulation loss scaling, while PR 43663 adds an overridable helper for signature-column filtering. They are related by file location only, not by the same bug or change.", + "confidence": 0.97, "canonical_issue_reason": null, - "canonical_pr_reason": "PR #43651 is the better anchor: it makes a concrete behavioral change in `training_step` with a clearly scoped bug/feature fix around gradient accumulation loss scaling.", + "canonical_pr_reason": "No single canonical PR: these are distinct trainer changes with different code paths and goals, so neither can stand in for the other as a duplicate.", "best_issue_reason": null, - "best_pr_reason": "PR #43651 is the strongest standalone PR in the cluster because it targets a specific training correctness issue rather than a refactor of trainer internals.", + "best_pr_reason": "PR 43651 is the more concrete, narrowly scoped behavior change, but it is still not a duplicate of PR 43663; keeping it separate best reflects the underlying work.", "soft_edge_verdicts": [ { "left": "pull_request:43651", "right": "pull_request:43663", "accept": false, - "reason": "Both touch `trainer.py` and mention custom trainers, but #43651 fixes loss normalization during gradient accumulation while #43663 refactors signature-column inspection. Different underlying behavior, not one mergeable change." + "reason": "Do not merge as duplicates: 43651 changes loss normalization during training_step, while 43663 refactors signature-column detection for column filtering. Same file, different functionality, different code paths." } ] }, @@ -23329,12 +23507,12 @@ "soft_pairs": [ "pull_request:43656|pull_request:43842" ], - "prepared_review_unit_hash": "e9aa2658a6859b904820d3810201648c9ecc38a38846fa655fad1c743beae328", + "prepared_review_unit_hash": "a830b2bd6c70b20253b079248963a9567f31ca65aa2d28c4060a6bdb536100a6", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 3664, + "serialized_chars": 3663, "estimated_input_tokens": 916, "estimated_eval_tokens": 2088 }, @@ -23342,7 +23520,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 3664, + "serialized_chars": 3663, "estimated_input_tokens": 916, "estimated_eval_tokens": 2088 }, @@ -23351,18 +23529,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Two open PRs touch the same file and both mention a TypeAdapter-related NameError, but they fix different typing/runtime issues: one adds future annotations and broader annotation cleanup, the other targets the missing-pydantic TypeAdapter reference. They are related, but not a clear duplicate pair.", - "confidence": 0.84, + "summary": "Two open PRs both touch `src/transformers/cli/serve.py` and both mention a `TypeAdapter` NameError, but they target different issues and appear to use different fixes for different trigger conditions. They should not be merged as duplicates.", + "confidence": 0.92, "canonical_issue_reason": null, - "canonical_pr_reason": null, + "canonical_pr_reason": "No canonical PR: the PRs address closely related but distinct `TypeAdapter`/typing problems in the CLI, with different issue targets and different corrective approaches.", "best_issue_reason": null, - "best_pr_reason": "PR #43842 is the more direct fix for the explicit pydantic-related TypeAdapter NameError, with a narrowly scoped change that matches the reported failure mode.", + "best_pr_reason": "Neither PR is a clear global winner because each is narrowly scoped to a different reported bug; keep both as separate PRs.", "soft_edge_verdicts": [ { "left": "pull_request:43656", "right": "pull_request:43842", "accept": false, - "reason": "Both are in transformers/cli/serve.py and both touch type annotations, but #43656 is a broader future-annotations/typing cleanup while #43842 specifically fixes the TypeAdapter reference when pydantic is absent. They overlap in area, not in the same concrete bug fix." + "reason": "Reject: despite the shared file and similar wording, these point to different underlying bugs/conditions (one is a general CLI `TypeAdapter` NameError fix, the other is specifically when `pydantic` is not installed) and the diffs show different remedies." } ] }, @@ -23509,8 +23687,8 @@ "pull_request:44085|pull_request:44086", "pull_request:44059|pull_request:44086", "pull_request:44098|pull_request:44101", - "pull_request:44059|pull_request:44722", "pull_request:44018|pull_request:44071", + "pull_request:44059|pull_request:44722", "pull_request:44017|pull_request:44161", "pull_request:44017|pull_request:44722", "pull_request:44017|pull_request:44073", @@ -23554,112 +23732,112 @@ "pull_request:44044|pull_request:44129", "pull_request:44044|pull_request:44154" ], - "prepared_review_unit_hash": "3f2bbcad7b198e1d4fc5e773dbc832b157c568700862b02c5dc1470a8fc3d177", + "prepared_review_unit_hash": "bd1bd52f02e9da9a8f849ea57eede3e79eac423dc274c17007d6f18ecdea5661", "component_budget": { "node_count": 28, "item_count": 28, "soft_pair_count": 78, - "serialized_chars": 84901, - "estimated_input_tokens": 21226, - "estimated_eval_tokens": 42708 + "serialized_chars": 84894, + "estimated_input_tokens": 21224, + "estimated_eval_tokens": 42704 }, "budget": { "node_count": 27, "item_count": 27, "soft_pair_count": 72, - "serialized_chars": 75096, - "estimated_input_tokens": 18774, - "estimated_eval_tokens": 37804 + "serialized_chars": 75088, + "estimated_input_tokens": 18772, + "estimated_eval_tokens": 37800 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly a family of output-tracing refactors, but the artifacts split cleanly by model family. Only a few pairs are true duplicates or near-duplicates within the same concrete code path (notably GPT-Neo and GPT-J variants); most cross-model pairs are just parallel refactors sharing the same umbrella issue.", - "confidence": 0.72, + "summary": "Mostly a heterogeneous set of model-specific output-tracing refactors. The only meaningful duplicate candidates are the GPT-J/CodeGen variants; the rest touch different model code paths and should stay separate.", + "confidence": 0.73, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 44066 is the clearest canonical representative: it is explicit about the shared issue target, covers the central GPT-J/CodeGen output-tracing refactor, and looks like the broadest coherent implementation among the closely related GPT-J variants.", + "canonical_pr_reason": "PR 44066 is the best canonical PR for the shared GPT-J output-tracing work: it is explicitly issue-linked, covers the core GPT-J change, and is the broadest representative of the overlapping GPT-J/CodeGen refactor family.", "best_issue_reason": null, - "best_pr_reason": "PR 44066 is the best single PR to keep as the cluster anchor because it is the most explicit and broadly representative of the repeated GPT-J/CodeGen tracing change, while still being a concrete, mergeable code-path refactor.", + "best_pr_reason": "PR 44066 is the best overall representative in this cluster because it is the clearest, issue-linked example of the shared standardized output-tracing refactor and overlaps with the other GPT-J variants.", "soft_edge_verdicts": [ { "left": "pull_request:43996", "right": "pull_request:44085", "accept": false, - "reason": "Different models and different code paths (CVT/FNet vs GPT-J); same umbrella topic is not enough." + "reason": "Different models and code paths (CVT/FNet vs GPT-J). Same broad refactor theme is not enough for duplication." }, { "left": "pull_request:43996", "right": "pull_request:44044", "accept": false, - "reason": "CVT/FNet output tracing is unrelated to DeBERTa's tracing refactor." + "reason": "CVT/FNet output-tracing cleanup is unrelated to DeBERTa v2's attention/output refactor." }, { "left": "pull_request:44066", "right": "pull_request:44085", "accept": true, - "reason": "Same GPT-J output-tracing refactor family on the same code path; plausibly one merged PR." + "reason": "Both are GPT-J output-tracing refactors in the same file and appear to be the same underlying decorator migration work." }, { "left": "pull_request:44007", "right": "pull_request:44072", "accept": false, - "reason": "RegNet/ResNet/RT-DETR refactor vs EfficientNet refactor; different model implementations." + "reason": "RegNet and EfficientNet are separate model implementations; these are similar migrations, not the same bug/change." }, { "left": "pull_request:44072", "right": "pull_request:44722", "accept": false, - "reason": "EfficientNet and GPT-J/CodeGen are unrelated model families." + "reason": "EfficientNet vs GPT-J/CodeGen; different model code paths." }, { "left": "pull_request:44066", "right": "pull_request:44072", "accept": false, - "reason": "GPT-J/CodeGen vs EfficientNet are different concrete model refactors." + "reason": "GPT-J/CodeGen output tracing is unrelated to EfficientNet output handling." }, { "left": "pull_request:44013", "right": "pull_request:44044", "accept": false, - "reason": "MobileNetV2 and DeBERTa are unrelated code paths." + "reason": "MobileNetV2 and DeBERTa v2 are different implementations with different tracing paths." }, { "left": "pull_request:44066", "right": "pull_request:44086", "accept": false, - "reason": "GPT-J/CodeGen output tracing and MGP-STR output tracing are different model changes." + "reason": "GPT-J/CodeGen versus MGP-STR; same pattern, different code paths." }, { "left": "pull_request:44066", "right": "pull_request:44071", "accept": false, - "reason": "GPT-J/CodeGen vs MPT; same pattern, different implementation." + "reason": "GPT-J/CodeGen and MPT are separate refactors, not one concrete change." }, { "left": "pull_request:44018", "right": "pull_request:44068", - "accept": true, - "reason": "Both are GPT-Neo output-tracing refactors on the same model code path." + "accept": false, + "reason": "GPT2 and GPT-Neo are distinct models; these are parallel refactors, not duplicates." }, { "left": "pull_request:44066", "right": "pull_request:44068", "accept": false, - "reason": "GPT-J/CodeGen and GPT-Neo are separate model implementations." + "reason": "GPT-J/CodeGen output-tracing work is separate from GPT-Neo's refactor." }, { "left": "pull_request:44018", "right": "pull_request:44066", "accept": false, - "reason": "GPT-Neo and GPT-J/CodeGen are not the same underlying change." + "reason": "GPT2 and GPT-J/CodeGen are different model paths." }, { "left": "pull_request:44019", "right": "pull_request:44722", "accept": false, - "reason": "ResNet refactor vs GPT-J/CodeGen refactor; not duplicates." + "reason": "ResNet output tracing is unrelated to GPT-J/CodeGen." }, { "left": "pull_request:44086", @@ -23677,277 +23855,265 @@ "left": "pull_request:44018", "right": "pull_request:44019", "accept": false, - "reason": "GPT-Neo vs ResNet; same theme, different models." + "reason": "GPT2 and ResNet are entirely different implementations." }, { "left": "pull_request:44019", "right": "pull_request:44071", "accept": false, - "reason": "ResNet and MPT are separate implementations." + "reason": "ResNet versus MPT; not the same bug/change." }, { "left": "pull_request:44068", "right": "pull_request:44722", "accept": false, - "reason": "GPT-Neo and GPT-J/CodeGen are different model refactors." + "reason": "GPT-Neo and GPT-J/CodeGen are different model code paths." }, { "left": "pull_request:44018", "right": "pull_request:44116", "accept": false, - "reason": "GPT-Neo and Flaubert/XLM are unrelated code paths." + "reason": "GPT2 and Flaubert are different model families." }, { "left": "pull_request:44068", "right": "pull_request:44116", "accept": false, - "reason": "GPT-Neo refactor vs Flaubert refactor; not the same bug/change." + "reason": "GPT-Neo and Flaubert are unrelated model implementations." }, { "left": "pull_request:44071", "right": "pull_request:44722", "accept": false, - "reason": "MPT and GPT-J/CodeGen are different model families." + "reason": "MPT and GPT-J/CodeGen are separate refactors." }, { "left": "pull_request:44018", "right": "pull_request:44085", "accept": false, - "reason": "GPT-Neo and GPT-J are not the same concrete change." + "reason": "GPT2 versus GPT-J; different concrete code paths." }, { "left": "pull_request:44018", "right": "pull_request:44086", "accept": false, - "reason": "GPT-Neo and MGP-STR are unrelated." + "reason": "GPT2 and MGP-STR are unrelated models." }, { "left": "pull_request:44068", "right": "pull_request:44086", "accept": false, - "reason": "GPT-Neo and MGP-STR are different implementations." + "reason": "GPT-Neo and MGP-STR do not share the same underlying change." }, { "left": "pull_request:44071", "right": "pull_request:44085", "accept": false, - "reason": "MPT and GPT-J are unrelated model code paths." + "reason": "MPT versus GPT-J; same pattern, different implementation." }, { "left": "pull_request:44085", "right": "pull_request:44086", "accept": false, - "reason": "GPT-J and MGP-STR are different model families." + "reason": "GPT-J and MGP-STR are distinct model code paths." }, { "left": "pull_request:44059", "right": "pull_request:44086", "accept": false, - "reason": "GPT-2 refactor vs MGP-STR refactor; not duplicates." + "reason": "GPT2 output-tracing refactor is unrelated to MGP-STR." }, { "left": "pull_request:44098", "right": "pull_request:44101", "accept": false, - "reason": "ViLT and Flaubert/XLM are separate model changes." - }, - { - "left": "pull_request:44059", - "right": "pull_request:44722", - "accept": false, - "reason": "GPT-2 output tracing vs GPT-J/CodeGen output tracing are different concrete code paths." - }, - { - "left": "pull_request:44018", - "right": "pull_request:44071", - "accept": false, - "reason": "GPT-Neo and MPT are different models." + "reason": "ViLT and Flaubert/XLM are separate model-specific changes." }, { "left": "pull_request:44017", "right": "pull_request:44161", "accept": false, - "reason": "Segformer and LongT5 are unrelated refactors." + "reason": "SegFormer and LongT5 are unrelated implementations." }, { "left": "pull_request:44017", "right": "pull_request:44722", "accept": false, - "reason": "Segformer and GPT-J/CodeGen are not the same underlying change." + "reason": "SegFormer output tracing is unrelated to GPT-J/CodeGen." }, { "left": "pull_request:44017", "right": "pull_request:44073", "accept": false, - "reason": "Segformer and VisualBERT are different model code paths." + "reason": "SegFormer and VisualBERT are different model paths." }, { - "left": "pull_request:43997", - "right": "pull_request:44073", + "left": "pull_request:44068", + "right": "pull_request:44071", "accept": false, - "reason": "RegNet and VisualBERT are unrelated refactors." + "reason": "GPT-Neo and MPT are separate model refactors." }, { - "left": "pull_request:43997", - "right": "pull_request:44018", + "left": "pull_request:44071", + "right": "pull_request:44076", "accept": false, - "reason": "RegNet and GPT-Neo are different model implementations." + "reason": "MPT and ImageGPT are different code paths." }, { - "left": "pull_request:43997", - "right": "pull_request:44056", + "left": "pull_request:44073", + "right": "pull_request:44101", "accept": false, - "reason": "RegNet and MPNet are distinct code paths." + "reason": "VisualBERT and Flaubert/XLM are unrelated model families." }, { - "left": "pull_request:44071", - "right": "pull_request:44076", + "left": "pull_request:44074", + "right": "pull_request:44101", "accept": false, - "reason": "MPT and ImageGPT are different models." + "reason": "TextNet and Flaubert/XLM are different model implementations." }, { - "left": "pull_request:44074", - "right": "pull_request:44101", + "left": "pull_request:43997", + "right": "pull_request:44017", "accept": false, - "reason": "TextNet and Flaubert/XLM are unrelated refactors." + "reason": "RegNet and SegFormer are different model code paths." }, { "left": "pull_request:43997", - "right": "pull_request:44161", + "right": "pull_request:44056", + "accept": false, + "reason": "RegNet and MPNet are unrelated refactors." + }, + { + "left": "pull_request:43997", + "right": "pull_request:44018", "accept": false, - "reason": "RegNet and LongT5 are different model families." + "reason": "RegNet and GPT2 are different models." }, { "left": "pull_request:43996", "right": "pull_request:44161", "accept": false, - "reason": "CVT/FNet and LongT5 are unrelated changes." + "reason": "CVT/FNet output tracing is unrelated to LongT5." }, { "left": "pull_request:43996", "right": "pull_request:44017", "accept": false, - "reason": "CVT/FNet and Segformer are different model implementations." + "reason": "CVT/FNet and SegFormer are different model implementations." }, { "left": "pull_request:44010", "right": "pull_request:44071", "accept": false, - "reason": "SqueezeBERT and MPT are unrelated model refactors." + "reason": "SqueezeBERT and MPT are unrelated code paths." }, { "left": "pull_request:44010", "right": "pull_request:44073", "accept": false, - "reason": "SqueezeBERT and VisualBERT are different code paths." + "reason": "SqueezeBERT and VisualBERT are different model families." }, { "left": "pull_request:43997", - "right": "pull_request:44018", - "accept": false, - "reason": "RegNet and GPT-Neo are not the same underlying bug/change." - }, - { - "left": "pull_request:44056", - "right": "pull_request:44161", + "right": "pull_request:44074", "accept": false, - "reason": "MPNet and LongT5 are different model families." + "reason": "RegNet and TextNet are separate model-specific changes." }, { "left": "pull_request:43997", - "right": "pull_request:44056", + "right": "pull_request:44161", "accept": false, - "reason": "RegNet and MPNet are unrelated refactors." + "reason": "RegNet and LongT5 are unrelated implementations." }, { - "left": "pull_request:44010", - "right": "pull_request:44018", + "left": "pull_request:44074", + "right": "pull_request:44161", "accept": false, - "reason": "SqueezeBERT and GPT-Neo are different models." + "reason": "TextNet and LongT5 are different model code paths." }, { - "left": "pull_request:44071", + "left": "pull_request:44018", "right": "pull_request:44129", "accept": false, - "reason": "MPT and SpeechT5 are different concrete code paths." + "reason": "GPT2 and SpeechT5 are unrelated changes." }, { - "left": "pull_request:43997", + "left": "pull_request:44071", "right": "pull_request:44129", "accept": false, - "reason": "RegNet and SpeechT5 are unrelated changes." + "reason": "MPT and SpeechT5 are different model families." }, { - "left": "pull_request:43996", + "left": "pull_request:44056", "right": "pull_request:44161", "accept": false, - "reason": "CVT/FNet and LongT5 are different model implementations." + "reason": "MPNet and LongT5 are separate refactors." }, { "left": "pull_request:43996", - "right": "pull_request:44017", + "right": "pull_request:44007", "accept": false, - "reason": "CVT/FNet and Segformer are separate refactors." + "reason": "CVT/FNet and RegNet are different model code paths." }, { - "left": "pull_request:44010", - "right": "pull_request:44044", + "left": "pull_request:44044", + "right": "pull_request:44129", "accept": false, - "reason": "SqueezeBERT and DeBERTa are unrelated model code paths." + "reason": "DeBERTa v2 and SpeechT5 are unrelated model implementations." }, { "left": "pull_request:44044", - "right": "pull_request:44074", + "right": "pull_request:44154", "accept": false, - "reason": "DeBERTa and TextNet are different concrete changes." + "reason": "DeBERTa v2 and VITS are different model families." }, { "left": "pull_request:44066", "right": "pull_request:44722", "accept": true, - "reason": "Same GPT-J/CodeGen output-tracing refactor family; the later PR looks like a narrower variant of the same change." + "reason": "Both touch GPT-J output tracing in the same core file and represent the same standardized decorator migration." }, { "left": "pull_request:44085", "right": "pull_request:44722", "accept": true, - "reason": "Both touch GPT-J output tracing on the same underlying model code path." + "reason": "Both are GPT-J output-tracing refactors on the same code path; they could plausibly be merged into one PR." }, { "left": "pull_request:44085", "right": "pull_request:44129", "accept": false, - "reason": "GPT-J and SpeechT5 are different model families." + "reason": "GPT-J and SpeechT5 are unrelated model code paths." }, { "left": "pull_request:43996", "right": "pull_request:44007", "accept": false, - "reason": "CVT/FNet refactor vs RegNet/ResNet/RT-DETR refactor; not duplicates." + "reason": "CVT/FNet and RegNet are different model implementations." }, { "left": "pull_request:43996", "right": "pull_request:44129", "accept": false, - "reason": "CVT/FNet and SpeechT5 are unrelated." + "reason": "CVT/FNet and SpeechT5 are unrelated refactors." }, { "left": "pull_request:44044", "right": "pull_request:44129", "accept": false, - "reason": "DeBERTa and SpeechT5 are different implementations." + "reason": "DeBERTa v2 and SpeechT5 are different models." }, { "left": "pull_request:44044", "right": "pull_request:44154", "accept": false, - "reason": "DeBERTa and Vits are separate model code paths." + "reason": "DeBERTa v2 and VITS do not share the same underlying code change." } ] }, "evaluator_result": { "accept": true, - "feedback": "The summary is generally grounded in the packet: the cluster is indeed dominated by output-tracing refactors, and the conservative cross-model rejections are supported by the filenames/diffs. The accepted same-family pairs (GPT-Neo and GPT-J/CodeGen variants) look plausibly mergeable into a single bugfix PR, so I don\u2019t see an evidence overstatement that requires rejection." + "feedback": "The summary is grounded in the packet and stays conservative about the non-GPT-J pairs. The three accepted GPT-J/CodeGen pairs are supported by shared issue target, shared file coverage, and clearly overlapping output-tracing refactors that look mergeable into one bugfix PR." }, "evaluator_used": true, "retried": false, @@ -23985,20 +24151,20 @@ "pull_request:43997|pull_request:44010", "pull_request:44010|pull_request:44129" ], - "prepared_review_unit_hash": "384826c814c47b3ca0d37e837e10f3fc30607be08b81d8885eb8d0d75a46578d", + "prepared_review_unit_hash": "e7b14536036cb341195189c03dc1e5142c65ebd98975f7c1699ba725f0edaf45", "component_budget": { "node_count": 28, "item_count": 28, "soft_pair_count": 78, - "serialized_chars": 84901, - "estimated_input_tokens": 21226, - "estimated_eval_tokens": 42708 + "serialized_chars": 84894, + "estimated_input_tokens": 21224, + "estimated_eval_tokens": 42704 }, "budget": { "node_count": 8, "item_count": 8, "soft_pair_count": 6, - "serialized_chars": 17126, + "serialized_chars": 17127, "estimated_input_tokens": 4282, "estimated_eval_tokens": 8820 }, @@ -24007,48 +24173,48 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "These PRs are all thematically similar refactors to standardized output tracing/capture, but they target different model implementations and code paths. They are related, not duplicates, so the soft edges should not be merged.", - "confidence": 0.86, + "summary": "These are all model-specific PRs that migrate different architectures to standardized output tracing/capture_outputs patterns. They are thematically related, but they do not fix the same concrete code path and would not plausibly be merged into one PR.", + "confidence": 0.95, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 44129 is the most representative single item: it has an explicit issue target, a concrete SpeechT5 code-path refactor, and clear output-capture changes. It best anchors the cluster without overstating cross-model duplication.", + "canonical_pr_reason": "PR 44073 is the clearest representative of the cluster: it explicitly refactors output tracing with both can_return_tuple and capture_outputs, plus _can_record_outputs wiring, making it the most complete and recognizable example.", "best_issue_reason": null, - "best_pr_reason": "PR 44129 is the strongest standalone candidate for triage because it is the most substantiated by an issue link and has a focused, model-specific implementation change.", + "best_pr_reason": "PR 44073 best represents the shared theme because its title and diff most directly describe the standardized output-tracing migration, while still remaining a single, concrete model implementation change.", "soft_edge_verdicts": [ { "left": "pull_request:44024", "right": "pull_request:44076", "accept": false, - "reason": "Both are standardized output-tracing refactors, but they touch different models (FocalNet vs ImageGPT) and different internals. Same theme, not the same underlying bug/change." + "reason": "Both are output-tracing refactors, but for different models (FocalNet vs ImageGPT). Same migration pattern, not the same underlying bug/change, and not mergeable into one PR." }, { "left": "pull_request:44073", "right": "pull_request:44074", "accept": false, - "reason": "VisualBert and TextNet are separate model-specific refactors. They share a decorator/capture pattern, but not a concrete shared code-path defect." + "reason": "Both standardize output capture, but one is VisualBert and the other is TextNet. Different model code paths and separate implementation changes." }, { "left": "pull_request:44056", "right": "pull_request:44073", "accept": false, - "reason": "MPNet and VisualBert are distinct implementations with different output-tracing hookups. Similar infrastructure direction, but not mergeable as one fix for one bug." + "reason": "Both deal with standardized output tracing, but MPNet and VisualBert are distinct model implementations with different hooks and outputs. Not the same concrete fix." }, { "left": "pull_request:44056", "right": "pull_request:44074", "accept": false, - "reason": "These are separate model refactors in different files and model classes. The overlap is only the standardized output-capture approach." + "reason": "Shared refactor theme only; MPNet and TextNet touch different model internals and are independent PRs." }, { "left": "pull_request:43997", "right": "pull_request:44010", "accept": false, - "reason": "RegNet and SqueezeBert are unrelated model implementations. They both migrate to output capture, but that is a broad refactor theme, not the same underlying issue." + "reason": "Both are model-level output tracing migrations, but RegNet and SqueezeBert are unrelated code paths. They do not appear to be duplicate fixes." }, { "left": "pull_request:44010", "right": "pull_request:44129", "accept": false, - "reason": "SqueezeBert and SpeechT5 are different code paths and likely different output signatures. Shared pattern alone is too weak to treat them as duplicates." + "reason": "Both relate to output capture/output tracing, but SqueezeBert and SpeechT5 are separate model changes; the SpeechT5 PR also includes its own cache/output signature adjustments." } ] }, @@ -24128,274 +24294,268 @@ "pull_request:44028|pull_request:44030", "pull_request:44029|pull_request:44030" ], - "prepared_review_unit_hash": "9f218fb860e58bc721a7f14900105e00292c8e0eee4a8474865c2833f4f9aab6", + "prepared_review_unit_hash": "7b141b41d230a881863fc2b9ac27b4ee34f1e247626620a30cd061e1defcbedd", "component_budget": { "node_count": 13, "item_count": 13, "soft_pair_count": 40, - "serialized_chars": 40177, - "estimated_input_tokens": 10045, - "estimated_eval_tokens": 20346 + "serialized_chars": 40183, + "estimated_input_tokens": 10046, + "estimated_eval_tokens": 20348 }, "budget": { "node_count": 13, "item_count": 13, "soft_pair_count": 40, - "serialized_chars": 40177, - "estimated_input_tokens": 10045, - "estimated_eval_tokens": 20346 + "serialized_chars": 40183, + "estimated_input_tokens": 10046, + "estimated_eval_tokens": 20348 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "This cluster is a set of per-model output-tracing refactors across different Transformers models. They share a common pattern, but the PRs usually touch distinct code paths, so they should not be deduped together.", - "confidence": 0.92, + "summary": "These PRs are all model-specific refactors around output tracing/can_return_tuple, but they target different files and forward paths, so they are not duplicates of one another.", + "confidence": 0.97, "canonical_issue_reason": null, - "canonical_pr_reason": "No issue is present. As a representative PR, 44025 is the broadest of the set and clearly shows the core output-tracing refactor pattern across closely related depth models.", + "canonical_pr_reason": null, "best_issue_reason": null, - "best_pr_reason": "44025 is the best single exemplar for the cluster: it is a concrete output-tracing refactor, spans multiple related files, and is representative of the overall theme without being tied to a narrower one-file cleanup.", + "best_pr_reason": "PR 44025 is the most representative anchor for the cluster because it is a fairly complete example of the shared output-tracing refactor pattern and spans two related model files, but it still does not subsume the other PRs as a duplicate.", "soft_edge_verdicts": [ { "left": "pull_request:43998", "right": "pull_request:44002", "accept": false, - "reason": "Same refactor theme, but different model implementations and different forward paths; not the same underlying change." + "reason": "Same refactor theme, but different model implementations and code paths; not the same concrete change." }, { "left": "pull_request:43998", "right": "pull_request:44000", "accept": false, - "reason": "Both are output-tracing cleanups, but they target unrelated model code paths." + "reason": "Both are output-tracing refactors, but they touch different models and forward paths." }, { "left": "pull_request:43998", "right": "pull_request:44001", "accept": false, - "reason": "Parallel API cleanup in a different model family; not a duplicate bug or change." + "reason": "Same broad pattern, but separate model code paths; not plausibly one PR." }, { "left": "pull_request:43998", "right": "pull_request:44003", "accept": false, - "reason": "Different architectures and files; overlap is only the generic refactor pattern." + "reason": "Different architectures and files; only the refactor style matches." }, { "left": "pull_request:43998", "right": "pull_request:44029", "accept": false, - "reason": "Same helper adoption theme, but unrelated model code paths." + "reason": "Different model family and code path; not the same underlying fix." }, { "left": "pull_request:43999", "right": "pull_request:44001", "accept": false, - "reason": "Both refactor outputs, but in different models with separate forward implementations." + "reason": "Both are tracing refactors, but for different models and outputs." }, { "left": "pull_request:43999", "right": "pull_request:44002", "accept": false, - "reason": "Similar output-tracing cleanup, not the same underlying model bug or code path." + "reason": "Different model implementations; no shared concrete bug/change." }, { "left": "pull_request:43999", "right": "pull_request:44025", "accept": false, - "reason": "Different model families and different forward/tracing logic; only the refactor style matches." + "reason": "Same refactor pattern, but distinct model files and behavior." }, { "left": "pull_request:43999", "right": "pull_request:44026", "accept": false, - "reason": "Related wrapper cleanup, but vision-text and vision-encoder-decoder are separate code paths." + "reason": "Different encoder-decoder path versus MobileNet; not duplicate changes." }, { "left": "pull_request:43999", "right": "pull_request:44027", "accept": false, - "reason": "Different encoder-decoder wrapper; not the same concrete change." + "reason": "Different model family and forward logic; only superficially similar." }, { "left": "pull_request:44000", "right": "pull_request:44001", "accept": false, - "reason": "Both are refactors to support tuple/output tracing, but they affect unrelated model implementations." + "reason": "Both are output-tracing refactors, but they affect unrelated model code paths." }, { "left": "pull_request:44000", "right": "pull_request:44002", "accept": false, - "reason": "Same broad cleanup pattern, but distinct model code paths." + "reason": "Same general theme, different model-specific implementation." }, { "left": "pull_request:44000", "right": "pull_request:44003", "accept": false, - "reason": "Different model families; only the naming/theme is shared." + "reason": "Distinct model families and source files; not the same change." }, { "left": "pull_request:44000", "right": "pull_request:44025", "accept": false, - "reason": "Output-tracing refactor in separate model stacks; not mergeable as one concrete fix." + "reason": "Refactor pattern matches, but the underlying code paths are different." }, { "left": "pull_request:44001", "right": "pull_request:44002", "accept": false, - "reason": "Parallel refactors in different models, not one shared bug." + "reason": "Different models and forward paths; not mergeable as one fix." }, { "left": "pull_request:44001", "right": "pull_request:44003", "accept": false, - "reason": "Different architectures and files; similar helper usage is not enough." + "reason": "Separate model implementations; only the tracing refactor is shared." }, { "left": "pull_request:44001", "right": "pull_request:44025", "accept": false, - "reason": "Same general refactor theme, but distinct model implementations." + "reason": "Same broad output-tracing cleanup, but for different models." }, { "left": "pull_request:44002", "right": "pull_request:44029", "accept": false, - "reason": "Not the same code path; just another output-tracing cleanup." + "reason": "Different architectures and code paths; not the same underlying bug/change." }, { "left": "pull_request:44002", "right": "pull_request:44003", "accept": false, - "reason": "Separate model code, no shared concrete defect or change." + "reason": "Only the refactor motif is shared; the actual model logic differs." }, { "left": "pull_request:44003", "right": "pull_request:44028", "accept": false, - "reason": "Falcon/Mamba and SuperPoint are unrelated; similarity is only superficial." + "reason": "Different model families and implementations; not a duplicate change." }, { "left": "pull_request:44003", "right": "pull_request:44004", "accept": false, - "reason": "Different model families and different forward logic; not a duplicate." + "reason": "Mamba/Falcon-Mamba versus CodeGen are unrelated code paths despite similar refactoring." }, { "left": "pull_request:43998", "right": "pull_request:44004", "accept": false, - "reason": "Unrelated model stacks; shared refactor vocabulary only." + "reason": "Different model code paths; the similarity is only in the output-tracing refactor pattern." }, { "left": "pull_request:44001", "right": "pull_request:44004", "accept": false, - "reason": "Different concrete implementations and code paths." + "reason": "Separate model implementations; not the same concrete bug or change." }, { "left": "pull_request:44002", "right": "pull_request:44004", "accept": false, - "reason": "Similar API cleanup, but not the same underlying change." - }, - { - "left": "pull_request:44000", - "right": "pull_request:44004", - "accept": false, - "reason": "Different architectures; overlap is only generic output-tracing refactor work." + "reason": "Same broad cleanup, but different model-specific forward logic." }, { "left": "pull_request:44025", "right": "pull_request:44026", "accept": false, - "reason": "Both are wrapper refactors, but they target different encoder-decoder families." + "reason": "Both are output-tracing refactors, but they apply to different model stacks and code paths." }, { "left": "pull_request:44025", "right": "pull_request:44027", "accept": false, - "reason": "Similar refactor pattern, but distinct models and code paths." + "reason": "Different encoder-decoder models; not the same underlying change." }, { "left": "pull_request:44025", "right": "pull_request:44028", "accept": false, - "reason": "Depth models versus SuperPoint are unrelated concrete changes." + "reason": "Same pattern, but SuperPoint and DepthAnything are unrelated implementations." }, { "left": "pull_request:44025", "right": "pull_request:44029", "accept": false, - "reason": "Different model implementations; same helper usage does not make them duplicates." + "reason": "Different model families and forward paths; not duplicate PRs." }, { "left": "pull_request:44025", "right": "pull_request:44030", "accept": false, - "reason": "Different model families and forward paths." + "reason": "Different model-specific tracing changes; not one mergeable fix." }, { "left": "pull_request:44026", "right": "pull_request:44029", "accept": false, - "reason": "Different wrapper/model code paths; not the same bug or change." + "reason": "Different model families and code paths; only the refactor theme matches." }, { "left": "pull_request:44026", "right": "pull_request:44027", "accept": false, - "reason": "Vision and speech encoder-decoder wrappers are separate implementations, not one shared code-path fix." + "reason": "Different encoder-decoder implementations; not the same change." }, { "left": "pull_request:44026", "right": "pull_request:44028", "accept": false, - "reason": "Unrelated model stacks; only the refactor motif matches." + "reason": "Distinct model code paths; not a duplicate fix." }, { "left": "pull_request:44026", "right": "pull_request:44030", "accept": false, - "reason": "Different model families and distinct forward implementations." + "reason": "Different architectures and implementation details; same broad pattern only." }, { "left": "pull_request:44027", "right": "pull_request:44028", "accept": false, - "reason": "Separate model code paths; similarity is limited to the refactor pattern." + "reason": "Different model implementations and forward paths; not the same underlying change." }, { "left": "pull_request:44027", "right": "pull_request:44029", "accept": false, - "reason": "Different models and concrete implementations." + "reason": "Separate model families; only the output-tracing refactor is shared." }, { "left": "pull_request:44027", "right": "pull_request:44030", "accept": false, - "reason": "Not the same underlying code-path change." + "reason": "Different code paths and model stacks; not duplicate PRs." }, { "left": "pull_request:44028", "right": "pull_request:44029", "accept": false, - "reason": "Different architectures and files; only a generic tracing refactor is shared." + "reason": "Unrelated models and implementations; not the same fix." }, { "left": "pull_request:44028", "right": "pull_request:44030", "accept": false, - "reason": "SuperPoint and DPR are unrelated code paths." + "reason": "Different model-specific refactors; no shared concrete bug." }, { "left": "pull_request:44029", "right": "pull_request:44030", "accept": false, - "reason": "Different model families; shared helper adoption is not enough to merge them as duplicates." + "reason": "Different architectures and forward paths; only the refactor pattern is common." } ] }, @@ -24425,12 +24585,12 @@ "soft_pairs": [ "pull_request:44662|pull_request:45401" ], - "prepared_review_unit_hash": "df69f6a7c8a7d4bd2beab2c7ba80b9be931907c7b4bbc0ba2b54297fe5ca639f", + "prepared_review_unit_hash": "3c1c968d7a6051791b1fdb57461c0e4df85bef375e0e79f46e0c11b289ad00ef", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 6551, + "serialized_chars": 6552, "estimated_input_tokens": 1638, "estimated_eval_tokens": 3532 }, @@ -24438,7 +24598,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 6551, + "serialized_chars": 6552, "estimated_input_tokens": 1638, "estimated_eval_tokens": 3532 }, @@ -24447,18 +24607,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are two unrelated model-implementation PRs: one adds PenguinVL, the other adds Voxtral TTS. They share some transformer registry/docs scaffolding, but they target different models and codepaths, so they should not be clustered as duplicates.", + "summary": "Both items are model-addition PRs that touch common Transformers registration/docs code, but they implement different models and different code paths: PenguinVL vs Voxtral TTS. They should not be deduplicated.", "confidence": 0.98, "canonical_issue_reason": null, - "canonical_pr_reason": null, + "canonical_pr_reason": "PR #44662 is the better representative: it is a more mature, non-draft model implementation with broader review activity, while #45401 is a draft and explicitly tied to a specific support request.", "best_issue_reason": null, - "best_pr_reason": null, + "best_pr_reason": "PR #44662 is the strongest standalone PR in the set because it contains a full model implementation with tests and documentation, whereas #45401 is narrower, draft, and targeted at a specific Voxtral TTS issue.", "soft_edge_verdicts": [ { "left": "pull_request:44662", "right": "pull_request:45401", "accept": false, - "reason": "Different underlying changes for different models (PenguinVL vs Voxtral TTS); shared files are generic registration/docs updates, not the same concrete fix or feature." + "reason": "Reject: these are different model integrations (PenguinVL vs Voxtral TTS). Overlap is limited to shared Transformers plumbing and docs registration, not the same underlying bug or change." } ] }, @@ -24488,12 +24648,12 @@ "soft_pairs": [ "pull_request:44965|pull_request:46845" ], - "prepared_review_unit_hash": "4cd5299058da650dc45266c2f99b067c61fda6a28caa535058849da780344b9c", + "prepared_review_unit_hash": "8aeb0d137a7a1cc533e3f9d0296ec7113e58328c8c9f859252e37885529ee134", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4219, + "serialized_chars": 4220, "estimated_input_tokens": 1055, "estimated_eval_tokens": 2366 }, @@ -24501,7 +24661,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4219, + "serialized_chars": 4220, "estimated_input_tokens": 1055, "estimated_eval_tokens": 2366 }, @@ -24510,18 +24670,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These two pull requests are not duplicates: #44965 is a workflow edit to change PRs to draft, while #46845 tweaks CI branching and includes an unrelated code change in modeling_utils.py.", - "confidence": 0.97, + "summary": "These two pull requests are not duplicates: one is about changing a PR documentation workflow to convert PRs to draft, while the other alters CI branch routing and makes unrelated edits in model code/tests. They do not fix the same concrete problem and would not plausibly merge into one PR.", + "confidence": 0.96, "canonical_issue_reason": null, - "canonical_pr_reason": "No canonical PR: #44965 and #46845 are different changes touching different code paths and should not be clustered as one duplicate change.", + "canonical_pr_reason": null, "best_issue_reason": null, - "best_pr_reason": "If forced to pick one representative, #46845 is the more substantial multi-file PR, but it is still not a good canonical match for #44965.", + "best_pr_reason": "PR 46845 is the more coherent and substantive change of the two, with a clear CI-related intent; PR 44965 looks like an unrelated workflow experiment and is not a better canonical representative.", "soft_edge_verdicts": [ { "left": "pull_request:44965", "right": "pull_request:46845", "accept": false, - "reason": "Different underlying changes: #44965 modifies a docs workflow to convert PRs to draft, while #46845 changes CI routing and adds an unrelated modeling_utils.py edit. They would not plausibly merge as one PR." + "reason": "Different changes with different goals: draft-conversion workflow vs CI branch/config tweaks plus unrelated code edits. They do not appear to address the same underlying bug or change." } ] }, @@ -24551,12 +24711,12 @@ "soft_pairs": [ "pull_request:44979|pull_request:45363" ], - "prepared_review_unit_hash": "1bd10efb07751dc2b78a85f2497ad47bc9b3407fcd2c6ea826e6f1c1a1b154d9", + "prepared_review_unit_hash": "23d13e5ce55be737676cce267bb69eb01557697a8a4b2be58a285754ad2c767f", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4408, + "serialized_chars": 4406, "estimated_input_tokens": 1102, "estimated_eval_tokens": 2460 }, @@ -24564,7 +24724,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4408, + "serialized_chars": 4406, "estimated_input_tokens": 1102, "estimated_eval_tokens": 2460 }, @@ -24573,18 +24733,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Related fusion work, but not duplicates: PR 44979 introduces a generic Module Fusion API, while PR 45363 wires a specific n-to-1 kernel-fusion path through KernelConfig and integrations. They target different layers and wouldn\u2019t be merged as one PR.", - "confidence": 0.9, + "summary": "These are related fusion-feature PRs, but they target different layers: one introduces a generic module-fusion API, while the other wires up n-to-1 kernel fusion through `KernelConfig` and model loading. They are not the same concrete change and should not be deduped.", + "confidence": 0.86, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 44979 is the better canonical representative because it defines the underlying module-fusion API; PR 45363 is a narrower, dependent application of that idea.", + "canonical_pr_reason": "PR #45363 is the more concrete, user-facing implementation: it threads kernel fusion through `KernelConfig`, `modeling_utils`, and integration code, making it the better representative of the cluster.", "best_issue_reason": null, - "best_pr_reason": "PR 44979 is the strongest standalone representative of the cluster since it introduces the core abstraction rather than one concrete fusion integration.", + "best_pr_reason": "PR #45363 is the strongest standalone candidate because it defines the actual end-to-end fusion behavior, whereas PR #44979 is mainly a lower-level API scaffold.", "soft_edge_verdicts": [ { "left": "pull_request:44979", "right": "pull_request:45363", "accept": false, - "reason": "Different change scope and code path: 44979 adds the reusable module-fusion API, while 45363 implements kernel-config-based n-to-1 fusion plumbing and example usage. Related, but not the same underlying fix/change." + "reason": "Both concern fusion, but #44979 adds a generic module-fusion API and tests, while #45363 implements n-to-1 kernel fusion via `KernelConfig` across loading/integration paths. Different concrete code paths, so not a duplicate." } ] }, @@ -24614,12 +24774,12 @@ "soft_pairs": [ "pull_request:45086|pull_request:45317" ], - "prepared_review_unit_hash": "18da8fdd1808986f2384e65fac09efae50b5f462d83c3502270cd34a04522ceb", + "prepared_review_unit_hash": "e57a0c6f49d394c5745f777e63c3168e8fc6a0a513035b7f892cf30b7c104d8d", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4414, + "serialized_chars": 4413, "estimated_input_tokens": 1104, "estimated_eval_tokens": 2464 }, @@ -24627,7 +24787,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4414, + "serialized_chars": 4413, "estimated_input_tokens": 1104, "estimated_eval_tokens": 2464 }, @@ -24636,18 +24796,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Both PRs fix the same AttributeError in `_patch_mistral_regex` by switching to `tokenizer.pre_tokenizer`; 45317 is the more complete version because it adds tests.", + "summary": "Both pull requests address the same concrete bug in `_patch_mistral_regex`: an `AttributeError` caused by using `tokenizer.backend_tokenizer.pre_tokenizer` instead of `tokenizer.pre_tokenizer`. The second PR is a more complete version with tests.", "confidence": 0.98, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 45317 is the best canonical representative: it contains the same code fix as 45086 and adds test coverage for the bug.", + "canonical_pr_reason": "PR 45317 is the stronger canonical PR because it fixes the same bug as 45086 but also adds test coverage, making it the more complete and reviewable patch.", "best_issue_reason": null, - "best_pr_reason": "45317 is the strongest PR for the cluster because it addresses the same concrete failure mode and includes validation, making it the most merge-ready candidate.", + "best_pr_reason": "PR 45317 is the best PR choice because it fully addresses the bug and includes tests, while 45086 is just a smaller variant of the same fix.", "soft_edge_verdicts": [ { "left": "pull_request:45086", "right": "pull_request:45317", "accept": true, - "reason": "Same underlying AttributeError fix in the same function and file; the second PR is essentially the first plus tests, so they are mergeable as one change." + "reason": "Same underlying AttributeError in the same code path (`_patch_mistral_regex`), same file, same fix direction, and they could plausibly be merged into one PR; 45317 simply adds tests." } ] }, @@ -24677,40 +24837,40 @@ "soft_pairs": [ "pull_request:45115|pull_request:45569" ], - "prepared_review_unit_hash": "a41fe8e63ac7ee4c332ad1a5d41047e62726ca03e9e4a67a03e97ee5ec6eddb7", + "prepared_review_unit_hash": "541843e562fa32b789ea91882acd66c9659cc35cc6aaeeac3e1c45c0de9bbdd2", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 6065, - "estimated_input_tokens": 1517, - "estimated_eval_tokens": 3290 + "serialized_chars": 6064, + "estimated_input_tokens": 1516, + "estimated_eval_tokens": 3288 }, "budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 6065, - "estimated_input_tokens": 1517, - "estimated_eval_tokens": 3290 + "serialized_chars": 6064, + "estimated_input_tokens": 1516, + "estimated_eval_tokens": 3288 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These PRs are related to NemotronH work, but they target different scopes: one is a refactor/inheritance and conversion-mapping update, while the other splits the architecture into new dense/sparse model families plus docs and auto-mapping changes. They should not be deduplicated.", - "confidence": 0.89, + "summary": "Two open NemotronH-related PRs overlap on model registry/config files, but they target different changes: one is a refactor to inherit GraniteMoeHybrid-compatible behavior, the other is a broader model-family split adding NemotronH dense/sparse docs and implementations.", + "confidence": 0.91, "canonical_issue_reason": null, - "canonical_pr_reason": "PR #45569 is the broader and more representative change for the NemotronH model-family restructuring, covering the new dense/sparse variants, docs, auto mappings, and related model files.", + "canonical_pr_reason": "PR #45569 is the broader, more representative NemotronH change set: it introduces the dense/sparse model family, updates docs, and touches the central auto/model wiring, making it the better cluster representative.", "best_issue_reason": null, - "best_pr_reason": "PR #45569 is the best overall representative because it captures the larger architectural split and new public model surface, whereas #45115 is a narrower refactor focused on inheritance and conversion mapping.", + "best_pr_reason": "PR #45569 is the best overall candidate because it covers the larger concrete feature/change surface and appears closer to the final intended architecture than the narrower refactor in #45115.", "soft_edge_verdicts": [ { "left": "pull_request:45115", "right": "pull_request:45569", "accept": false, - "reason": "Not the same underlying change. #45115 is a targeted refactor around NemotronH inheritance and conversion mapping; #45569 introduces separate NemotronH dense/sparse model packages, docs, and auto-mapping updates. They overlap in the same subsystem but would not plausibly be merged as one PR." + "reason": "They share NemotronH infrastructure files, but the actual changes differ: #45115 is a conversion/refactor toward GraniteMoeHybrid-compatible weight naming, while #45569 splits NemotronH into dense/sparse variants and adds docs/model registry updates. Same subsystem, not the same underlying change." } ] }, @@ -24740,7 +24900,7 @@ "soft_pairs": [ "pull_request:45332|pull_request:45333" ], - "prepared_review_unit_hash": "10e40d33e2f59ca2773f71422205f29a824d5706a6b3c2d7d82985509f960409", + "prepared_review_unit_hash": "e97b6d069fab3669d692330d4fc4feba0413a0d96ae828f92ee3ddd752af9f6e", "component_budget": { "node_count": 2, "item_count": 2, @@ -24762,18 +24922,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Two closely related PRs for heterogeneous model support, but one is a broader end-to-end implementation and the other is a narrower config-only slice. They overlap heavily in docs/configuration code, yet they are not the same concrete change set.", - "confidence": 0.88, + "summary": "Two overlapping PRs in the heterogeneous-modeling area, but they are not the same change: one is an end-to-end heterogeneous model support PR, the other is a configuration-only subset.", + "confidence": 0.94, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 45332 is the better canonical representative because it covers the full heterogeneous-model feature stack: per-layer config plus modeling, cache/masking updates, integrations, and tests. PR 45333 is a narrower config-only subset.", + "canonical_pr_reason": "PR #45332 is the broader, more complete change: it adds heterogeneous modeling support plus cache, masking, integration, and test updates, so it best represents the cluster.", "best_issue_reason": null, - "best_pr_reason": "PR 45332 is the strongest overall match for the cluster\u2019s theme and the most complete implementation of the underlying feature.", + "best_pr_reason": "PR #45332 is the best representative because it covers the full feature scope, while #45333 is limited to per-layer configuration utilities and docs.", "soft_edge_verdicts": [ { "left": "pull_request:45332", "right": "pull_request:45333", "accept": false, - "reason": "Related but not duplicates: 45333 is focused on per-layer configuration only, while 45332 also changes modeling, cache, masking, and integration code. They share files and topic, but they are different scope and could reasonably be separate PRs." + "reason": "They share docs and config files, but #45332 implements full heterogeneous model support (modeling, masking, cache, integrations) whereas #45333 only adds configuration support. Different concrete changes, so not a duplicate PR pair." } ] }, @@ -24803,7 +24963,7 @@ "soft_pairs": [ "pull_request:45415|pull_request:45425" ], - "prepared_review_unit_hash": "a9de0c9f8f3666b43d5468e7ec8f2ca9835fbc6f2a1736eea32b8897d0a9eb9f", + "prepared_review_unit_hash": "2093ad51836a73cd385e362d8b446b7459a937d953c8c29e2f0ddc40902e28f5", "component_budget": { "node_count": 2, "item_count": 2, @@ -24825,18 +24985,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Both items are type-checking/typing-focused PRs in the transformers codebase, but they address different changes: one is a broad sweep over many `src/transformers/*py` files and mypy/pyproject config, while the other adds typing protocols and related adjustments around `modeling_utils` and friends. The overlap in filenames reflects shared typing infrastructure, not the same underlying bug or fix.", - "confidence": 0.91, + "summary": "These are two different typing-focused PRs: one is a broad mechanical type-check cleanup across many transformer modules, and the other adds new typing protocols/aliases for modeling_utils, PEFT, and related helpers.", + "confidence": 0.9, "canonical_issue_reason": null, - "canonical_pr_reason": "No clear duplicate PR here: the scopes differ materially, with one being a broad repo-wide typing cleanup and the other introducing specific typing aliases/protocols for internal interfaces.", + "canonical_pr_reason": "PR 45425 is the more coherent, self-contained change set with concrete typing API additions; PR 45415 is a broader mechanical enablement pass across many files.", "best_issue_reason": null, - "best_pr_reason": "PR 45425 is the better representative of the cluster because it is more focused on a concrete typing-interface change, whereas 45415 is a wider mechanical typing pass.", + "best_pr_reason": "PR 45425 is the best representative of the cluster because it makes targeted typing/interface changes rather than a wide lint/config sweep.", "soft_edge_verdicts": [ { "left": "pull_request:45415", "right": "pull_request:45425", "accept": false, - "reason": "Shared typing-related files are not enough to make these the same change; the PRs target different typing problems and would not plausibly be merged as one fix." + "reason": "Both are about typing, but they address different underlying changes: 45415 is broad type-check cleanup and config relaxation, while 45425 introduces specific typing protocols/aliases for model and PEFT integration. They would not plausibly merge into one PR without becoming a much larger unrelated bundle." } ] }, @@ -24866,12 +25026,12 @@ "soft_pairs": [ "pull_request:45548|pull_request:45662" ], - "prepared_review_unit_hash": "997fefa9eb85a0783848f17675afa77db7604efba53b322de68237abc357fd61", + "prepared_review_unit_hash": "c15cf677d62616d12c453f71657d0160dc1a31012ab1ebbfeadfd4a3bfbd0a9d", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4508, + "serialized_chars": 4507, "estimated_input_tokens": 1127, "estimated_eval_tokens": 2510 }, @@ -24879,7 +25039,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4508, + "serialized_chars": 4507, "estimated_input_tokens": 1127, "estimated_eval_tokens": 2510 }, @@ -24888,18 +25048,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Two distinct expert-parallelism fixes: one targets DeepSpeed ZeRO-3/accelerate loading and serialized distributed_config handling, the other targets FSDP2 rank-0 broadcast overwriting expert parameters. Related area, but different bugs and fixes.", - "confidence": 0.95, + "summary": "Both PRs concern expert parallelism, but they fix different failure modes: one is about EP config/ZeRO-3 loading via accelerate, the other about FSDP2 expert parameters being overwritten during load. They share a subsystem, not the same bug or change.", + "confidence": 0.94, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 45662 is the better representative of its own bug: it has the more concrete failure mode (experts overwritten by rank-0 broadcast) and the broader code-path fix.", + "canonical_pr_reason": "No single canonical PR for duplication: these are separate EP loading bugs with different triggers and fixes (DeepSpeed ZeRO-3/config serialization vs FSDP2 rank-0 broadcast overwrite).", "best_issue_reason": null, - "best_pr_reason": "PR 45662 is the strongest standalone PR in this cluster because it addresses a specific, reproducible expert-loading corruption path with coordinated changes across loading and parallelism code.", + "best_pr_reason": "45662 is the better representative of the cluster because it fixes a concrete, narrowly defined EP loading corruption path with explicit code-path changes in the loader and MoE handling.", "soft_edge_verdicts": [ { "left": "pull_request:45548", "right": "pull_request:45662", "accept": false, - "reason": "Both are EP-related, but they fix different problems: 45548 is about accelerate/ZeRO-3 loading and config serialization, while 45662 prevents expert weights from being overwritten during FSDP2 rank-0 broadcast." + "reason": "Reject: same broad EP area, but the PRs target different concrete bugs and loading paths; they would not plausibly merge into one change without conflating separate fixes." } ] }, @@ -24929,40 +25089,40 @@ "soft_pairs": [ "pull_request:45639|pull_request:45645" ], - "prepared_review_unit_hash": "7757c23db54c97af9b40f75b4a78ddbb0118bf3b3c93f0570174853c61e73df6", + "prepared_review_unit_hash": "76c4d9b1ab6f9f7c1efeb5ba69e045777ca380f7062e20e01c5c9e246460b25a", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4673, - "estimated_input_tokens": 1169, - "estimated_eval_tokens": 2594 + "serialized_chars": 4672, + "estimated_input_tokens": 1168, + "estimated_eval_tokens": 2592 }, "budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4673, - "estimated_input_tokens": 1169, - "estimated_eval_tokens": 2594 + "serialized_chars": 4672, + "estimated_input_tokens": 1168, + "estimated_eval_tokens": 2592 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Both PRs fix the same captured_info/xdist collision problem in transformers testing utilities; 45645 is the more complete superset because it includes the core file-splitting fix plus CI log aggregation changes.", - "confidence": 0.87, + "summary": "Both PRs target the same xdist-related captured_info logging collision around the testing debug artifacts, but PR 45645 is the broader, more complete fix because it also updates CI log collection to handle multiple per-worker files.", + "confidence": 0.94, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 45645 is the better canonical representative because it contains the main xdist-safe debug-log fix and also updates CI to collect multiple captured_info files, making it the broader end-to-end solution.", + "canonical_pr_reason": "PR 45645 is the better canonical PR because it covers the underlying xdist-safe artifact handling end-to-end: the runtime filename fix plus the CI workflow update needed to read multiple captured_info files.", "best_issue_reason": null, - "best_pr_reason": "PR 45645 is the best overall PR for the cluster since it addresses the underlying bug and the downstream artifact handling, whereas 45639 only covers the core library-side change.", + "best_pr_reason": "PR 45645 is the strongest representative of the cluster since it addresses the same concrete failure mode as 45639 and additionally preserves the debug logs in CI, making it the more complete fix.", "soft_edge_verdicts": [ { "left": "pull_request:45639", "right": "pull_request:45645", "accept": true, - "reason": "Same underlying xdist-safe captured_info bug in testing_utils; 45645 extends the exact same fix with workflow support, so they are mergeable as one change." + "reason": "Both PRs fix the same captured_info xdist collision problem in the testing debug logging path; 45645 appears to subsume 45639 and could plausibly be merged as one combined change." } ] }, @@ -24992,7 +25152,7 @@ "soft_pairs": [ "pull_request:45720|pull_request:46259" ], - "prepared_review_unit_hash": "afe027662d53836aa1d6f25219dd9ba1297467d6041349d3b33adf406055b761", + "prepared_review_unit_hash": "2a4cb0f8d4a50e298434bd6139d58e5df35d189559395a962a3339827a20d2c8", "component_budget": { "node_count": 2, "item_count": 2, @@ -25014,18 +25174,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are two unrelated linter-rule PRs that happen to touch overlapping transformer model files. They do not fix the same concrete bug or implement the same change, so they should not be merged as duplicates.", - "confidence": 0.97, + "summary": "The two PRs overlap in several transformer model files, but they address different linter rules and different behaviors: 45720 updates forward return handling for TRF003, while 46259 changes weight initialization logic for TRF018. They are not duplicates.", + "confidence": 0.98, "canonical_issue_reason": null, - "canonical_pr_reason": "PR #46259 is the broader, more central modeling-rule change and is the better representative of this cluster, but it is not a duplicate of #45720.", + "canonical_pr_reason": "No canonical duplicate PR: 45720 targets forward/return_dict behavior, while 46259 targets _init_weights initialization behavior. The overlapping filenames are incidental.", "best_issue_reason": null, - "best_pr_reason": "PR #46259 is the strongest representative because it covers a larger linter-rule implementation across many modeling files; #45720 is a narrower, different rule change.", + "best_pr_reason": "46259 is the broader and more substantial PR overall, but it is not a duplicate of 45720.", "soft_edge_verdicts": [ { "left": "pull_request:45720", "right": "pull_request:46259", "accept": false, - "reason": "Reject: the PRs target different lint rules and different code changes. #45720 implements part of TRF003 by adding `can_return_tuple`/forward-signature changes, while #46259 adds TRF018 modeling-rule updates around weight initialization. Shared filenames are incidental to broad model coverage, not evidence of the same fix." + "reason": "Different underlying changes and code paths: TRF003 modifies forward return handling, while TRF018 modifies weight initialization. Shared model files do not make them the same fix." } ] }, @@ -25121,40 +25281,40 @@ "soft_pairs": [ "pull_request:45932|pull_request:46100" ], - "prepared_review_unit_hash": "afe9fb6324ade85d1ac0078584d2dcd3f042fd39d0eebaf1524a7a77f22873c0", + "prepared_review_unit_hash": "c2cee72e260cfd846889ce2d4e492dfed15202861e1173872863eef2542adecc", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 5160, - "estimated_input_tokens": 1290, - "estimated_eval_tokens": 2836 + "serialized_chars": 5161, + "estimated_input_tokens": 1291, + "estimated_eval_tokens": 2838 }, "budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 5160, - "estimated_input_tokens": 1290, - "estimated_eval_tokens": 2836 + "serialized_chars": 5161, + "estimated_input_tokens": 1291, + "estimated_eval_tokens": 2838 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are both open, unrelated PRs with broad test/maintenance themes, but they do not fix the same underlying problem and should not be clustered as duplicates.", - "confidence": 0.95, + "summary": "These are both open PRs touching testing-related areas, but they address different changes: one is a mixed CI/model patch, the other is a focused test-migration PR. They are not the same underlying change.", + "confidence": 0.94, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 45932 is the more representative PR in the cluster because it contains substantive model/config and CI workflow changes, whereas PR 46100 is primarily a test migration/addition PR.", + "canonical_pr_reason": "pull_request:46100 is the more coherent and self-contained change set; pull_request:45932 mixes CI secrets and multiple model/config edits without a single clear theme.", "best_issue_reason": null, - "best_pr_reason": "PR 45932 is the best single representative of the cluster because it is the broader, more central code change; PR 46100 is narrower and mostly test infrastructure.", + "best_pr_reason": "pull_request:46100 is the best representative of the cluster because it has a narrowly defined purpose (test migration) and a clearer mergeable scope than pull_request:45932.", "soft_edge_verdicts": [ { "left": "pull_request:45932", "right": "pull_request:46100", "accept": false, - "reason": "Reject: the PRs address different changes. 45932 tweaks CI secrets and model/config internals across multiple files; 46100 adds kernel tests and a DDP test config for AltCLIP/ViT migration. They do not appear to fix the same code-path bug or be plausibly merged into one PR." + "reason": "Both mention testing, but 45932 is a mixed grafana/CI/model patch while 46100 is a dedicated test migration; they do not appear to fix the same concrete problem and would not plausibly be merged as one PR." } ] }, @@ -25184,12 +25344,12 @@ "soft_pairs": [ "pull_request:46126|pull_request:46269" ], - "prepared_review_unit_hash": "ae1de4a7a58d9f9d2de84fadb5134320d1c8e3f945b73f6451e69d0f7d1038fc", + "prepared_review_unit_hash": "73ea664d7038f01ef1181390ccf390bd83e76aa9ce60c89ba2ad17b02210e6d2", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 5525, + "serialized_chars": 5526, "estimated_input_tokens": 1382, "estimated_eval_tokens": 3020 }, @@ -25197,7 +25357,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 5525, + "serialized_chars": 5526, "estimated_input_tokens": 1382, "estimated_eval_tokens": 3020 }, @@ -25206,18 +25366,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These two pull requests overlap in distributed/MoE code, but they are not the same change: #46126 is a focused EP + Trainer integration patch, while #46269 is a much broader distributed branch/base refactor touching many unrelated areas. The shared filenames suggest branch lineage or subsystem overlap, not a mergeable duplicate PR.", - "confidence": 0.93, + "summary": "These are related distributed-training pull requests, but not the same underlying change: one is a focused EP + Trainer integration, while the other is a broad distributed branch base covering many subsystems. The shared files are not enough to treat them as duplicates.", + "confidence": 0.83, "canonical_issue_reason": null, - "canonical_pr_reason": "#46269 is the broader umbrella change and best represents the cluster\u2019s distributed branch work; #46126 is a narrower follow-up focused on EP + Trainer integration.", + "canonical_pr_reason": "PR 46269 is the broader umbrella branch base touching many distributed components, so it is the better canonical representative of this cluster than the narrower EP/Trainer integration PR.", "best_issue_reason": null, - "best_pr_reason": "#46269 is the best representative PR because it spans the larger distributed foundation and includes the overlapping files, whereas #46126 is a specific feature integration on top of that base.", + "best_pr_reason": "PR 46269 is the best representative because it is the more comprehensive distributed baseline and subsumes a wider set of related changes.", "soft_edge_verdicts": [ { "left": "pull_request:46126", "right": "pull_request:46269", "accept": false, - "reason": "Shared distributed filenames are not enough to treat these as duplicates. #46126 is a targeted EP/Trainer integration change, while #46269 is a broad distributed branch base touching many additional components; they are not the same concrete code-path fix and would not plausibly be merged as one PR." + "reason": "Reject: they overlap in distributed infrastructure files, but 46126 is a specific EP + Trainer integration change while 46269 is a broad distributed branch base with many unrelated edits. They do not look like the same concrete fix and would not plausibly be merged as one PR." } ] }, @@ -25285,232 +25445,220 @@ "pull_request:46504|pull_request:46716", "pull_request:46432|pull_request:46716" ], - "prepared_review_unit_hash": "b6374718386e3a67f7e036e82b666afdf24c7de2052592de2b4d6bb1dad95da1", + "prepared_review_unit_hash": "42358afa722ee686b323565b3f3ae86afef7a8cb9c2165b1caf7c1167499d2b4", "component_budget": { "node_count": 10, "item_count": 10, "soft_pair_count": 31, - "serialized_chars": 26679, - "estimated_input_tokens": 6670, - "estimated_eval_tokens": 13596 + "serialized_chars": 26674, + "estimated_input_tokens": 6669, + "estimated_eval_tokens": 13594 }, "budget": { "node_count": 10, "item_count": 10, "soft_pair_count": 31, - "serialized_chars": 26679, - "estimated_input_tokens": 6670, - "estimated_eval_tokens": 13596 + "serialized_chars": 26674, + "estimated_input_tokens": 6669, + "estimated_eval_tokens": 13594 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These PRs are mostly unrelated workflow maintenance changes. The only apparent subgroup is several updates to pr-ci-security-gate.yml, but they address different security/permission behaviors and should not be treated as duplicates.", - "confidence": 0.93, + "summary": "No true duplicates here. The PRs fall into separate workflow-maintenance changes, and the pr-ci-security-gate.yml edits are related in theme but implement different security-hardening fixes, so I would not merge them as duplicates.", + "confidence": 0.74, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 46554 is the strongest representative of the only related subgroup: it makes the most substantive functional change to pr-ci-security-gate.yml by adding real security checks, whereas the nearby PRs are narrower permission or wiring tweaks.", + "canonical_pr_reason": "If one PR must represent the related security-gate subgroup, 46554 is the best candidate because it makes the most concrete self-contained code-path change (checkout plus diff-based checks), while the others are separate workflow adjustments.", "best_issue_reason": null, - "best_pr_reason": "PR 46554 is the best overall representative because it is the most concrete and central change in the cluster, and the other PRs are either unrelated workflow edits or separate follow-up adjustments in the same file.", + "best_pr_reason": "46554 is the strongest standalone PR in the only partially related subgroup: it addresses a specific security-gate behavior with an explicit logic change and clearer implementation than the follow-up permission/secret/reusable-workflow tweaks.", "soft_edge_verdicts": [ { "left": "pull_request:46138", "right": "pull_request:46554", "accept": false, - "reason": "Different workflows and different code paths: self-comment CI stdin handling vs pr-ci-security-gate security logic." + "reason": "Different workflow areas: self-comment CI command piping vs pr-ci security gate logic." }, { "left": "pull_request:46138", "right": "pull_request:46558", "accept": false, - "reason": "Unrelated workflow changes; one updates self-comment CI command input, the other changes PR CI security-gate permissions/approval flow." + "reason": "Different workflow changes in unrelated files and code paths." }, { "left": "pull_request:46138", "right": "pull_request:46563", "accept": false, - "reason": "Different PR CI workflows and different intents; no shared underlying bug or change." + "reason": "Unrelated workflow update; no shared concrete bug or fix." }, { "left": "pull_request:46138", "right": "pull_request:46716", "accept": false, - "reason": "Different workflows and different fixes; one is self-comment CI, the other is PR CI security-gate secrets handling." + "reason": "Self-comment CI update vs pr-ci security-gate secrets handling; not the same change." }, { "left": "pull_request:46138", "right": "pull_request:46504", "accept": false, - "reason": "Both are workflow maintenance, but they fix different scripts in different files and are not the same underlying change." + "reason": "Different workflow files with different behavior; not a duplicate fix." }, { "left": "pull_request:46432", "right": "pull_request:46503", "accept": false, - "reason": "SSH runner hardening is unrelated to benchmark_v2 dispatch behavior." + "reason": "SSH runner hardening vs benchmark trigger change; separate workflow concerns." }, { "left": "pull_request:46432", "right": "pull_request:46501", "accept": false, - "reason": "Different workflow files and unrelated changes: SSH runner secret handling vs anti-slop permission scoping." + "reason": "SSH runner update and anti-slop permissions change are unrelated." }, { "left": "pull_request:46432", "right": "pull_request:46502", "accept": false, - "reason": "SSH runner hardening is unrelated to assign-reviewers checkout behavior." + "reason": "Different workflows and purposes; no same underlying bug." }, { "left": "pull_request:46432", "right": "pull_request:46554", "accept": false, - "reason": "Different workflows and different code paths; ssh-runner hardening is not the same as PR CI security-gate checks." + "reason": "SSH runner environment cleanup vs security-gate logic; distinct changes." }, { "left": "pull_request:46432", "right": "pull_request:46558", "accept": false, - "reason": "Different workflow files and unrelated security adjustments." - }, - { - "left": "pull_request:46432", - "right": "pull_request:46563", - "accept": false, - "reason": "SSH runner changes and PR CI security-gate reuse changes are distinct fixes." - }, - { - "left": "pull_request:46432", - "right": "pull_request:46716", - "accept": false, - "reason": "Different workflows and unrelated security changes; no common bug/change." + "reason": "SSH runner update vs security-gate permissions/approval tweak; not duplicates." }, { "left": "pull_request:46501", "right": "pull_request:46503", "accept": false, - "reason": "Different workflow files and different behavior changes: anti-slop permissions vs benchmark_v2 trigger conditions." + "reason": "Both are workflow edits, but they change different jobs and behavior." }, { "left": "pull_request:46501", "right": "pull_request:46502", "accept": false, - "reason": "Different workflow files and unrelated fixes; permissions in anti-slop is not the same as checkout ref in assign-reviewers." + "reason": "Anti-slop permissions vs assign-reviewers checkout ref; unrelated fixes." }, { "left": "pull_request:46501", "right": "pull_request:46504", "accept": false, - "reason": "Different workflow files and different mechanics: anti-slop permission scoping vs shell-safe COMMIT_MESSAGE handling." + "reason": "Different workflow files and unrelated logic changes." }, { "left": "pull_request:46501", "right": "pull_request:46554", "accept": false, - "reason": "Same broad CI area, but different workflows and unrelated changes." + "reason": "Anti-slop permission scoping vs security-gate checkout/diff checks; separate issues." }, { "left": "pull_request:46502", "right": "pull_request:46503", "accept": false, - "reason": "Different workflow files and distinct behavior changes; not the same underlying bug." + "reason": "Assign-reviewers checkout behavior vs benchmark trigger condition; different changes." }, { "left": "pull_request:46502", "right": "pull_request:46504", "accept": false, - "reason": "Assign-reviewers checkout ref and build-ci-docker-images message handling are unrelated." + "reason": "Unrelated workflow updates in different files." }, { "left": "pull_request:46502", "right": "pull_request:46554", "accept": false, - "reason": "Different workflows and different problems; no mergeable duplicate relationship." + "reason": "Different workflow code paths; not the same bug or fix." }, { "left": "pull_request:46503", "right": "pull_request:46504", "accept": false, - "reason": "Benchmark dispatch trigger and build-ci image tag parsing are separate workflow changes." + "reason": "Benchmark trigger logic vs build-image commit-message parsing; separate changes." }, { "left": "pull_request:46503", "right": "pull_request:46554", "accept": false, - "reason": "Different workflow files and different concerns; benchmark triggering vs PR CI security gating." + "reason": "Benchmark workflow trigger update vs security-gate enforcement; unrelated." }, { "left": "pull_request:46504", "right": "pull_request:46554", "accept": false, - "reason": "Both touch CI workflows, but one fixes shell parsing in build-ci-docker-images while the other changes PR CI security gating." + "reason": "Build-image tag detection vs pr-ci security-gate logic; not the same change." }, { "left": "pull_request:46504", "right": "pull_request:46558", "accept": false, - "reason": "Different files and different workflow behavior; not the same change." + "reason": "Different workflow logic: commit-message handling vs permissions/approval gating." }, { "left": "pull_request:46554", "right": "pull_request:46558", "accept": false, - "reason": "Same file, but different underlying changes: 46554 adds security checks, while 46558 tightens permissions/approval gating. They could coexist in one file but are not duplicates." + "reason": "Same file, but different concrete fixes: 46554 adds checkout/diff checks; 46558 tightens permissions and approval flow." }, { "left": "pull_request:46554", "right": "pull_request:46563", "accept": false, - "reason": "Same workflow file, but different code paths: actual security checks versus switching to a reusable workflow wrapper." + "reason": "Different fixes in the same workflow area: one changes the security check implementation, the other swaps to a reusable workflow and adds notes." }, { "left": "pull_request:46554", "right": "pull_request:46716", "accept": false, - "reason": "Same file, but different change scope: security checks versus secrets handling on the reusable workflow call." + "reason": "46554 hardens the security gate logic; 46716 changes secrets passing on the reusable workflow. Different code-path problems." }, { "left": "pull_request:46558", "right": "pull_request:46563", "accept": false, - "reason": "Both touch the same workflow file, but they change different behaviors and are not the same underlying fix." + "reason": "Permissions/approval gating vs reusable-workflow substitution; distinct changes." }, { "left": "pull_request:46558", "right": "pull_request:46716", "accept": false, - "reason": "Permissions/approval gating and secrets inheritance are distinct security tweaks, not duplicates." + "reason": "Permissions and approval guard vs secrets inheritance removal; not the same underlying fix." }, { "left": "pull_request:46563", "right": "pull_request:46716", "accept": false, - "reason": "Both are about pr-ci-security-gate.yml, but one switches to a reusable workflow and the other changes secrets passing; different changes." + "reason": "Reusable workflow call/comment change vs secrets handling change; separate PRs." }, { "left": "pull_request:46504", "right": "pull_request:46563", "accept": false, - "reason": "Different workflow files and different concerns; shell-safe env parsing vs PR CI security gate wiring." + "reason": "Build-image workflow change vs pr-ci-security-gate reusable workflow note; unrelated." }, { "left": "pull_request:46432", "right": "pull_request:46563", "accept": false, - "reason": "SSH runner hardening is unrelated to PR CI security-gate workflow reuse." + "reason": "SSH runner hardening vs security-gate reusable-workflow change; different subsystems and fixes." }, { "left": "pull_request:46504", "right": "pull_request:46716", "accept": false, - "reason": "Different files and different fixes; build-ci message parsing is unrelated to PR CI secrets handling." + "reason": "Build-ci docker workflow update vs security-gate secrets change; not duplicates." }, { "left": "pull_request:46432", "right": "pull_request:46716", "accept": false, - "reason": "Different workflows and unrelated security hardening changes." + "reason": "SSH runner update vs security-gate secrets handling; separate changes." } ] }, @@ -25603,12 +25751,12 @@ "soft_pairs": [ "pull_request:46281|pull_request:46890" ], - "prepared_review_unit_hash": "fed8477c05d8e74318f4cf9fb4cf02cff6cf89711041cca613f456abdeccc0af", + "prepared_review_unit_hash": "5cd5b95aae12c0bd33443866e33fbf581bbac5dbf0e70a8267a2b7e8056754f8", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4279, + "serialized_chars": 4278, "estimated_input_tokens": 1070, "estimated_eval_tokens": 2396 }, @@ -25616,7 +25764,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4279, + "serialized_chars": 4278, "estimated_input_tokens": 1070, "estimated_eval_tokens": 2396 }, @@ -25625,18 +25773,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These PRs are thematically similar security hardening changes, but they fix different code paths: one validates shard filenames from a checkpoint index in `hub.py`, the other rejects path traversal in the `transformers_weights` config field in `modeling_utils.py`. They are not duplicates.", + "summary": "Two security-related PRs, but they address different path-traversal entry points in different loader code paths. They are thematically similar, not duplicates.", "confidence": 0.96, "canonical_issue_reason": null, - "canonical_pr_reason": "No single canonical PR: the changes target different loaders, different trusted inputs, and different filenames, so they should remain separate.", + "canonical_pr_reason": "PR #46890 is the better cluster representative because it targets a clearly defined, config-driven path-traversal check in the core checkpoint resolution flow and has a self-contained fix in `modeling_utils.py`.", "best_issue_reason": null, - "best_pr_reason": "Neither PR is a better duplicate representative of the other; both are standalone fixes for distinct path-traversal vectors.", + "best_pr_reason": "PR #46890 is the strongest representative for this cluster: it fixes a concrete unsafe file-path resolution bug, has a clear rejection condition, and is easier to describe as the canonical security hardening change.", "soft_edge_verdicts": [ { "left": "pull_request:46281", "right": "pull_request:46890", "accept": false, - "reason": "Both are security-related path validation fixes, but they address different bugs in different call sites (`weight_map` shard loading vs `transformers_weights` config loading) and would not plausibly be merged into one PR." + "reason": "Different bugs and code paths: #46281 guards sharded-checkpoint filenames from the index `weight_map` in `hub.py`, while #46890 validates `transformers_weights` / explicit checkpoint filenames in `modeling_utils.py`. Same broad path-traversal theme, but not the same underlying fix and not mergeable as one PR." } ] }, @@ -25666,7 +25814,7 @@ "soft_pairs": [ "pull_request:46310|pull_request:47010" ], - "prepared_review_unit_hash": "7581c757eb6aa6d843a95c9fcd396b4cf3242270ba7f156fe1cd6481937ead49", + "prepared_review_unit_hash": "3e01bc3b8ecef7c3a9e1f9cd452c02a0146787169c22b819133f5c18bb1a4f92", "component_budget": { "node_count": 2, "item_count": 2, @@ -25688,18 +25836,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Two open PRs address an off-by-one in sliding_window_mask_function right-window handling; one is a Moonshine-only fix, the other also includes Gemma4. They are closely related but not the same PR change set.", - "confidence": 0.91, + "summary": "Both pull requests fix the same off-by-one bug in the sliding window mask\u2019s right boundary (`-dist <= right_window_size` instead of `<`). PR 47010 is the broader, more representative patch because it applies the fix in both Gemma4 and Moonshine Streaming, while 46310 covers only Moonshine Streaming.", + "confidence": 0.98, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 47010 is the broader, more complete change set: it fixes the same boundary bug in Moonshine streaming and also applies the same correction in Gemma4, making it the better representative of the cluster.", + "canonical_pr_reason": "PR 47010 is the best canonical PR because it addresses the same concrete boundary bug and includes the fuller cross-model fix, making it the most representative version of the change.", "best_issue_reason": null, - "best_pr_reason": "PR 47010 is the strongest candidate overall because it captures the shared off-by-one fix across multiple affected implementations, while PR 46310 is a narrower Moonshine-only subset.", + "best_pr_reason": "PR 47010 is the strongest standalone candidate: it fixes the exact code-path bug and covers all shown affected locations, whereas PR 46310 is a narrower subset.", "soft_edge_verdicts": [ { "left": "pull_request:46310", "right": "pull_request:47010", - "accept": false, - "reason": "Both change the right-window boundary from < to <=, but 47010 also modifies Gemma4 and is not just the same concrete PR; it is a broader multi-file fix rather than a duplicate of 46310." + "accept": true, + "reason": "Accept: both PRs fix the same off-by-one in `sliding_window_mask_function` right-window inclusion on the same code path; 47010 is essentially a superset of 46310, so they plausibly merge into one change." } ] }, @@ -25729,7 +25877,7 @@ "soft_pairs": [ "pull_request:46376|pull_request:46417" ], - "prepared_review_unit_hash": "c40598589d2d24317f5ad8a7fa8b557c358c50ab468c17d55b3dc25e73a5499b", + "prepared_review_unit_hash": "4f4efbb4b257021dde9b3f12a4185f39f438656fa42d5fe00bf406b02d966dbe", "component_budget": { "node_count": 2, "item_count": 2, @@ -25751,18 +25899,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These two pull requests are unrelated: one adds multimodal support to continuous batching, while the other adds the HunYuan VL model and its docs/code. They should not be clustered as duplicates or merged as one change.", + "summary": "The two pull requests are unrelated: one implements multimodal support for continuous batching, while the other adds the HunYuan VL model stack and docs. They do not target the same bug or code change and should not be deduplicated.", "confidence": 0.98, "canonical_issue_reason": null, - "canonical_pr_reason": null, + "canonical_pr_reason": "No clear canonical PR for a duplicate cluster; if forced, PR #46376 is the more infrastructure-focused change, but it is not a duplicate of PR #46417.", "best_issue_reason": null, - "best_pr_reason": null, + "best_pr_reason": "PR #46376 is the better representative of a concrete runtime code-path change; PR #46417 is a broader model-addition/documentation PR and is clearly separate.", "soft_edge_verdicts": [ { "left": "pull_request:46376", "right": "pull_request:46417", "accept": false, - "reason": "Different underlying changes and code paths: continuous batching infrastructure vs. a new multimodal model implementation. They do not appear to be the same bug or feature and would not plausibly be merged into one PR." + "reason": "Different underlying work: #46376 adds multimodal support to continuous batching internals, while #46417 adds the HunYuan VL model, processors, docs, and auto-mapping entries. They do not share a concrete code-path fix and would not plausibly be merged into one PR." } ] }, @@ -25792,7 +25940,7 @@ "soft_pairs": [ "pull_request:46574|pull_request:46580" ], - "prepared_review_unit_hash": "74cc8f237da39f0cdd451392138f6ee9135acde2a147b6a8268d30c94fa266d0", + "prepared_review_unit_hash": "7fb932b4185f683ed2dac3b2780642ef377d33283eb4ebff815e56541ff5a2e9", "component_budget": { "node_count": 2, "item_count": 2, @@ -25814,18 +25962,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These two PRs are both tokenizer-related but fix different bugs in different code paths: CodeLlama decode stripping versus DeepSeek-Coder/LLM tokenizer-class selection. They should not be treated as duplicates or merged into one change.", - "confidence": 0.96, + "summary": "These are two unrelated PRs: #46574 fixes CodeLlama decode stripping leading spaces, while #46580 fixes DeepSeek-Coder/DeepSeek-LLM being assigned the wrong tokenizer class in AutoTokenizer. They do not look like the same underlying bug or change.", + "confidence": 0.99, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 46574 is the more self-contained concrete bug fix: it targets a specific decode-path regression in CodeLlama and includes focused test coverage.", + "canonical_pr_reason": null, "best_issue_reason": null, - "best_pr_reason": "PR 46574 is the best representative of the cluster because it is a narrowly scoped, directly reproducible tokenizer bug fix rather than a broader registry/dispatch adjustment.", + "best_pr_reason": "#46574 is the tighter representative of its bugfix class: it has a single concrete tokenizer-decoding issue, an explicit issue link, and a narrowly scoped code/test change.", "soft_edge_verdicts": [ { "left": "pull_request:46574", "right": "pull_request:46580", "accept": false, - "reason": "Different underlying problems and code paths: one fixes CodeLlama decoder stripping real leading spaces, the other fixes model-to-tokenizer-class routing for DeepSeek checkpoints. They are not plausibly one merged PR." + "reason": "Different failure modes and code paths: one changes CodeLlama decoder setup to preserve leading spaces, the other changes AutoTokenizer checkpoint-prefix routing for DeepSeek models. They are not mergeable as one fix." } ] }, @@ -25855,12 +26003,12 @@ "soft_pairs": [ "pull_request:46847|pull_request:47009" ], - "prepared_review_unit_hash": "89957a77a61e12cee2aab25476b1649244f9ac97dfb0a7d68d130daedd278750", + "prepared_review_unit_hash": "66783a7c52659440be3dd8a54877ab0ee59de25c65133baeffcce0459576c028", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4640, + "serialized_chars": 4639, "estimated_input_tokens": 1160, "estimated_eval_tokens": 2576 }, @@ -25868,7 +26016,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4640, + "serialized_chars": 4639, "estimated_input_tokens": 1160, "estimated_eval_tokens": 2576 }, @@ -25877,18 +26025,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Both pull requests are near-duplicates for the same Qwen3.5 tensor-parallel plan change: they touch the same files, add the same `linear_attn` TP entries, and target the same issue. Treat them as the same underlying PR work item.", + "summary": "Both PRs make the same Qwen3.5 tensor-parallel plan update for linear_attn, touching the same files with the same added mappings and the same tracked issue. 46847 appears to be the earlier canonical duplicate.", "confidence": 0.99, "canonical_issue_reason": null, - "canonical_pr_reason": "PR #46847 is the better canonical PR because it is the earlier of the two, has more discussion/review activity, and carries the same code changes and same explicit issue target as #47009.", + "canonical_pr_reason": "PR 46847 is the earlier instance of the identical fix, with the same target issue, same files, and the same configuration changes.", "best_issue_reason": null, - "best_pr_reason": "PR #46847 is the strongest representative of the cluster: same diff scope, same filenames, same issue target, and slightly more established through inbound references and review activity.", + "best_pr_reason": "PR 46847 is the best representative because it is earlier and has the more established review/inbound activity while implementing the same concrete code-path change as 47009.", "soft_edge_verdicts": [ { "left": "pull_request:46847", "right": "pull_request:47009", "accept": true, - "reason": "They are effectively the same patch: identical changed files, identical `linear_attn` TP-plan additions, and the same explicit issue target. They could plausibly be merged into one PR." + "reason": "Same underlying change: identical Qwen3.5 linear_attn tp-plan entries in the same files, with the same issue target and matching diff shape." } ] }, @@ -25918,12 +26066,12 @@ "soft_pairs": [ "pull_request:46993|pull_request:46994" ], - "prepared_review_unit_hash": "c4af77e0f2fff280e6710547029f50ed97edec8a01794711002ec8df152d33fc", + "prepared_review_unit_hash": "10cdd3de01bfd45178654e7f3fda490993621f3a5e16ca6120d4a398feb1005e", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4310, + "serialized_chars": 4309, "estimated_input_tokens": 1078, "estimated_eval_tokens": 2412 }, @@ -25931,7 +26079,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4310, + "serialized_chars": 4309, "estimated_input_tokens": 1078, "estimated_eval_tokens": 2412 }, @@ -25940,18 +26088,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "The two pull requests are unrelated and should not be deduplicated: one fixes prompt lookup decoding overshooting max_length, the other adds validation for empty token lists in logits processor configuration.", - "confidence": 0.99, + "summary": "These two pull requests are not duplicates: one fixes prompt lookup decoding exceeding max_length, while the other tightens validation for empty bad_words_ids / sequence_bias inputs in logits processing.", + "confidence": 0.98, "canonical_issue_reason": null, "canonical_pr_reason": null, "best_issue_reason": null, - "best_pr_reason": null, + "best_pr_reason": "Neither PR is a good canonical representative because they address unrelated bugs in different generation subsystems and would not plausibly be merged into one change.", "soft_edge_verdicts": [ { "left": "pull_request:46993", "right": "pull_request:46994", "accept": false, - "reason": "Different bug classes and code paths: PR 46993 changes generation candidate length capping in candidate_generator.py, while PR 46994 adds input validation in logits_process.py for empty bad_words_ids/sequence_bias lists. They would not plausibly be merged into one PR." + "reason": "Different concrete problems and code paths: PR 46993 changes candidate generation length capping; PR 46994 adds argument validation for empty token lists in logits processors. They are not the same underlying bug or a single mergeable fix." } ] }, diff --git a/analysis/current/manifest.json b/analysis/current/manifest.json index 2d8534c0db6a230de43a847a8fd72d03b9c05a2e..084131ccc852e9d02ff59a444e78799e5520c93a 100644 --- a/analysis/current/manifest.json +++ b/analysis/current/manifest.json @@ -1,8 +1,8 @@ { - "analysis_id": "hybrid-model-20260702t000029z", + "analysis_id": "hybrid-model-20260702t060019z", "archived_artifacts": { - "hybrid": "snapshots/20260702T000029Z/analysis-runs/hybrid-model-20260702t000029z/analysis-report-hybrid.json", - "hybrid_reviews": "snapshots/20260702T000029Z/analysis-runs/hybrid-model-20260702t000029z/analysis-report-hybrid.llm-reviews.json" + "hybrid": "snapshots/20260702T060019Z/analysis-runs/hybrid-model-20260702t060019z/analysis-report-hybrid.json", + "hybrid_reviews": "snapshots/20260702T060019Z/analysis-runs/hybrid-model-20260702t060019z/analysis-report-hybrid.llm-reviews.json" }, "artifacts": { "hybrid": "analysis/current/analysis-report-hybrid.json", @@ -10,9 +10,9 @@ }, "channel": "canonical", "model": null, - "published_at": "2026-07-02T00:13:12Z", + "published_at": "2026-07-02T06:11:20Z", "repo": "huggingface/transformers", "schema_version": 1, - "snapshot_id": "20260702T000029Z", + "snapshot_id": "20260702T060019Z", "variant": "hybrid" } diff --git a/snapshots/20260702T060019Z/analysis-runs/hybrid-model-20260702t060019z/analysis-report-hybrid.json b/snapshots/20260702T060019Z/analysis-runs/hybrid-model-20260702t060019z/analysis-report-hybrid.json new file mode 100644 index 0000000000000000000000000000000000000000..f6f4746ec4b8f75a8b2f93fc44623b93280b517a --- /dev/null +++ b/snapshots/20260702T060019Z/analysis-runs/hybrid-model-20260702t060019z/analysis-report-hybrid.json @@ -0,0 +1,2565 @@ +{ + "schema_version": "1.0", + "repo": "huggingface/transformers", + "snapshot_id": "20260702T060019Z", + "generated_at": "2026-07-02T06:11:19Z", + "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-46894-6", + "summary": "Cluster of 3 issues and 3 PRs centered on issue #46894.", + "status": "open", + "confidence": 0.6, + "canonical_issue_number": 46894, + "canonical_pr_number": 46903, + "issue_numbers": [ + 46894, + 46897, + 46901 + ], + "pr_numbers": [ + 46895, + 46903, + 46958 + ], + "evidence_types": [ + "closing_reference", + "soft_similarity" + ], + "pr_comparisons": [ + { + "left_pr_number": 46895, + "right_pr_number": 46903, + "code_similarity": 0.362, + "size_similarity": 0.443, + "file_overlap": 0.333, + "area_overlap": 0.419, + "patch_similarity": 0.178, + "shared_filenames": [ + "src/transformers/models/cohere_asr/modeling_cohere_asr.py", + "src/transformers/models/cohere_asr/modular_cohere_asr.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/models/cohere_asr/modeling_cohere_asr.py", + "left_ranges": [ + [ + 615, + 627 + ], + [ + 639, + 654 + ] + ], + "right_ranges": [ + [ + 637, + 651 + ] + ] + }, + { + "filename": "src/transformers/models/cohere_asr/modular_cohere_asr.py", + "left_ranges": [ + [ + 478, + 490 + ], + [ + 502, + 517 + ] + ], + "right_ranges": [ + [ + 500, + 514 + ] + ] + } + ] + }, + { + "left_pr_number": 46895, + "right_pr_number": 46958, + "code_similarity": 0.715, + "size_similarity": 0.833, + "file_overlap": 1.0, + "area_overlap": 0.366, + "patch_similarity": 0.802, + "shared_filenames": [ + "src/transformers/models/cohere_asr/modeling_cohere_asr.py", + "src/transformers/models/cohere_asr/modular_cohere_asr.py", + "tests/models/cohere_asr/test_modeling_cohere_asr.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/models/cohere_asr/modeling_cohere_asr.py", + "left_ranges": [ + [ + 615, + 627 + ], + [ + 639, + 654 + ] + ], + "right_ranges": [ + [ + 22, + 28 + ], + [ + 638, + 645 + ] + ] + }, + { + "filename": "src/transformers/models/cohere_asr/modular_cohere_asr.py", + "left_ranges": [ + [ + 478, + 490 + ], + [ + 502, + 517 + ] + ], + "right_ranges": [ + [ + 16, + 22 + ], + [ + 501, + 508 + ] + ] + }, + { + "filename": "tests/models/cohere_asr/test_modeling_cohere_asr.py", + "left_ranges": [ + [ + 35, + 41 + ], + [ + 176, + 214 + ] + ], + "right_ranges": [ + [ + 15, + 21 + ], + [ + 180, + 221 + ] + ] + } + ] + }, + { + "left_pr_number": 46903, + "right_pr_number": 46958, + "code_similarity": 0.325, + "size_similarity": 0.369, + "file_overlap": 0.333, + "area_overlap": 0.364, + "patch_similarity": 0.156, + "shared_filenames": [ + "src/transformers/models/cohere_asr/modeling_cohere_asr.py", + "src/transformers/models/cohere_asr/modular_cohere_asr.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/models/cohere_asr/modeling_cohere_asr.py", + "left_ranges": [ + [ + 637, + 651 + ] + ], + "right_ranges": [ + [ + 22, + 28 + ], + [ + 638, + 645 + ] + ] + }, + { + "filename": "src/transformers/models/cohere_asr/modular_cohere_asr.py", + "left_ranges": [ + [ + 500, + 514 + ] + ], + "right_ranges": [ + [ + 16, + 22 + ], + [ + 501, + 508 + ] + ] + } + ] + } + ] + }, + { + "cluster_id": "cluster-46899-4", + "summary": "Cluster of 2 issues and 2 PRs centered on issue #46932.", + "status": "open", + "confidence": 0.8, + "canonical_issue_number": 46932, + "canonical_pr_number": 46933, + "issue_numbers": [ + 46899, + 46932 + ], + "pr_numbers": [ + 46933, + 47008 + ], + "evidence_types": [ + "closing_reference", + "shared_issue_target", + "soft_similarity" + ], + "pr_comparisons": [ + { + "left_pr_number": 46933, + "right_pr_number": 47008, + "code_similarity": 0.706, + "size_similarity": 1.0, + "file_overlap": 1.0, + "area_overlap": 0.161, + "patch_similarity": 1.0, + "shared_filenames": [ + "src/transformers/models/gemma4/modeling_gemma4.py", + "src/transformers/models/gemma4/modular_gemma4.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/models/gemma4/modeling_gemma4.py", + "left_ranges": [ + [ + 369, + 377 + ], + [ + 612, + 620 + ] + ], + "right_ranges": [ + [ + 373, + 381 + ], + [ + 616, + 624 + ] + ] + }, + { + "filename": "src/transformers/models/gemma4/modular_gemma4.py", + "left_ranges": [ + [ + 343, + 351 + ], + [ + 586, + 594 + ] + ], + "right_ranges": [ + [ + 402, + 410 + ], + [ + 645, + 653 + ] + ] + } + ] + } + ] + }, + { + "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-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-46869-3", + "summary": "Cluster of 1 issues and 2 PRs centered on issue #46869.", + "status": "open", + "confidence": 0.75, + "canonical_issue_number": 46869, + "canonical_pr_number": 46886, + "issue_numbers": [ + 46869 + ], + "pr_numbers": [ + 46886, + 46922 + ], + "evidence_types": [ + "closing_reference", + "shared_issue_target" + ], + "pr_comparisons": [ + { + "left_pr_number": 46886, + "right_pr_number": 46922, + "code_similarity": 0.711, + "size_similarity": 0.857, + "file_overlap": 1.0, + "area_overlap": 0.5, + "patch_similarity": 0.429, + "shared_filenames": [ + "src/transformers/models/yolos/modeling_yolos.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/models/yolos/modeling_yolos.py", + "left_ranges": [ + [ + 446, + 452 + ] + ], + "right_ranges": [ + [ + 434, + 440 + ], + [ + 446, + 452 + ] + ] + } + ] + } + ] + }, + { + "cluster_id": "cluster-46159-3", + "summary": "Cluster of 1 issues and 2 PRs centered on issue #46159.", + "status": "open", + "confidence": 0.55, + "canonical_issue_number": 46159, + "canonical_pr_number": 46174, + "issue_numbers": [ + 46159 + ], + "pr_numbers": [ + 46174, + 46780 + ], + "evidence_types": [ + "closing_reference" + ], + "pr_comparisons": [ + { + "left_pr_number": 46174, + "right_pr_number": 46780, + "code_similarity": 0.841, + "size_similarity": 0.87, + "file_overlap": 1.0, + "area_overlap": 0.767, + "patch_similarity": 0.657, + "shared_filenames": [ + "src/transformers/models/aimv2/modeling_aimv2.py", + "src/transformers/models/d_fine/modeling_d_fine.py", + "src/transformers/models/deimv2/modeling_deimv2.py", + "src/transformers/models/pp_doclayout_v2/modeling_pp_doclayout_v2.py", + "src/transformers/models/pp_doclayout_v3/modeling_pp_doclayout_v3.py", + "src/transformers/models/rt_detr/modeling_rt_detr.py", + "src/transformers/models/rt_detr_v2/modeling_rt_detr_v2.py", + "src/transformers/models/vit_mae/modeling_vit_mae.py", + "src/transformers/models/vit_mae/modular_vit_mae.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/models/aimv2/modeling_aimv2.py", + "left_ranges": [ + [ + 134, + 141 + ], + [ + 144, + 154 + ], + [ + 157, + 165 + ] + ], + "right_ranges": [ + [ + 134, + 140 + ], + [ + 143, + 153 + ], + [ + 156, + 162 + ] + ] + }, + { + "filename": "src/transformers/models/d_fine/modeling_d_fine.py", + "left_ranges": [ + [ + 631, + 638 + ], + [ + 641, + 651 + ], + [ + 654, + 662 + ] + ], + "right_ranges": [ + [ + 631, + 637 + ], + [ + 640, + 650 + ], + [ + 653, + 659 + ] + ] + }, + { + "filename": "src/transformers/models/deimv2/modeling_deimv2.py", + "left_ranges": [ + [ + 703, + 710 + ], + [ + 713, + 723 + ], + [ + 726, + 734 + ] + ], + "right_ranges": [ + [ + 703, + 709 + ], + [ + 712, + 722 + ], + [ + 725, + 731 + ] + ] + }, + { + "filename": "src/transformers/models/pp_doclayout_v2/modeling_pp_doclayout_v2.py", + "left_ranges": [ + [ + 1551, + 1558 + ], + [ + 1561, + 1571 + ], + [ + 1574, + 1582 + ] + ], + "right_ranges": [ + [ + 1551, + 1557 + ], + [ + 1560, + 1570 + ], + [ + 1573, + 1579 + ] + ] + }, + { + "filename": "src/transformers/models/pp_doclayout_v3/modeling_pp_doclayout_v3.py", + "left_ranges": [ + [ + 804, + 811 + ], + [ + 814, + 824 + ], + [ + 827, + 835 + ] + ], + "right_ranges": [ + [ + 804, + 810 + ], + [ + 813, + 823 + ], + [ + 826, + 832 + ] + ] + }, + { + "filename": "src/transformers/models/rt_detr/modeling_rt_detr.py", + "left_ranges": [ + [ + 847, + 854 + ], + [ + 857, + 867 + ], + [ + 870, + 878 + ] + ], + "right_ranges": [ + [ + 847, + 853 + ], + [ + 856, + 866 + ], + [ + 869, + 875 + ] + ] + }, + { + "filename": "src/transformers/models/rt_detr_v2/modeling_rt_detr_v2.py", + "left_ranges": [ + [ + 976, + 983 + ], + [ + 986, + 996 + ], + [ + 999, + 1007 + ] + ], + "right_ranges": [ + [ + 976, + 982 + ], + [ + 985, + 995 + ], + [ + 998, + 1004 + ] + ] + }, + { + "filename": "src/transformers/models/vit_mae/modeling_vit_mae.py", + "left_ranges": [ + [ + 152, + 159 + ], + [ + 162, + 172 + ], + [ + 175, + 183 + ] + ], + "right_ranges": [ + [ + 152, + 158 + ], + [ + 161, + 171 + ], + [ + 174, + 180 + ] + ] + }, + { + "filename": "src/transformers/models/vit_mae/modular_vit_mae.py", + "left_ranges": [ + [ + 62, + 69 + ], + [ + 72, + 82 + ], + [ + 85, + 93 + ] + ], + "right_ranges": [ + [ + 62, + 68 + ], + [ + 71, + 81 + ], + [ + 84, + 90 + ] + ] + } + ] + } + ] + }, + { + "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": 0.721, + "size_similarity": 0.973, + "file_overlap": 0.75, + "area_overlap": 0.563, + "patch_similarity": 0.696, + "shared_filenames": [ + "src/transformers/testing_utils.py", + "tests/utils/test_testing_utils.py", + "utils/notification_service.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/testing_utils.py", + "left_ranges": [ + [ + 3611, + 3618 + ], + [ + 3835, + 3869 + ] + ], + "right_ranges": [ + [ + 3525, + 3558 + ], + [ + 3782, + 3788 + ] + ] + }, + { + "filename": "tests/utils/test_testing_utils.py", + "left_ranges": [ + [ + 1, + 144 + ] + ], + "right_ranges": [ + [ + 1, + 114 + ] + ] + }, + { + "filename": "utils/notification_service.py", + "left_ranges": [ + [ + 935, + 967 + ] + ], + "right_ranges": [ + [ + 935, + 967 + ] + ] + } + ] + } + ] + }, + { + "cluster_id": "cluster-46846-3", + "summary": "Cluster of 1 issues and 2 PRs centered on issue #46846.", + "status": "open", + "confidence": 0.8, + "canonical_issue_number": 46846, + "canonical_pr_number": 46847, + "issue_numbers": [ + 46846 + ], + "pr_numbers": [ + 46847, + 47009 + ], + "evidence_types": [ + "closing_reference", + "shared_issue_target", + "soft_similarity" + ], + "pr_comparisons": [ + { + "left_pr_number": 46847, + "right_pr_number": 47009, + "code_similarity": 1.0, + "size_similarity": 1.0, + "file_overlap": 1.0, + "area_overlap": 1.0, + "patch_similarity": 1.0, + "shared_filenames": [ + "src/transformers/models/qwen3_5/configuration_qwen3_5.py", + "src/transformers/models/qwen3_5/modular_qwen3_5.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/models/qwen3_5/configuration_qwen3_5.py", + "left_ranges": [ + [ + 66, + 76 + ] + ], + "right_ranges": [ + [ + 66, + 76 + ] + ] + }, + { + "filename": "src/transformers/models/qwen3_5/modular_qwen3_5.py", + "left_ranges": [ + [ + 103, + 113 + ] + ], + "right_ranges": [ + [ + 103, + 113 + ] + ] + } + ] + } + ] + }, + { + "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-46305-3", + "summary": "Cluster of 1 issues and 2 PRs centered on issue #46305.", + "status": "open", + "confidence": 0.8, + "canonical_issue_number": 46305, + "canonical_pr_number": 46310, + "issue_numbers": [ + 46305 + ], + "pr_numbers": [ + 46310, + 47010 + ], + "evidence_types": [ + "closing_reference", + "shared_issue_target", + "soft_similarity" + ], + "pr_comparisons": [ + { + "left_pr_number": 46310, + "right_pr_number": 47010, + "code_similarity": 0.833, + "size_similarity": 0.667, + "file_overlap": 0.667, + "area_overlap": 1.0, + "patch_similarity": 1.0, + "shared_filenames": [ + "src/transformers/models/moonshine_streaming/modeling_moonshine_streaming.py", + "src/transformers/models/moonshine_streaming/modular_moonshine_streaming.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/models/moonshine_streaming/modeling_moonshine_streaming.py", + "left_ranges": [ + [ + 356, + 362 + ] + ], + "right_ranges": [ + [ + 356, + 362 + ] + ] + }, + { + "filename": "src/transformers/models/moonshine_streaming/modular_moonshine_streaming.py", + "left_ranges": [ + [ + 269, + 275 + ] + ], + "right_ranges": [ + [ + 269, + 275 + ] + ] + } + ] + } + ] + }, + { + "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-46970-3", + "summary": "Cluster of 3 related pull requests linked by soft_similarity.", + "status": "open", + "confidence": 0.5, + "canonical_issue_number": null, + "canonical_pr_number": 46986, + "issue_numbers": [], + "pr_numbers": [ + 46970, + 46986, + 47006 + ], + "evidence_types": [ + "soft_similarity" + ], + "pr_comparisons": [ + { + "left_pr_number": 46970, + "right_pr_number": 46986, + "code_similarity": 0.01, + "size_similarity": 0.029, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.026, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 46970, + "right_pr_number": 47006, + "code_similarity": 0.008, + "size_similarity": 0.02, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.026, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 46986, + "right_pr_number": 47006, + "code_similarity": 0.144, + "size_similarity": 0.697, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.028, + "shared_filenames": [], + "shared_file_areas": [] + } + ] + }, + { + "cluster_id": "cluster-46908-2", + "summary": "Cluster of 2 related pull requests linked by soft_similarity.", + "status": "open", + "confidence": 0.5, + "canonical_issue_number": null, + "canonical_pr_number": 46926, + "issue_numbers": [], + "pr_numbers": [ + 46908, + 46926 + ], + "evidence_types": [ + "soft_similarity" + ], + "pr_comparisons": [ + { + "left_pr_number": 46908, + "right_pr_number": 46926, + "code_similarity": 0.31, + "size_similarity": 0.8, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 1.0, + "shared_filenames": [], + "shared_file_areas": [] + } + ] + }, + { + "cluster_id": "cluster-46927-2", + "summary": "Cluster of 2 related pull requests linked by soft_similarity.", + "status": "open", + "confidence": 0.5, + "canonical_issue_number": null, + "canonical_pr_number": 46927, + "issue_numbers": [], + "pr_numbers": [ + 46927, + 46971 + ], + "evidence_types": [ + "soft_similarity" + ], + "pr_comparisons": [ + { + "left_pr_number": 46927, + "right_pr_number": 46971, + "code_similarity": 0.162, + "size_similarity": 0.786, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.029, + "shared_filenames": [], + "shared_file_areas": [] + } + ] + } + ], + "duplicate_issues": [ + { + "cluster_id": "cluster-46894-6", + "canonical_issue_number": 46894, + "duplicate_issue_numbers": [ + 46897, + 46901 + ], + "reason": "Issues in cluster-46894-6 are treated as duplicates because they share closing_reference, soft_similarity evidence." + }, + { + "cluster_id": "cluster-46899-4", + "canonical_issue_number": 46932, + "duplicate_issue_numbers": [ + 46899 + ], + "reason": "Issues in cluster-46899-4 are treated as duplicates because they share closing_reference, shared_issue_target, soft_similarity evidence." + } + ], + "duplicate_prs": [ + { + "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-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-46159-3", + "canonical_pr_number": 46174, + "duplicate_pr_numbers": [ + 46780 + ], + "target_issue_number": 46159, + "reason": "PRs in cluster-46159-3 are treated as duplicates because they converge on issue #46159 with closing_reference evidence." + }, + { + "cluster_id": "cluster-46305-3", + "canonical_pr_number": 46310, + "duplicate_pr_numbers": [ + 47010 + ], + "target_issue_number": 46305, + "reason": "PRs in cluster-46305-3 are treated as duplicates because they converge on issue #46305 with closing_reference, shared_issue_target, soft_similarity evidence." + }, + { + "cluster_id": "cluster-46846-3", + "canonical_pr_number": 46847, + "duplicate_pr_numbers": [ + 47009 + ], + "target_issue_number": 46846, + "reason": "PRs in cluster-46846-3 are treated as duplicates because they converge on issue #46846 with closing_reference, shared_issue_target, soft_similarity evidence." + }, + { + "cluster_id": "cluster-46869-3", + "canonical_pr_number": 46886, + "duplicate_pr_numbers": [ + 46922 + ], + "target_issue_number": 46869, + "reason": "PRs in cluster-46869-3 are treated as duplicates because they converge on issue #46869 with closing_reference, shared_issue_target evidence." + }, + { + "cluster_id": "cluster-46894-6", + "canonical_pr_number": 46903, + "duplicate_pr_numbers": [ + 46895, + 46958 + ], + "target_issue_number": 46894, + "reason": "PRs in cluster-46894-6 are treated as duplicates because they converge on issue #46894 with closing_reference, soft_similarity evidence." + }, + { + "cluster_id": "cluster-46899-4", + "canonical_pr_number": 46933, + "duplicate_pr_numbers": [ + 47008 + ], + "target_issue_number": 46932, + "reason": "PRs in cluster-46899-4 are treated as duplicates because they converge on issue #46932 with closing_reference, shared_issue_target, soft_similarity evidence." + }, + { + "cluster_id": "cluster-46908-2", + "canonical_pr_number": 46926, + "duplicate_pr_numbers": [ + 46908 + ], + "target_issue_number": null, + "reason": "PRs in cluster-46908-2 are treated as duplicates because they share soft_similarity evidence." + }, + { + "cluster_id": "cluster-46927-2", + "canonical_pr_number": 46927, + "duplicate_pr_numbers": [ + 46971 + ], + "target_issue_number": null, + "reason": "PRs in cluster-46927-2 are treated as duplicates because they share soft_similarity evidence." + }, + { + "cluster_id": "cluster-46970-3", + "canonical_pr_number": 46986, + "duplicate_pr_numbers": [ + 46970, + 47006 + ], + "target_issue_number": null, + "reason": "PRs in cluster-46970-3 are treated as duplicates because they share soft_similarity evidence." + } + ], + "best_issue": { + "cluster_id": "cluster-46894-6", + "issue_number": 46894, + "reason": "Issue #46894 is the strongest global issue candidate because it is open, belongs to a cluster with 6 artifacts, and carries 0 discussion comments plus 6 inbound references.", + "score": 166.0 + }, + "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/20260702T060019Z/analysis-runs/hybrid-model-20260702t060019z/analysis-report-hybrid.llm-reviews.json b/snapshots/20260702T060019Z/analysis-runs/hybrid-model-20260702t060019z/analysis-report-hybrid.llm-reviews.json new file mode 100644 index 0000000000000000000000000000000000000000..326f67316993bb422f2e8d1fbd66a810869e47cd --- /dev/null +++ b/snapshots/20260702T060019Z/analysis-runs/hybrid-model-20260702t060019z/analysis-report-hybrid.llm-reviews.json @@ -0,0 +1,26114 @@ +{ + "schema_version": "1.0", + "repo": "huggingface/transformers", + "snapshot_id": "20260702T060019Z", + "generated_at": "2026-07-02T06:11:19Z", + "model": "gpt-5.4-mini?service_tier=flex", + "reviews": [ + { + "label": "issue", + "component_index": 1, + "component_count": 26, + "review_unit_index": 1, + "review_unit_count": 61, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-505", + "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:43596", + "issue:43638", + "issue:43716", + "issue:43827", + "issue:43828", + "issue:43975", + "issue:44292", + "issue:44322", + "issue:44457", + "issue:44509", + "issue:44512", + "issue:44560", + "issue:44805", + "issue:44841", + "issue:44918", + "issue:45081", + "issue:45161", + "issue:45237", + "issue:45464", + "issue:45507", + "issue:45685", + "issue:45800", + "issue:45823", + "issue:45859", + "issue:45874", + "issue:45901", + "issue:45902", + "issue:46015", + "issue:46165", + "issue:46294", + "issue:46468", + "issue:46515", + "issue:46661", + "issue:46869" + ], + "soft_pairs": [ + "issue:44322|issue:45464", + "issue:45237|issue:45685", + "issue:43716|issue:45237", + "issue:43828|issue:45237", + "issue:45237|issue:46468", + "issue:45237|issue:45507", + "issue:30277|issue:35545", + "issue:35545|issue:42915", + "issue:35545|issue:41720", + "issue:45901|issue:46015", + "issue:35545|issue:36010", + "issue:44509|issue:44512", + "issue:44918|issue:46165", + "issue:46165|issue:46869", + "issue:44805|issue:45823", + "issue:43638|issue:44805", + "issue:43638|issue:46165", + "issue:43596|issue:44805", + "issue:44805|issue:46294", + "issue:44805|issue:45161", + "issue:46165|issue:46468", + "issue:44560|issue:46165", + "issue:34634|issue:35545", + "issue:30277|issue:43638", + "issue:43596|issue:43638", + "issue:43039|issue:45161", + "issue:43975|issue:44457", + "issue:44292|issue:44457", + "issue:44841|issue:46661", + "issue:43716|issue:46869", + "issue:24643|issue:30277", + "issue:34928|issue:44805", + "issue:45685|issue:46869", + "issue:45859|issue:45902", + "issue:43638|issue:45161", + "issue:34928|issue:36331", + "issue:43596|issue:45161", + "issue:43716|issue:45507", + "issue:45874|issue:46294", + "issue:44292|issue:45237", + "issue:39290|issue:45081", + "issue:45823|issue:45874", + "issue:45800|issue:46515", + "issue:34928|issue:41720", + "issue:43425|issue:44292", + "issue:43828|issue:45161", + "issue:43541|issue:43716", + "issue:43828|issue:44292", + "issue:43716|issue:46468", + "issue:44292|issue:44918", + "issue:45507|issue:45685", + "issue:43716|issue:44292", + "issue:43716|issue:45685", + "issue:44918|issue:45161", + "issue:43039|issue:44918", + "issue:41762|issue:43638", + "issue:43827|issue:44512", + "issue:43828|issue:45685", + "issue:43039|issue:43828", + "issue:30277|issue:36331", + "issue:34928|issue:36010", + "issue:45237|issue:46015", + "issue:43039|issue:44805" + ], + "prepared_review_unit_hash": "064d6365c1d30a398675df748be644a53c9e808dd7e34c410a3bc80383480118", + "component_budget": { + "node_count": 505, + "item_count": 505, + "soft_pair_count": 1854, + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 63, + "serialized_chars": 46576, + "estimated_input_tokens": 11644, + "estimated_eval_tokens": 23544 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "Only one clear duplicate pair appears in this cluster: the DeepSpeed ZeRO-3 BertModel IndexError reports. The rest are mostly loosely related model/loading/device-mismatch/doc issues and should stay separate.", + "confidence": 0.84, + "canonical_issue_reason": "issue:43638 is the best canonical issue because it captures the same ZeRO-3 Bert IndexError as issue:43596, but with a more specific repro and more discussion.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43638 is the strongest representative of the only real duplicate set in the cluster and is the most specific of the two Bert ZeRO-3 reports.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44322", + "right": "issue:45464", + "accept": false, + "reason": "Different failures: missing Qwen3.5 config fields versus streaming chat/completions API behavior." + }, + { + "left": "issue:45237", + "right": "issue:45685", + "accept": false, + "reason": "Both mention runtime/device problems, but one is AMD GPU support for GPT-OSS and the other is an MPS histogram error in MoE code; different bugs." + }, + { + "left": "issue:43716", + "right": "issue:45237", + "accept": false, + "reason": "Different code paths: image preprocessor dtype mismatch versus GPT-OSS failing on AMD GPUs." + }, + { + "left": "issue:43828", + "right": "issue:45237", + "accept": false, + "reason": "Both are dtype-related, but one is an autocast mismatch and the other is an AMD GPU compatibility issue; not the same bug." + }, + { + "left": "issue:45237", + "right": "issue:46468", + "accept": false, + "reason": "Different models and symptoms: GPT-OSS on AMD GPUs versus a broken official Nemotron example." + }, + { + "left": "issue:45237", + "right": "issue:45507", + "accept": false, + "reason": "Different failures in different model families; no shared concrete bug." + }, + { + "left": "issue:30277", + "right": "issue:35545", + "accept": false, + "reason": "DeepSpeed ZeRO-3 collective mismatch is unrelated to ModernBERT ONNX export." + }, + { + "left": "issue:35545", + "right": "issue:42915", + "accept": false, + "reason": "ONNX export error and FineGrainedFP8 failure are unrelated model/runtime issues." + }, + { + "left": "issue:35545", + "right": "issue:41720", + "accept": false, + "reason": "Different problems: ONNX export versus Qwen3 auto device mapping cudaErrorAssert." + }, + { + "left": "issue:45901", + "right": "issue:46015", + "accept": false, + "reason": "A task crash is not the same as outdated RAG examples documentation." + }, + { + "left": "issue:35545", + "right": "issue:36010", + "accept": false, + "reason": "Different failures: ONNX export versus GenerationMixin import regression." + }, + { + "left": "issue:44509", + "right": "issue:44512", + "accept": false, + "reason": "Both are docs cleanups for v5 removals, but they target different deprecated content and are not the same change." + }, + { + "left": "issue:44918", + "right": "issue:46165", + "accept": false, + "reason": "TRL SFT embedding unpacking failure is unrelated to resuming training with a locally downloaded remote-code model." + }, + { + "left": "issue:46165", + "right": "issue:46869", + "accept": false, + "reason": "Different scenarios and code paths: resume-training/remote-code versus YOLO multi-GPU device-map auto." + }, + { + "left": "issue:44805", + "right": "issue:45823", + "accept": false, + "reason": "Mask-shape IndexError is unrelated to Gemma4 device-map forward mismatch." + }, + { + "left": "issue:43638", + "right": "issue:44805", + "accept": false, + "reason": "Both are IndexErrors, but one is ZeRO-3 Bert initialization and the other is a mask/tensor shape mismatch; different bugs." + }, + { + "left": "issue:43596", + "right": "issue:44805", + "accept": false, + "reason": "Different failure modes and subsystems: ZeRO-3 Bert init versus tensor mask shape mismatch." + }, + { + "left": "issue:44805", + "right": "issue:46294", + "accept": false, + "reason": "Shape-mismatch IndexError is unrelated to Qwen3.5 video token/template mismatch." + }, + { + "left": "issue:44805", + "right": "issue:45161", + "accept": false, + "reason": "Different models and bugs: mask/tensor shape mismatch versus GPT-OSS MoE tensor-parallel failure." + }, + { + "left": "issue:46165", + "right": "issue:46468", + "accept": false, + "reason": "Resume-training remote-code issue is unrelated to the broken Nemotron example." + }, + { + "left": "issue:44560", + "right": "issue:46165", + "accept": false, + "reason": "Video StopIteration in chat template handling is unrelated to resuming training from a local model." + }, + { + "left": "issue:34634", + "right": "issue:35545", + "accept": false, + "reason": "BarkProcessor voice_preset parsing and ModernBERT ONNX export are unrelated." + }, + { + "left": "issue:30277", + "right": "issue:43638", + "accept": false, + "reason": "DeepSpeed collective mismatch on Jamba-v01 is not the same as the Bert ZeRO-3 IndexError." + }, + { + "left": "issue:43596", + "right": "issue:43638", + "accept": true, + "reason": "These describe the same ZeRO-3 BertModel IndexError; the non-pretrained Bert repro is just a more specific instance." + }, + { + "left": "issue:43039", + "right": "issue:45161", + "accept": false, + "reason": "Liger Kernel cross-entropy routing and GPT-OSS tensor-parallel failure are different issues." + }, + { + "left": "issue:43975", + "right": "issue:44457", + "accept": false, + "reason": "Incorrect detokenization and LoRA merge/save inconsistency are unrelated bugs." + }, + { + "left": "issue:44292", + "right": "issue:44457", + "accept": false, + "reason": "Qwen-3-8B-NVFP4 runtime error is unrelated to LoRA weight merge/reload mismatch." + }, + { + "left": "issue:44841", + "right": "issue:46661", + "accept": false, + "reason": "Processor failure for Voxtral and remote-code config_class AttributeError are different failures." + }, + { + "left": "issue:43716", + "right": "issue:46869", + "accept": false, + "reason": "Mistral dtype mismatch and YOLO multi-GPU device-map failure are unrelated." + }, + { + "left": "issue:24643", + "right": "issue:30277", + "accept": false, + "reason": "Both involve DeepSpeed, but the concrete failures are different: 2-D weight error versus collective mismatch." + }, + { + "left": "issue:34928", + "right": "issue:44805", + "accept": false, + "reason": "Activation checkpointing/FSDP recomputation mismatch is a different bug from the mask-shape IndexError." + }, + { + "left": "issue:45685", + "right": "issue:46869", + "accept": false, + "reason": "MPS histogram not implemented for Int is unrelated to YOLO device-map auto failures." + }, + { + "left": "issue:45859", + "right": "issue:45902", + "accept": false, + "reason": "Same module family, but one is CPU offload incompatibility and the other is uninitialized memory after materialization; not clearly the same bug." + }, + { + "left": "issue:43638", + "right": "issue:45161", + "accept": false, + "reason": "ZeRO-3 Bert initialization and GPT-OSS tensor-parallel issues are unrelated." + }, + { + "left": "issue:34928", + "right": "issue:36331", + "accept": false, + "reason": "FSDP activation-checkpointing tensor recomputation and CustomTrainer compute_loss signature mismatch are unrelated." + }, + { + "left": "issue:43596", + "right": "issue:45161", + "accept": false, + "reason": "Different model and stack: Bert ZeRO-3 init versus GPT-OSS TP failure." + }, + { + "left": "issue:43716", + "right": "issue:45507", + "accept": false, + "reason": "Mistral dtype mismatch and GraniteMoEHybrid invalid method do not share the same underlying bug." + }, + { + "left": "issue:45874", + "right": "issue:46294", + "accept": false, + "reason": "Gemma4 reverse-embedding lookup and Qwen3.5 video token mismatch are unrelated." + }, + { + "left": "issue:44292", + "right": "issue:45237", + "accept": false, + "reason": "Qwen-3-8B-NVFP4 error and GPT-OSS AMD GPU failure are different model-specific problems." + }, + { + "left": "issue:39290", + "right": "issue:45081", + "accept": false, + "reason": "Gemma3TextConfig missing attribute with vLLM is unrelated to the Mistral regex patch crash." + }, + { + "left": "issue:45823", + "right": "issue:45874", + "accept": false, + "reason": "Both are Gemma4 issues, but one is device-mismatch in forward and the other is expensive reverse embedding lookup; different code paths." + }, + { + "left": "issue:45800", + "right": "issue:46515", + "accept": false, + "reason": "Both mention torch compatibility, but they are different reported incompatibilities and are not the same bug." + }, + { + "left": "issue:34928", + "right": "issue:41720", + "accept": false, + "reason": "FSDP checkpointing tensor mismatch and Qwen3 auto-device-map cudaErrorAssert are unrelated." + }, + { + "left": "issue:43425", + "right": "issue:44292", + "accept": false, + "reason": "General Torch 2.10 incompatibility is not the same as the Qwen-3-8B-NVFP4 runtime error." + }, + { + "left": "issue:43828", + "right": "issue:45161", + "accept": false, + "reason": "Autocast dtype mismatch and GPT-OSS tensor-parallel failure are distinct issues." + }, + { + "left": "issue:43541", + "right": "issue:43716", + "accept": false, + "reason": "Mixtral torch dynamo grouped_mm tracing error is unrelated to Mistral image-preprocessor dtype mismatch." + }, + { + "left": "issue:43828", + "right": "issue:44292", + "accept": false, + "reason": "Different failures: autocast dtype mismatch versus Qwen-3-8B-NVFP4 runtime error." + }, + { + "left": "issue:43716", + "right": "issue:46468", + "accept": false, + "reason": "Mistral dtype mismatch and broken Nemotron example are unrelated." + }, + { + "left": "issue:44292", + "right": "issue:44918", + "accept": false, + "reason": "Qwen-3-8B-NVFP4 runtime error is unrelated to TRL SFT embedding unpacking." + }, + { + "left": "issue:45507", + "right": "issue:45685", + "accept": false, + "reason": "Different model/runtime bugs: GraniteMoEHybrid invalid method versus MPS histogram limitation." + }, + { + "left": "issue:43716", + "right": "issue:44292", + "accept": false, + "reason": "Different model families and failure modes." + }, + { + "left": "issue:43716", + "right": "issue:45685", + "accept": false, + "reason": "Image-preprocessor dtype mismatch is unrelated to an MPS integer histogram error." + }, + { + "left": "issue:44918", + "right": "issue:45161", + "accept": false, + "reason": "TRl embedding unpacking and GPT-OSS tensor-parallel failure are different issues." + }, + { + "left": "issue:43039", + "right": "issue:44918", + "accept": false, + "reason": "Liger Kernel cross-entropy routing is unrelated to TRL SFT embedding unpacking." + }, + { + "left": "issue:41762", + "right": "issue:43638", + "accept": false, + "reason": "Gemma3 ZeRO-3 loading IndexError is a different bug from the Bert ZeRO-3 IndexError." + }, + { + "left": "issue:43827", + "right": "issue:44512", + "accept": false, + "reason": "Both are docs regressions, but they refer to different removed v5 features and should not be merged as the same change." + }, + { + "left": "issue:43828", + "right": "issue:45685", + "accept": false, + "reason": "Autocast dtype mismatch and MPS histogram not implemented are unrelated." + }, + { + "left": "issue:43039", + "right": "issue:43828", + "accept": false, + "reason": "Liger Kernel cross-entropy being called and autocast dtype mismatch are different bugs." + }, + { + "left": "issue:30277", + "right": "issue:36331", + "accept": false, + "reason": "DeepSpeed collective mismatch is unrelated to the CustomTrainer compute_loss signature change." + }, + { + "left": "issue:34928", + "right": "issue:36010", + "accept": false, + "reason": "Activation-checkpointing/FSDP mismatch and GenerationMixin import failure are unrelated." + }, + { + "left": "issue:45237", + "right": "issue:46015", + "accept": false, + "reason": "GPU/model runtime failure and outdated RAG docs are not the same underlying issue." + }, + { + "left": "issue:43039", + "right": "issue:44805", + "accept": false, + "reason": "Liger Kernel cross-entropy routing is unrelated to the tensor mask shape IndexError." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The analysis is grounded in the packet and stays conservative on the soft edges. The only clearly supported duplicate pair is the ZeRO-3 Bert IndexError reports (43596/43638), and the rest are reasonably kept separate given the titles and available evidence." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 26, + "review_unit_index": 2, + "review_unit_count": 61, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-505", + "nodes": [ + "issue:36010", + "issue:39290", + "issue:41093", + "issue:41720", + "issue:41762", + "issue:43039", + "issue:43296", + "issue:43404", + "issue:43425", + "issue:43531", + "issue:43541", + "issue:43572", + "issue:43596", + "issue:43638", + "issue:43792", + "issue:43828", + "issue:43854", + "issue:43866", + "issue:43872", + "issue:43901", + "issue:43975", + "issue:44292", + "issue:44512", + "issue:44560", + "issue:44589", + "issue:44841", + "issue:44863", + "issue:45020", + "issue:45070", + "issue:45084", + "issue:45237", + "issue:45362", + "issue:45399", + "issue:45464", + "issue:45507", + "issue:45685", + "issue:45800", + "issue:45812", + "issue:45823", + "issue:45901", + "issue:45920", + "issue:46015", + "issue:46294", + "issue:46304", + "issue:46402", + "issue:46661", + "issue:46772", + "issue:46869" + ], + "soft_pairs": [ + "issue:45800|issue:46304", + "issue:43296|issue:46304", + "issue:43866|issue:46402", + "issue:41762|issue:43596", + "issue:39290|issue:41720", + "issue:43901|issue:44512", + "issue:44589|issue:46304", + "issue:45237|issue:45901", + "issue:43975|issue:46015", + "issue:43039|issue:44560", + "issue:45823|issue:46294", + "issue:44841|issue:45823", + "issue:43866|issue:46772", + "issue:44863|issue:46015", + "issue:45020|issue:46304", + "issue:45812|issue:45920", + "issue:44292|issue:46015", + "issue:43866|issue:44863", + "issue:43872|issue:46304", + "issue:39290|issue:43531", + "issue:41093|issue:43638", + "issue:43792|issue:45901", + "issue:43854|issue:43866", + "issue:43425|issue:45070", + "issue:43975|issue:45901", + "issue:45362|issue:45464", + "issue:43792|issue:45070", + "issue:43596|issue:45800", + "issue:43541|issue:46869", + "issue:44292|issue:45901", + "issue:39290|issue:45823", + "issue:45464|issue:45685", + "issue:36010|issue:39290", + "issue:43541|issue:45685", + "issue:43541|issue:43828", + "issue:44841|issue:45084", + "issue:43404|issue:43866", + "issue:45507|issue:46869", + "issue:41093|issue:41762", + "issue:45399|issue:46661", + "issue:43572|issue:45070", + "issue:45800|issue:45823" + ], + "prepared_review_unit_hash": "b5f4c4219911ff1fcade2a7ac1aca13a313ba26f548eaaf30defeb93f85b6547", + "component_budget": { + "node_count": 505, + "item_count": 505, + "soft_pair_count": 1854, + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 42, + "serialized_chars": 40042, + "estimated_input_tokens": 10011, + "estimated_eval_tokens": 20278 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "Most items are unrelated. I only accept two soft duplicates: the ZeRO-3 load-time index-out-of-bounds reports, and the AutoTokenizer wrong-token-IDs regression reports.", + "confidence": 0.9, + "canonical_issue_reason": "issue:45920 is the best canonical issue: it describes the broad v5 AutoTokenizer wrong-token-ID regression across multiple model families, making it the clearest umbrella for the narrower Granite-specific report.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45920 is the best overall issue candidate because it is the broadest, most representative regression report in the set and likely subsumes issue:45812.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45800", + "right": "issue:46304", + "accept": false, + "reason": "Both are torch/version compatibility complaints, but they are about different breakages and code paths; not the same bug." + }, + { + "left": "issue:43296", + "right": "issue:46304", + "accept": false, + "reason": "Different failures: PaddleOCR-VL/vLLM model loading vs USE_DISTRIBUTED=0 support regression." + }, + { + "left": "issue:43866", + "right": "issue:46402", + "accept": false, + "reason": "Corrupted checkpoint vs weights not loaded; same area, different underlying problem." + }, + { + "left": "issue:41762", + "right": "issue:43596", + "accept": true, + "reason": "Both report the same ZeRO-3 load-time index-out-of-bounds failure from an empty/invalid parameter path; model names differ, but the bug looks identical." + }, + { + "left": "issue:39290", + "right": "issue:41720", + "accept": false, + "reason": "Gemma3 sliding_window_pattern missing in vLLM vs Qwen3 auto device-map CUDA assert are unrelated." + }, + { + "left": "issue:43901", + "right": "issue:44512", + "accept": false, + "reason": "Both are docs issues, but they cover different documentation problems." + }, + { + "left": "issue:44589", + "right": "issue:46304", + "accept": false, + "reason": "Float8 storage lookup failure is unrelated to the torch distributed compatibility breakage." + }, + { + "left": "issue:45237", + "right": "issue:45901", + "accept": false, + "reason": "Different subsystems and symptoms: AMD GPU support vs table-question-answering crash." + }, + { + "left": "issue:43975", + "right": "issue:46015", + "accept": false, + "reason": "DeepSeek detokenization bug vs outdated RAG examples; not the same issue." + }, + { + "left": "issue:43039", + "right": "issue:44560", + "accept": false, + "reason": "Liger Kernel cross_entropy fallback vs Qwen3-VL video StopIteration are unrelated." + }, + { + "left": "issue:45823", + "right": "issue:46294", + "accept": false, + "reason": "Different model-specific bugs: Gemma4 device mismatch vs Qwen3.5 video token/feature mismatch." + }, + { + "left": "issue:44841", + "right": "issue:45823", + "accept": false, + "reason": "Processor failure for Voxtral vs Gemma4 device mismatch; no common code-path bug shown." + }, + { + "left": "issue:43866", + "right": "issue:46772", + "accept": false, + "reason": "Checkpoint corruption and NVFP4 compatibility are distinct problems." + }, + { + "left": "issue:44863", + "right": "issue:46015", + "accept": false, + "reason": "NemotronH loading failure vs RAG docs drift are unrelated." + }, + { + "left": "issue:45020", + "right": "issue:46304", + "accept": false, + "reason": "Remote-code model breakage and torch distributed support regression are different changes." + }, + { + "left": "issue:45812", + "right": "issue:45920", + "accept": true, + "reason": "Same tokenizer regression: wrong token IDs in v5, with 45920 providing the broader multi-model scope and 45812 the Granite-specific instance." + }, + { + "left": "issue:44292", + "right": "issue:46015", + "accept": false, + "reason": "Qwen NVFP4 runtime error vs outdated RAG examples; unrelated." + }, + { + "left": "issue:43866", + "right": "issue:44863", + "accept": false, + "reason": "Checkpoint corruption vs inability to load NemotronH checkpoints are different issues." + }, + { + "left": "issue:43872", + "right": "issue:46304", + "accept": false, + "reason": "bitsandbytes constructor incompatibility is unrelated to the torch distributed build/runtime regression." + }, + { + "left": "issue:39290", + "right": "issue:43531", + "accept": false, + "reason": "Both mention sliding_window, but the affected models and failure modes differ." + }, + { + "left": "issue:41093", + "right": "issue:43638", + "accept": false, + "reason": "Both are index errors, but one is a mask/tensor shape mismatch and the other is a ZeRO-3 empty-tensor load issue." + }, + { + "left": "issue:43792", + "right": "issue:45901", + "accept": false, + "reason": "Whisper runtime failure vs table QA crash; unrelated." + }, + { + "left": "issue:43854", + "right": "issue:43866", + "accept": false, + "reason": "Unit-test load failure and corrupted checkpoint are not the same bug." + }, + { + "left": "issue:43425", + "right": "issue:45070", + "accept": false, + "reason": "Torch 2.10 incompatibility and pydantic PretrainedConfig breakage are different regressions." + }, + { + "left": "issue:43975", + "right": "issue:45901", + "accept": false, + "reason": "Detokenization bug vs table QA crash; no shared underlying change." + }, + { + "left": "issue:45362", + "right": "issue:45464", + "accept": false, + "reason": "Both involve Qwen3.5, but one is chat crashes and the other is streaming API failure; not clearly the same bug." + }, + { + "left": "issue:43792", + "right": "issue:45070", + "accept": false, + "reason": "Whisper cannot run and a pydantic config regression are unrelated." + }, + { + "left": "issue:43596", + "right": "issue:45800", + "accept": false, + "reason": "ZeRO-3 load index error vs torch 2.4.1 compatibility complaint are different issues." + }, + { + "left": "issue:43541", + "right": "issue:46869", + "accept": false, + "reason": "Different models and failure modes: torch dynamo/grouped_mm vs YOLO multi-GPU auto device-map." + }, + { + "left": "issue:44292", + "right": "issue:45901", + "accept": false, + "reason": "Qwen NVFP4 runtime failure vs table QA crash are unrelated." + }, + { + "left": "issue:39290", + "right": "issue:45823", + "accept": false, + "reason": "Different models and symptoms; no evidence of one shared bug." + }, + { + "left": "issue:45464", + "right": "issue:45685", + "accept": false, + "reason": "Chat/completions streaming failure and MPS histogram int error are unrelated." + }, + { + "left": "issue:36010", + "right": "issue:39290", + "accept": false, + "reason": "ImportError for GenerationMixin and Gemma3 config attribute missing are separate regressions." + }, + { + "left": "issue:43541", + "right": "issue:45685", + "accept": false, + "reason": "MoE-related titles are not enough; the concrete failures differ." + }, + { + "left": "issue:43541", + "right": "issue:43828", + "accept": false, + "reason": "Grouped_mm tracing error vs autocast dtype mismatch are different concrete bugs." + }, + { + "left": "issue:44841", + "right": "issue:45084", + "accept": false, + "reason": "Processor failure and template-node compile error are unrelated." + }, + { + "left": "issue:43404", + "right": "issue:43866", + "accept": false, + "reason": "Weight-tying bug in Mistral3 is unrelated to a corrupted checkpoint report." + }, + { + "left": "issue:45507", + "right": "issue:46869", + "accept": false, + "reason": "Invalid method call in GraniteMoEHybrid vs YOLO multi-GPU device-map failure are unrelated." + }, + { + "left": "issue:45399", + "right": "issue:46661", + "accept": false, + "reason": "Flash-Attn fallback checks and remote-code config_class errors are different code paths." + }, + { + "left": "issue:43572", + "right": "issue:45070", + "accept": false, + "reason": "Missing pad_token_idx in StableLmConfig is not the same as the pydantic PretrainedConfig field regression." + }, + { + "left": "issue:45800", + "right": "issue:45823", + "accept": false, + "reason": "Torch version compatibility and Gemma4 device mismatch are unrelated." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet, and the two accepted soft duplicates are conservative and well-supported by the issue titles: the ZeRO-3 index-out-of-bounds pair and the AutoTokenizer wrong-token-ID pair. The other verdicts stay cautious rather than overclaiming, and the canonical issue choice is reasonable." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 26, + "review_unit_index": 3, + "review_unit_count": 61, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-505", + "nodes": [ + "issue:39186", + "issue:40990", + "issue:41093", + "issue:42915", + "issue:43054", + "issue:43296", + "issue:43366", + "issue:43404", + "issue:43531", + "issue:43541", + "issue:43596", + "issue:43645", + "issue:43716", + "issue:43828", + "issue:43873", + "issue:43994", + "issue:44291", + "issue:44360", + "issue:44387", + "issue:44479", + "issue:44492", + "issue:44512", + "issue:44530", + "issue:44805", + "issue:44863", + "issue:44918", + "issue:44960", + "issue:45005", + "issue:45020", + "issue:45070", + "issue:45161", + "issue:45183", + "issue:45399", + "issue:45464", + "issue:45507", + "issue:45632", + "issue:45698", + "issue:45759", + "issue:45800", + "issue:45902", + "issue:46165", + "issue:46294", + "issue:46463", + "issue:46468", + "issue:46640", + "issue:46661", + "issue:46869" + ], + "soft_pairs": [ + "issue:43296|issue:43366", + "issue:44805|issue:44918", + "issue:43541|issue:45161", + "issue:43828|issue:45464", + "issue:44387|issue:45005", + "issue:43366|issue:45464", + "issue:44291|issue:45070", + "issue:43531|issue:45070", + "issue:41093|issue:43596", + "issue:43716|issue:46294", + "issue:45020|issue:45800", + "issue:43054|issue:43994", + "issue:44479|issue:46294", + "issue:46468|issue:46869", + "issue:45759|issue:45902", + "issue:43596|issue:46165", + "issue:46640|issue:46661", + "issue:43716|issue:44918", + "issue:39186|issue:44918", + "issue:40990|issue:44960", + "issue:43645|issue:45399", + "issue:45632|issue:45698", + "issue:43404|issue:45005", + "issue:42915|issue:45399", + "issue:43873|issue:45005", + "issue:44360|issue:44512", + "issue:44530|issue:45005", + "issue:44492|issue:44512", + "issue:44863|issue:45005", + "issue:45183|issue:46463", + "issue:44918|issue:45507" + ], + "prepared_review_unit_hash": "c3209c118ae64a33c897f277cdf4f5e57c6395584fe5f7a9ac3af1a0c402da11", + "component_budget": { + "node_count": 505, + "item_count": 505, + "soft_pair_count": 1854, + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 31, + "serialized_chars": 36237, + "estimated_input_tokens": 9060, + "estimated_eval_tokens": 18376 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is heterogeneous and does not look like a true duplicate set. The soft pairs mostly share broad symptoms or subsystem labels, but they describe different concrete bugs, models, or code paths, so none should be merged as duplicates.", + "confidence": 0.95, + "canonical_issue_reason": "No single issue is a safe canonical representative: the items span unrelated regressions (FSDP, MoE, video I/O, remote_code, quantization, docs, etc.) and the soft pairs do not match on a concrete underlying bug.", + "canonical_pr_reason": "No pull requests are present in this cluster.", + "best_issue_reason": "No issue is a good global best representative because the cluster is not centered on one bug; choosing any one would mischaracterize the rest of the items.", + "best_pr_reason": "No pull requests are present in this cluster.", + "soft_edge_verdicts": [ + { + "left": "issue:43296", + "right": "issue:43366", + "accept": false, + "reason": "Different problems: PaddleOCR-VL loading failure vs GGUF/gpt-oss architecture support." + }, + { + "left": "issue:44805", + "right": "issue:44918", + "accept": false, + "reason": "Both are shape-related errors, but one is a generic mask/tensor mismatch and the other is Qwen3.5 embedding unpacking with TRL SFT; not the same bug." + }, + { + "left": "issue:43541", + "right": "issue:45161", + "accept": false, + "reason": "Different failures: torch dynamo grouped_mm tracing on Mixtral vs tensor-parallel-only GPT-OSS MoE behavior." + }, + { + "left": "issue:43828", + "right": "issue:45464", + "accept": false, + "reason": "Different code paths and models: autocast dtype mismatch in Phi-tiny-MoE vs Qwen3.5 streaming chat/completions regression." + }, + { + "left": "issue:44387", + "right": "issue:45005", + "accept": false, + "reason": "OOM under int4 quantization is unrelated to tied-weight issues in translation models." + }, + { + "left": "issue:43366", + "right": "issue:45464", + "accept": false, + "reason": "GGUF gpt-oss support is a model-format request, not the Qwen3.5 streaming inference failure." + }, + { + "left": "issue:44291", + "right": "issue:45070", + "accept": false, + "reason": "Init-empty-weights/_is_hf_initialized TypeError is unrelated to the PretrainedConfig pydantic regression." + }, + { + "left": "issue:43531", + "right": "issue:45070", + "accept": false, + "reason": "Qwen3-MoE sliding_window behavior and pydantic PretrainedConfig breakage are different bugs." + }, + { + "left": "issue:41093", + "right": "issue:43596", + "accept": false, + "reason": "Both raise IndexError, but one is a mask/tensor size mismatch and the other is a DeepSpeed ZeRO-3/BertModel init failure." + }, + { + "left": "issue:43716", + "right": "issue:46294", + "accept": false, + "reason": "Different multimodal mismatches: image-preprocessor dtype issue vs video-token/feature count mismatch in Qwen3.5." + }, + { + "left": "issue:45020", + "right": "issue:45800", + "accept": false, + "reason": "Remote-code compatibility regression and torch 2.4.1/transformers 5.8.0 incompatibility are distinct issues." + }, + { + "left": "issue:43054", + "right": "issue:43994", + "accept": false, + "reason": "Both involve SigLIP2, but one is an embedding-quality complaint and the other is a loading/pipeline correctness bug." + }, + { + "left": "issue:44479", + "right": "issue:46294", + "accept": false, + "reason": "Broad video-input regression across several models is not the same concrete bug as Qwen3.5 chat_template video feature/token mismatch." + }, + { + "left": "issue:46468", + "right": "issue:46869", + "accept": false, + "reason": "Nemotron example breakage and YOLO multi-GPU device-map failure are unrelated." + }, + { + "left": "issue:45759", + "right": "issue:45902", + "accept": false, + "reason": "Composite Qwen config unwrapping and uninitialized inv_freq after to_empty(cuda) are different Qwen issues." + }, + { + "left": "issue:43596", + "right": "issue:46165", + "accept": false, + "reason": "DeepSpeed zero3 init failure with BertModel is unrelated to resuming training with trust_remote_code." + }, + { + "left": "issue:46640", + "right": "issue:46661", + "accept": false, + "reason": "Flash Attention 2 issues and remote-code config_class AttributeError are separate code paths." + }, + { + "left": "issue:43716", + "right": "issue:44918", + "accept": false, + "reason": "Mistral-3 image dtype mismatch and Qwen3.5 embedding unpacking with TRL SFT are unrelated." + }, + { + "left": "issue:39186", + "right": "issue:44918", + "accept": false, + "reason": "FSDP weight dimensionality error and Qwen3.5/TRL embedding unpacking are not the same bug." + }, + { + "left": "issue:40990", + "right": "issue:44960", + "accept": false, + "reason": "High perplexity on gpt-oss-20b and GLM5 are unrelated reports." + }, + { + "left": "issue:43645", + "right": "issue:45399", + "accept": false, + "reason": "Custom-model notebook initialization breakage and flash-attn2 fallback gating are different issues." + }, + { + "left": "issue:45632", + "right": "issue:45698", + "accept": false, + "reason": "Both concern remote-code/custom models, but one is cache-path collision and the other is wrong module selection after save/load; not the same concrete defect." + }, + { + "left": "issue:43404", + "right": "issue:45005", + "accept": false, + "reason": "Tied lm_head not tied in Mistral3 is different from tied-weight issues in translation models." + }, + { + "left": "issue:42915", + "right": "issue:45399", + "accept": false, + "reason": "Qwen3MoE with FineGrainedFP8Config and flash-attn2 fallback gating are unrelated." + }, + { + "left": "issue:43873", + "right": "issue:45005", + "accept": false, + "reason": "Quantization/offloading behavior and tied-weight handling are different problems." + }, + { + "left": "issue:44360", + "right": "issue:44512", + "accept": false, + "reason": "DSA indexer ReLU discussion and removed run-command documentation are unrelated." + }, + { + "left": "issue:44530", + "right": "issue:45005", + "accept": false, + "reason": "PagedAttentionCache linear_attention crash on Qwen3.5 is not the same as translation-model tied-weight bugs." + }, + { + "left": "issue:44492", + "right": "issue:44512", + "accept": false, + "reason": "A cache-strategy typo and outdated docs about the removed run command are unrelated." + }, + { + "left": "issue:44863", + "right": "issue:45005", + "accept": false, + "reason": "NemotronH checkpoint loading failure is unrelated to tied weights in translation models." + }, + { + "left": "issue:45183", + "right": "issue:46463", + "accept": false, + "reason": "Both are validation/error-message issues, but one is input_ids/inputs_embeds XOR logic and the other is T5 decoder-input validation; not the same bug." + }, + { + "left": "issue:44918", + "right": "issue:45507", + "accept": false, + "reason": "Qwen3.5 embedding unpacking with TRL SFT and GraniteMoEHybrid invalid-method calls 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": 26, + "review_unit_index": 4, + "review_unit_count": 61, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-505", + "nodes": [ + "issue:24643", + "issue:28530", + "issue:36331", + "issue:38175", + "issue:39692", + "issue:41553", + "issue:41762", + "issue:42915", + "issue:42947", + "issue:43054", + "issue:43493", + "issue:43643", + "issue:43644", + "issue:43782", + "issue:43824", + "issue:43828", + "issue:43856", + "issue:43872", + "issue:43881", + "issue:43883", + "issue:43950", + "issue:43975", + "issue:44292", + "issue:44387", + "issue:44403", + "issue:44451", + "issue:44488", + "issue:44534", + "issue:44589", + "issue:44910", + "issue:44928", + "issue:44960", + "issue:44991", + "issue:45020", + "issue:45399", + "issue:45520", + "issue:45663", + "issue:45701", + "issue:45859", + "issue:45902", + "issue:45920", + "issue:45987", + "issue:46489", + "issue:46491", + "issue:46858", + "issue:46899", + "issue:46932" + ], + "soft_pairs": [ + "issue:43975|issue:44292", + "issue:46489|issue:46491", + "issue:43950|issue:44534", + "issue:39692|issue:43054", + "issue:43644|issue:43950", + "issue:36331|issue:41762", + "issue:46899|issue:46932", + "issue:43872|issue:44589", + "issue:43828|issue:44928", + "issue:45920|issue:46489", + "issue:43054|issue:45701", + "issue:45987|issue:46858", + "issue:38175|issue:43054", + "issue:43950|issue:45902", + "issue:44910|issue:45859", + "issue:41553|issue:43883", + "issue:43054|issue:43493", + "issue:42947|issue:44387", + "issue:45663|issue:45902", + "issue:44403|issue:44451", + "issue:43975|issue:45520", + "issue:24643|issue:28530", + "issue:42947|issue:43856", + "issue:44403|issue:44488", + "issue:43643|issue:45020", + "issue:41762|issue:42915", + "issue:41553|issue:43881", + "issue:44991|issue:45399", + "issue:43782|issue:44960", + "issue:43824|issue:43872" + ], + "prepared_review_unit_hash": "c81dd76313996f9095ebf799cc756c799d481ee1021ec4d62e05eac36e008ab4", + "component_budget": { + "node_count": 505, + "item_count": 505, + "soft_pair_count": 1854, + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 30, + "serialized_chars": 36283, + "estimated_input_tokens": 9071, + "estimated_eval_tokens": 18398 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The cluster has a few solid duplicates around the same concrete regressions, especially the v5 non-persistent-buffer corruption and the StaticCache return-type bug. The SigLIP2 pair is related but not strong enough to merge as the same bug from the available titles alone: one is a broad quality regression and the other is a specific HF-vs-JAX discrepancy. Overall this looks like a mostly conservative cluster with a few clear matches, not a set of indisputable duplicates.", + "confidence": 0.74, + "canonical_issue_reason": "#43950 is the clearest anchor for the v5 non-persistent-buffer corruption regression, and its wording best covers the same issue seen in #43644/#44534.", + "canonical_pr_reason": null, + "best_issue_reason": "#43950 most directly names the concrete bug and regression scope, making it the best duplicate target among the issues here.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43975", + "right": "issue:44292", + "accept": false, + "reason": "Different models and different failure modes; the titles do not indicate one shared bug." + }, + { + "left": "issue:46489", + "right": "issue:46491", + "accept": false, + "reason": "Both are tokenizer regressions, but the affected models and concrete symptoms differ too much to treat as one bug." + }, + { + "left": "issue:43950", + "right": "issue:44534", + "accept": true, + "reason": "Same v5 regression corrupting non-persistent buffers / filling them with junk." + }, + { + "left": "issue:39692", + "right": "issue:43054", + "accept": false, + "reason": "Doc-example errors vs. a SigLIP2 quality regression; not the same concrete bug." + }, + { + "left": "issue:43644", + "right": "issue:43950", + "accept": true, + "reason": "Both report the same non-persistent-buffer junk/corruption regression in Transformers 5.x." + }, + { + "left": "issue:36331", + "right": "issue:41762", + "accept": false, + "reason": "Unrelated training API signature break vs. DeepSpeed loading index error." + }, + { + "left": "issue:46899", + "right": "issue:46932", + "accept": true, + "reason": "Same Gemma4 multimodal 4-bit quantization dtype-casting bug." + }, + { + "left": "issue:43872", + "right": "issue:44589", + "accept": false, + "reason": "Different quantization/storage errors affecting different code paths." + }, + { + "left": "issue:43828", + "right": "issue:44928", + "accept": false, + "reason": "Different models and different numerical issues; not the same failure." + }, + { + "left": "issue:45920", + "right": "issue:46489", + "accept": false, + "reason": "Tokenizer regressions in different model families, but not one concrete shared bug." + }, + { + "left": "issue:43054", + "right": "issue:45701", + "accept": false, + "reason": "SigLIP2 embedding-quality issue vs. a broad tokenization-change report; too different to merge." + }, + { + "left": "issue:45987", + "right": "issue:46858", + "accept": true, + "reason": "Same StaticCache get_seq_length return-type/behavior bug; one describes the contract violation, the other the downstream breakage." + }, + { + "left": "issue:38175", + "right": "issue:43054", + "accept": false, + "reason": "Related SigLIP2 symptom area, but no clear shared underlying bug from the titles alone." + }, + { + "left": "issue:43950", + "right": "issue:45902", + "accept": false, + "reason": "Both mention bad initialization-like behavior, but they are different code paths and different objects." + }, + { + "left": "issue:44910", + "right": "issue:45859", + "accept": false, + "reason": "Flash-Attention packed-sequence crash vs. CPU-offload incompatibility; unrelated issues." + }, + { + "left": "issue:41553", + "right": "issue:43883", + "accept": false, + "reason": "Bad error message vs. missing attribute; different failures." + }, + { + "left": "issue:43054", + "right": "issue:43493", + "accept": false, + "reason": "Related SigLIP2 area, but one is a broad text-embedding quality regression and the other is a specific HF-vs-JAX discrepancy; titles alone do not prove the same concrete bug." + }, + { + "left": "issue:42947", + "right": "issue:44387", + "accept": false, + "reason": "Different training/memory regressions with no shared concrete code-path." + }, + { + "left": "issue:45663", + "right": "issue:45902", + "accept": false, + "reason": "Different model internals and different memory/state bugs." + }, + { + "left": "issue:44403", + "right": "issue:44451", + "accept": false, + "reason": "Generic load-noise complaint vs. a specific model load failure; not the same bug." + }, + { + "left": "issue:43975", + "right": "issue:45520", + "accept": false, + "reason": "Tokenizer/detokenization issue vs. import-utils flash_attn KeyError; unrelated." + }, + { + "left": "issue:24643", + "right": "issue:28530", + "accept": false, + "reason": "DeepSpeed training runtime error vs. early-stopping metric configuration problem." + }, + { + "left": "issue:42947", + "right": "issue:43856", + "accept": false, + "reason": "Both are memory-related, but they describe different training settings and different failure modes." + }, + { + "left": "issue:44403", + "right": "issue:44488", + "accept": false, + "reason": "Both mention loading, but they concern different models and no common bug is clear." + }, + { + "left": "issue:43643", + "right": "issue:45020", + "accept": false, + "reason": "trust_remote_code missing-fields bug vs. a broad remote_code regression; too general to merge." + }, + { + "left": "issue:41762", + "right": "issue:42915", + "accept": false, + "reason": "Different models and different failure classes (ZeRO-3 load vs. FineGrainedFP8Config)." + }, + { + "left": "issue:41553", + "right": "issue:43881", + "accept": false, + "reason": "Different model load failures with no shared concrete cause." + }, + { + "left": "issue:44991", + "right": "issue:45399", + "accept": false, + "reason": "Tokenizer loading regression vs. flash-attn fallback logic; unrelated." + }, + { + "left": "issue:43782", + "right": "issue:44960", + "accept": false, + "reason": "Different model families and different errors; no shared code-path bug." + }, + { + "left": "issue:43824", + "right": "issue:43872", + "accept": false, + "reason": "Import error for a missing class vs. bitsandbytes constructor mismatch; unrelated." + } + ] + }, + "evaluator_result": { + "accept": false, + "feedback": "Most of the cluster calls are grounded, but the SigLIP2 accept (#43054/#43493) is a bit too strong for the packet: one issue is a broad quality regression and the other is a specific HF-vs-JAX discrepancy, and the titles alone do not prove they are the same bug. The summary\u2019s wording about \u2018genuine duplicate clusters\u2019 is therefore slightly overconfident. The other accepted pairs look conservative enough." + }, + "evaluator_used": true, + "retried": true, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 26, + "review_unit_index": 5, + "review_unit_count": 61, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-505", + "nodes": [ + "issue:24643", + "issue:30064", + "issue:36331", + "issue:41093", + "issue:41553", + "issue:42491", + "issue:42947", + "issue:43257", + "issue:43404", + "issue:43792", + "issue:43827", + "issue:43828", + "issue:43854", + "issue:43872", + "issue:43881", + "issue:43883", + "issue:43940", + "issue:43950", + "issue:44220", + "issue:44291", + "issue:44368", + "issue:44479", + "issue:44509", + "issue:44560", + "issue:44661", + "issue:44805", + "issue:44863", + "issue:44928", + "issue:44991", + "issue:45020", + "issue:45081", + "issue:45305", + "issue:45313", + "issue:45356", + "issue:45399", + "issue:45488", + "issue:45588", + "issue:45685", + "issue:45715", + "issue:45759", + "issue:45823", + "issue:45923", + "issue:46315", + "issue:46489", + "issue:46661", + "issue:46869", + "issue:46894", + "issue:46897" + ], + "soft_pairs": [ + "issue:45685|issue:45823", + "issue:45081|issue:45399", + "issue:45020|issue:45823", + "issue:44661|issue:45020", + "issue:43881|issue:43883", + "issue:43828|issue:45823", + "issue:42491|issue:44368", + "issue:45020|issue:45356", + "issue:45715|issue:46315", + "issue:43872|issue:46869", + "issue:45488|issue:46489", + "issue:45759|issue:46661", + "issue:43404|issue:43950", + "issue:43827|issue:44509", + "issue:46894|issue:46897", + "issue:43872|issue:44291", + "issue:45020|issue:46661", + "issue:43792|issue:44220", + "issue:43257|issue:45305", + "issue:42947|issue:45923", + "issue:36331|issue:41093", + "issue:41553|issue:44991", + "issue:43940|issue:45313", + "issue:43854|issue:44863", + "issue:44928|issue:45588", + "issue:41093|issue:44805", + "issue:44479|issue:44560", + "issue:24643|issue:30064" + ], + "prepared_review_unit_hash": "774bba1b5d32f5ab124400a05017e313cd4ee918d401efdf5f3206bf238b48d6", + "component_budget": { + "node_count": 505, + "item_count": 505, + "soft_pair_count": 1854, + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 28, + "serialized_chars": 36753, + "estimated_input_tokens": 9189, + "estimated_eval_tokens": 18634 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "Most pairs are only broad thematic matches, not true duplicates. Clear duplicate-looking groups appear for docs mentioning removed pipeline tasks, the `_is_hf_initialized` TypeError, Qwen3-family DeepSpeed ZeRO-3 load failures, the identical mask/tensor shape IndexError, and the training-loss double-shift bug.", + "confidence": 0.73, + "canonical_issue_reason": "Issue 44291 is the broadest/root-cause report for the `_is_hf_initialized` loading regression, making it the best canonical issue among the duplicates in this set.", + "canonical_pr_reason": null, + "best_issue_reason": "44291 is the most reusable representative because it describes the underlying TypeError generically rather than a single downstream symptom.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45685", + "right": "issue:45823", + "accept": false, + "reason": "Different failures: MPS histogram/int dtype error vs Gemma4 PLE device placement mismatch." + }, + { + "left": "issue:45081", + "right": "issue:45399", + "accept": false, + "reason": "Both involve tokenizer/model loading, but one is a Mistral regex crash and the other is flash-attn2 fallback gating; not the same bug." + }, + { + "left": "issue:45020", + "right": "issue:45823", + "accept": false, + "reason": "Remote-code model breakage is a different problem from a Gemma4 forward device mismatch." + }, + { + "left": "issue:44661", + "right": "issue:45020", + "accept": false, + "reason": "Add-new-model-like TOKENIZER_MAPPING_NAMES failure is unrelated to the broader remote-code regression." + }, + { + "left": "issue:43881", + "right": "issue:43883", + "accept": false, + "reason": "Different model-loading errors for different models; only loosely similar." + }, + { + "left": "issue:43828", + "right": "issue:45823", + "accept": false, + "reason": "Autocast dtype mismatch on Phi MoE is not the same as a device-map forward mismatch." + }, + { + "left": "issue:42491", + "right": "issue:44368", + "accept": false, + "reason": "Both mention LoRA/Qwen embedding tying, but they describe different symptoms and code paths." + }, + { + "left": "issue:45020", + "right": "issue:45356", + "accept": false, + "reason": "Broad remote-code breakage vs a specific Kimi-K2.5 tokenizer regression; not the same underlying bug." + }, + { + "left": "issue:45715", + "right": "issue:46315", + "accept": false, + "reason": "These are inverse tokenizer-performance regressions; similar pattern, but not the same defect." + }, + { + "left": "issue:43872", + "right": "issue:46869", + "accept": false, + "reason": "Bitsandbytes/`_is_hf_initialized` incompatibility is unrelated to YOLO multi-GPU device-map failure." + }, + { + "left": "issue:45488", + "right": "issue:46489", + "accept": false, + "reason": "Both concern tokenizer regressions in DeepSeek-family models, but the described bugs and code paths differ." + }, + { + "left": "issue:45759", + "right": "issue:46661", + "accept": false, + "reason": "Both involve model/config loading APIs, but one is `text_config` unwrapping and the other is missing `config_class` on remote-code models." + }, + { + "left": "issue:43404", + "right": "issue:43950", + "accept": false, + "reason": "Untied lm_head weights and silent non-persistent buffer corruption are different regressions." + }, + { + "left": "issue:43827", + "right": "issue:44509", + "accept": true, + "reason": "Same docs regression: both say removed pipeline tasks are still referenced after v5 removals." + }, + { + "left": "issue:46894", + "right": "issue:46897", + "accept": true, + "reason": "Same underlying training-loss double-shift bug, reported for two model families." + }, + { + "left": "issue:43872", + "right": "issue:44291", + "accept": true, + "reason": "Same `_is_hf_initialized` TypeError during model loading; 44291 gives the generic root-cause framing." + }, + { + "left": "issue:45020", + "right": "issue:46661", + "accept": false, + "reason": "Remote-code models failing generally is broader than the specific missing `config_class` AttributeError." + }, + { + "left": "issue:43792", + "right": "issue:44220", + "accept": false, + "reason": "Whisper model run failure and `_torch_extract_fbank_features()` are not the same issue." + }, + { + "left": "issue:43257", + "right": "issue:45305", + "accept": false, + "reason": "Both involve DeepSpeed/Qwen, but one is weight conversion and the other is gradient averaging; different bugs." + }, + { + "left": "issue:42947", + "right": "issue:45923", + "accept": false, + "reason": "Gradient checkpointing with PEFT LoRA is unrelated to a missing supports_gradient_checkpointing flag on a trust_remote_code variant." + }, + { + "left": "issue:36331", + "right": "issue:41093", + "accept": false, + "reason": "Unexpected `num_items_in_batch` kwarg is unrelated to the mask/index shape mismatch." + }, + { + "left": "issue:41553", + "right": "issue:44991", + "accept": false, + "reason": "Both are tokenizer-related, but the Voxtral error-message issue and est-roberta tokenizer loading failure are different bugs." + }, + { + "left": "issue:43940", + "right": "issue:45313", + "accept": true, + "reason": "Same Qwen3-family DeepSpeed ZeRO-3 weight-loading failure; different variants of the same missing-weights path." + }, + { + "left": "issue:43854", + "right": "issue:44863", + "accept": false, + "reason": "GLM-4.7-Flash test loading failure and NemotronH checkpoint loading failure are unrelated." + }, + { + "left": "issue:44928", + "right": "issue:45588", + "accept": false, + "reason": "RLHF gradient explosion from 3D position_ids is unrelated to flash_attention s_aux handling." + }, + { + "left": "issue:41093", + "right": "issue:44805", + "accept": true, + "reason": "Same mask-vs-indexed-tensor shape mismatch error; 44805 is the generic follow-up report." + }, + { + "left": "issue:44479", + "right": "issue:44560", + "accept": false, + "reason": "Both are video-input regressions, but they affect different model families and failure modes." + }, + { + "left": "issue:24643", + "right": "issue:30064", + "accept": false, + "reason": "DeepSpeed training weight-shape runtime error is unrelated to void segmentation map processing." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The grouping is grounded in the packet and the soft-edge verdicts are suitably conservative. The accepted pairs are defensible as same-bug or same-regression matches, and the rejected pairs avoid overstating similarity." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 26, + "review_unit_index": 6, + "review_unit_count": 61, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-505", + "nodes": [ + "issue:36010", + "issue:38175", + "issue:41553", + "issue:42491", + "issue:42617", + "issue:42915", + "issue:43366", + "issue:43531", + "issue:43550", + "issue:43596", + "issue:43632", + "issue:43638", + "issue:43645", + "issue:43646", + "issue:43716", + "issue:43720", + "issue:43824", + "issue:43828", + "issue:43950", + "issue:43957", + "issue:43994", + "issue:44155", + "issue:44488", + "issue:44530", + "issue:44560", + "issue:44589", + "issue:44661", + "issue:44683", + "issue:44805", + "issue:44863", + "issue:44877", + "issue:44910", + "issue:44918", + "issue:45030", + "issue:45325", + "issue:45584", + "issue:45663", + "issue:45672", + "issue:45685", + "issue:45812", + "issue:45823", + "issue:45834", + "issue:45902", + "issue:46402", + "issue:46489", + "issue:46515", + "issue:46858", + "issue:46899" + ], + "soft_pairs": [ + "issue:44530|issue:44863", + "issue:41553|issue:44488", + "issue:44863|issue:46402", + "issue:43366|issue:43828", + "issue:45325|issue:45902", + "issue:43366|issue:43531", + "issue:43596|issue:46515", + "issue:44155|issue:45584", + "issue:45663|issue:45823", + "issue:43638|issue:46515", + "issue:43366|issue:44589", + "issue:43950|issue:45834", + "issue:43824|issue:43957", + "issue:42915|issue:43957", + "issue:43716|issue:44560", + "issue:44910|issue:46858", + "issue:43645|issue:43646", + "issue:43550|issue:44910", + "issue:44560|issue:45685", + "issue:42617|issue:43366", + "issue:36010|issue:42915", + "issue:44560|issue:44918", + "issue:44683|issue:46515", + "issue:45812|issue:46489", + "issue:45823|issue:46515", + "issue:44661|issue:45823", + "issue:38175|issue:43994", + "issue:44560|issue:44805", + "issue:45672|issue:46899", + "issue:43632|issue:45834", + "issue:44877|issue:45030", + "issue:43638|issue:44661", + "issue:42491|issue:43720", + "issue:43596|issue:44661" + ], + "prepared_review_unit_hash": "e74c5f95845bfd863fc633d7c50b7e5f56721e755ef12dfb8d6e5280f7681009", + "component_budget": { + "node_count": 505, + "item_count": 505, + "soft_pair_count": 1854, + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 34, + "serialized_chars": 38059, + "estimated_input_tokens": 9515, + "estimated_eval_tokens": 19286 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The cluster is mostly heterogeneous. Only two soft-edge pairs are accepted as true duplicates-like matches: SigLIP2 inference regression (38175/43994) and the Transformers 5 custom-model initialization regression (43645/43646). The Bert/DeepSpeed zero3 issues are not supported as a duplicate pair.", + "confidence": 0.91, + "canonical_issue_reason": "issue:43646 is the best canonical issue because it states the broad Transformers 5 custom-model initialization regression in a generic, reusable way; issue:43645 is the narrower notebook-flavored manifestation.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43646 is the strongest cluster representative: concise, generic, and centered on the underlying regression rather than one environment-specific symptom.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44530", + "right": "issue:44863", + "accept": false, + "reason": "Different bugs: Qwen3.5 PagedAttentionCache linear_attention crash vs NemotronH checkpoint loading failure." + }, + { + "left": "issue:41553", + "right": "issue:44488", + "accept": false, + "reason": "Both are loading problems, but they concern different models and different failure modes." + }, + { + "left": "issue:44863", + "right": "issue:46402", + "accept": false, + "reason": "Different model-loading regressions: NemotronH checkpoint compatibility vs Sam3Model text encoder weights not loading." + }, + { + "left": "issue:43366", + "right": "issue:43828", + "accept": false, + "reason": "Unrelated issues: gpt-oss GGUF support request vs Phi-tiny-MoE autocast dtype mismatch." + }, + { + "left": "issue:45325", + "right": "issue:45902", + "accept": false, + "reason": "Different Qwen internals and different defects: rope index scaling vs inv_freq materialization after meta->to_empty(cuda)." + }, + { + "left": "issue:43366", + "right": "issue:43531", + "accept": false, + "reason": "Different problems: gpt-oss support request vs sliding_window bug in Qwen3-MoE." + }, + { + "left": "issue:43596", + "right": "issue:46515", + "accept": false, + "reason": "Bert/DeepSpeed zero3 IndexError is a different bug from a generic torch 2.4.1 incompatibility report." + }, + { + "left": "issue:44155", + "right": "issue:45584", + "accept": false, + "reason": "Audio-related but distinct code paths: track-to-track token leak vs empty-transcription handling after align_special_tokens." + }, + { + "left": "issue:45663", + "right": "issue:45823", + "accept": false, + "reason": "Same model family, but different bugs: per-layer shared_kv_states rebuild under FSDP2 vs device mismatch with device_map=\"auto\"." + }, + { + "left": "issue:43638", + "right": "issue:46515", + "accept": false, + "reason": "The Bert zero3 training IndexError is not the same as a generic torch version incompatibility issue." + }, + { + "left": "issue:43366", + "right": "issue:44589", + "accept": false, + "reason": "Unrelated: gpt-oss GGUF support vs Float8_e4m3fnStorage lookup failure." + }, + { + "left": "issue:43950", + "right": "issue:45834", + "accept": false, + "reason": "Different regressions: non-persistent buffer corruption in from_pretrained() vs Kosmos2.5 long-ocr index error." + }, + { + "left": "issue:43824", + "right": "issue:43957", + "accept": false, + "reason": "Different loading failures: missing Qwen2_5_VL export vs meta-device loading breakage on some models." + }, + { + "left": "issue:42915", + "right": "issue:43957", + "accept": false, + "reason": "Different regressions: Qwen3Moe FineGrainedFP8Config failure vs meta-device model loading breakage." + }, + { + "left": "issue:43716", + "right": "issue:44560", + "accept": false, + "reason": "Different models and failure modes: Mistral-3 image-preprocessor dtype mismatch vs Qwen3-vl-embedding video StopIteration." + }, + { + "left": "issue:44910", + "right": "issue:46858", + "accept": false, + "reason": "Both involve cache/positions, but these are distinct bugs with different triggers and backends." + }, + { + "left": "issue:43645", + "right": "issue:43646", + "accept": true, + "reason": "Same Transformers 5 custom-model initialization regression; 43645 is the notebook-specific symptom of the broader bug in 43646." + }, + { + "left": "issue:43550", + "right": "issue:44910", + "accept": false, + "reason": "Different Qwen/Bamba failures: torch.compile with SDPA on Bamba-9B-v2 vs flash-attention crash from Qwen3.5 3D position_ids handling." + }, + { + "left": "issue:44560", + "right": "issue:45685", + "accept": false, + "reason": "Different platform/model bugs: Qwen3-vl video StopIteration vs MPS histogram not implemented for Int." + }, + { + "left": "issue:42617", + "right": "issue:43366", + "accept": false, + "reason": "3d_parallel.py execution failure is unrelated to gpt-oss GGUF architecture support." + }, + { + "left": "issue:36010", + "right": "issue:42915", + "accept": false, + "reason": "ImportError for GenerationMixin and Qwen3Moe FineGrainedFP8Config failure are unrelated bugs." + }, + { + "left": "issue:44560", + "right": "issue:44918", + "accept": false, + "reason": "Different Qwen3-related failures: video StopIteration vs TRL SFT trainer embedding unpacking." + }, + { + "left": "issue:44683", + "right": "issue:46515", + "accept": false, + "reason": "Both are torch-version-related, but the concrete failures are different." + }, + { + "left": "issue:45812", + "right": "issue:46489", + "accept": false, + "reason": "Both are tokenizer regressions, but they affect different model families and appear to have different root causes." + }, + { + "left": "issue:45823", + "right": "issue:46515", + "accept": false, + "reason": "Gemma4 device mismatch with device_map=\"auto\" is a different bug from a generic torch 2.4.1 incompatibility." + }, + { + "left": "issue:44661", + "right": "issue:45823", + "accept": false, + "reason": "Tokenizer-mapping/add-new-model-like failure is unrelated to Gemma4 PLE device mismatch." + }, + { + "left": "issue:38175", + "right": "issue:43994", + "accept": true, + "reason": "Same SigLIP2 inference regression on google/siglip2-base-patch16-224; \"zero probabilities\" and \"nonsensical results\" point to the same underlying bug." + }, + { + "left": "issue:44560", + "right": "issue:44805", + "accept": false, + "reason": "Different failures: Qwen3-vl embedding video StopIteration vs a generic mask-shape IndexError." + }, + { + "left": "issue:45672", + "right": "issue:46899", + "accept": false, + "reason": "Both are Gemma4 quantization-related, but they are different concrete bugs in different modules." + }, + { + "left": "issue:43632", + "right": "issue:45834", + "accept": false, + "reason": "_is_hf_initialized flag regression and Kosmos2.5 long OCR index error are different issues." + }, + { + "left": "issue:44877", + "right": "issue:45030", + "accept": false, + "reason": "Both are config-validation problems, but they target different models and different validation changes." + }, + { + "left": "issue:43638", + "right": "issue:44661", + "accept": false, + "reason": "Bert zero3 IndexError and tokenizer-mapping add-new-model-like failure are unrelated defects." + }, + { + "left": "issue:42491", + "right": "issue:43720", + "accept": false, + "reason": "Different subsystems and failure modes: Qwen3_MOE LoRA incompatibility across hf4.x/hf5.x vs BitNet packed-weight loading in accelerate." + }, + { + "left": "issue:43596", + "right": "issue:44661", + "accept": false, + "reason": "Different bugs: Bert zero3 IndexError vs tokenizer-mapping/add-new-model-like failure." + } + ] + }, + "evaluator_result": { + "accept": false, + "feedback": "The analysis is mostly careful on the soft-edge verdicts, but the summary overstates the evidence by claiming \u201cthree true duplicate-like issue pairs.\u201d In the packet, only two pairs are actually accepted as duplicates (38175/43994 and 43645/43646); the Bert/DeepSpeed zero3 issues are not supported as a true duplicate pair. Please keep the cluster summary tighter and grounded in the explicit soft-edge outcomes." + }, + "evaluator_used": true, + "retried": true, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 26, + "review_unit_index": 7, + "review_unit_count": 61, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-505", + "nodes": [ + "issue:38175", + "issue:39692", + "issue:41720", + "issue:42491", + "issue:42915", + "issue:43257", + "issue:43317", + "issue:43644", + "issue:43645", + "issue:43708", + "issue:43854", + "issue:43856", + "issue:43873", + "issue:43931", + "issue:43994", + "issue:44075", + "issue:44164", + "issue:44291", + "issue:44387", + "issue:44451", + "issue:44488", + "issue:44534", + "issue:44661", + "issue:44756", + "issue:44805", + "issue:44863", + "issue:44910", + "issue:44960", + "issue:44977", + "issue:44991", + "issue:45276", + "issue:45292", + "issue:45313", + "issue:45335", + "issue:45357", + "issue:45362", + "issue:45406", + "issue:45488", + "issue:45507", + "issue:45588", + "issue:45812", + "issue:46050", + "issue:46459", + "issue:46468", + "issue:46512", + "issue:46846", + "issue:46897", + "issue:46901" + ], + "soft_pairs": [ + "issue:43317|issue:44756", + "issue:44661|issue:44805", + "issue:38175|issue:39692", + "issue:43644|issue:44534", + "issue:41720|issue:42915", + "issue:42915|issue:44756", + "issue:42491|issue:45276", + "issue:44960|issue:44977", + "issue:44451|issue:44488", + "issue:44863|issue:46846", + "issue:46459|issue:46512", + "issue:45313|issue:45357", + "issue:45507|issue:46468", + "issue:44488|issue:44991", + "issue:43644|issue:43645", + "issue:43257|issue:43854", + "issue:39692|issue:43994", + "issue:43856|issue:43873", + "issue:43854|issue:43931", + "issue:42491|issue:45292", + "issue:43856|issue:44387", + "issue:43708|issue:44075", + "issue:42491|issue:45335", + "issue:44960|issue:46050", + "issue:44164|issue:44291", + "issue:46897|issue:46901", + "issue:45362|issue:45406", + "issue:45488|issue:45812", + "issue:45276|issue:45335", + "issue:44910|issue:45588" + ], + "prepared_review_unit_hash": "a6d6ec9bbb8a7d90ec4c081687246d952bb5285ff923b1be74ac67d4e68f1ddc", + "component_budget": { + "node_count": 505, + "item_count": 505, + "soft_pair_count": 1854, + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 30, + "serialized_chars": 36654, + "estimated_input_tokens": 9164, + "estimated_eval_tokens": 18584 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "Most items are distinct model- or subsystem-specific regressions. I only accept a few soft duplicates where the titles point to the same concrete bug: non-persistent buffers being junk-filled in v5, resize_token_embeddings not propagating to output embeddings in related model variants, and the training-loss double-shift bug.", + "confidence": 0.9, + "canonical_issue_reason": "issue:43644 is the clearest, most concrete regression report in the set and has the exact duplicate-style paraphrase in issue:44534.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43644 is the best representative issue here because it states the bug precisely and matches the closest duplicate candidate without ambiguity.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43317", + "right": "issue:44756", + "accept": false, + "reason": "Different failures: device_map/offload loading error vs mmap/OOM mitigation on Strix Halo." + }, + { + "left": "issue:44661", + "right": "issue:44805", + "accept": false, + "reason": "Unrelated bugs: tokenizer mapping/model addition logic vs mask shape mismatch at runtime." + }, + { + "left": "issue:38175", + "right": "issue:39692", + "accept": false, + "reason": "Same model family, but one is zero probabilities and the other is doc-example/model-processor/quantization errors; not the same concrete bug." + }, + { + "left": "issue:43644", + "right": "issue:44534", + "accept": true, + "reason": "Same regression: both report Transformers v5 filling non-persistent buffers with junk." + }, + { + "left": "issue:41720", + "right": "issue:42915", + "accept": false, + "reason": "Different Qwen3 issues with different failure modes and code paths." + }, + { + "left": "issue:42915", + "right": "issue:44756", + "accept": false, + "reason": "Unrelated: Qwen3 MoE FP8 failure vs mmap/OOM on Strix Halo." + }, + { + "left": "issue:42491", + "right": "issue:45276", + "accept": false, + "reason": "Different problems: LoRA compatibility across versions vs resize_token_embeddings propagation." + }, + { + "left": "issue:44960", + "right": "issue:44977", + "accept": false, + "reason": "Both are GLM/Qwen-related, but the reported failures are different model/runtime issues." + }, + { + "left": "issue:44451", + "right": "issue:44488", + "accept": false, + "reason": "Both are model-loading complaints, but they concern different checkpoints and no shared concrete root cause is evident." + }, + { + "left": "issue:44863", + "right": "issue:46846", + "accept": false, + "reason": "Different models and different failures: NemotronH checkpoint loading vs missing linear_attn TP plan entries." + }, + { + "left": "issue:46459", + "right": "issue:46512", + "accept": false, + "reason": "Related around dtype handling, but one is loading composite checkpoints and the other is from_config/config-state mismatch." + }, + { + "left": "issue:45313", + "right": "issue:45357", + "accept": false, + "reason": "Different bugs: DeepSpeed ZeRO-3 loading vs incorrect visual encoder keys on save_pretrained." + }, + { + "left": "issue:45507", + "right": "issue:46468", + "accept": false, + "reason": "Different model implementations and different invalid-method/example failures." + }, + { + "left": "issue:44488", + "right": "issue:44991", + "accept": false, + "reason": "Both are tokenizer/model loading issues, but they affect different repos and no same underlying bug is shown." + }, + { + "left": "issue:43644", + "right": "issue:43645", + "accept": false, + "reason": "One is the non-persistent-buffer junk regression; the other is Jupyter custom-model initialization, a different issue." + }, + { + "left": "issue:43257", + "right": "issue:43854", + "accept": false, + "reason": "Different loading paths and failure symptoms: accelerate+deepspeed MoE weight conversion vs unit-test model loading." + }, + { + "left": "issue:39692", + "right": "issue:43994", + "accept": false, + "reason": "Same SigLIP2 family, but doc-example mistakes are not the same as the nonsensical-output regression." + }, + { + "left": "issue:43856", + "right": "issue:43873", + "accept": false, + "reason": "Both mention memory/offloading around Qwen3 MoE, but they are different problems with different concrete causes." + }, + { + "left": "issue:43854", + "right": "issue:43931", + "accept": false, + "reason": "Different models and errors: GLM-4.7 test loading vs Qwen3-VL shape mismatch." + }, + { + "left": "issue:42491", + "right": "issue:45292", + "accept": false, + "reason": "Different underlying bugs; only loosely related by model state/embedding behavior." + }, + { + "left": "issue:43856", + "right": "issue:44387", + "accept": false, + "reason": "Different memory issues: training efficiency vs int4 reserved-memory/OOM regression." + }, + { + "left": "issue:43708", + "right": "issue:44075", + "accept": false, + "reason": "Trainer checkpoint-resume math bug is unrelated to SGD optimizer args not being used." + }, + { + "left": "issue:42491", + "right": "issue:45335", + "accept": false, + "reason": "LoRA/version compatibility issue is unrelated to resize_token_embeddings behavior." + }, + { + "left": "issue:44960", + "right": "issue:46050", + "accept": false, + "reason": "Different bugs: GLM5 model issue vs quantized model save/load corruption." + }, + { + "left": "issue:44164", + "right": "issue:44291", + "accept": false, + "reason": "Different save/load failures: extra_state handling vs unexpected _is_hf_initialized argument under init_empty_weights." + }, + { + "left": "issue:46897", + "right": "issue:46901", + "accept": true, + "reason": "Same bug pattern: training loss is being double-shifted instead of using the labels as-is." + }, + { + "left": "issue:45362", + "right": "issue:45406", + "accept": false, + "reason": "Different components and failures: chat crash for Qwen3.5-35B vs Gemma4Processor missing _tokenizer in serve." + }, + { + "left": "issue:45488", + "right": "issue:45812", + "accept": false, + "reason": "Both are tokenizer regressions, but the affected families and failure mechanisms differ." + }, + { + "left": "issue:45276", + "right": "issue:45335", + "accept": true, + "reason": "Same concrete bug: resize_token_embeddings does not propagate updated embeddings to output/decoder embeddings across related model variants." + }, + { + "left": "issue:44910", + "right": "issue:45588", + "accept": false, + "reason": "Both involve flash attention, but the crashes come from different triggers and code paths." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet and the conservative rejections are mostly well supported. The accepted soft duplicates also look plausibly mergeable/same-bug from the titles: non-persistent buffer junk in v5, resize_token_embeddings propagation, and the training-loss double-shift pattern. No clear 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": 26, + "review_unit_index": 8, + "review_unit_count": 61, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-505", + "nodes": [ + "issue:36010", + "issue:39404", + "issue:39692", + "issue:41720", + "issue:42796", + "issue:42947", + "issue:43335", + "issue:43381", + "issue:43404", + "issue:43425", + "issue:43454", + "issue:43645", + "issue:43646", + "issue:43708", + "issue:43828", + "issue:43854", + "issue:43873", + "issue:43957", + "issue:44079", + "issue:44155", + "issue:44164", + "issue:44291", + "issue:44292", + "issue:44387", + "issue:44402", + "issue:44479", + "issue:44488", + "issue:44756", + "issue:44863", + "issue:44877", + "issue:44912", + "issue:44928", + "issue:44960", + "issue:45216", + "issue:45276", + "issue:45292", + "issue:45335", + "issue:45588", + "issue:45663", + "issue:45684", + "issue:45698", + "issue:45704", + "issue:45759", + "issue:45823", + "issue:45859", + "issue:46326", + "issue:46468", + "issue:46829" + ], + "soft_pairs": [ + "issue:44402|issue:44488", + "issue:39692|issue:44960", + "issue:41720|issue:44155", + "issue:43425|issue:43957", + "issue:45663|issue:45759", + "issue:42947|issue:45704", + "issue:43404|issue:43454", + "issue:44291|issue:44756", + "issue:43873|issue:44387", + "issue:44079|issue:46829", + "issue:45663|issue:45859", + "issue:44292|issue:44912", + "issue:36010|issue:39404", + "issue:44155|issue:45823", + "issue:44387|issue:44756", + "issue:42947|issue:43381", + "issue:43957|issue:44756", + "issue:43335|issue:45663", + "issue:44863|issue:46468", + "issue:45276|issue:45292", + "issue:44479|issue:44877", + "issue:43381|issue:44928", + "issue:44164|issue:44756", + "issue:42796|issue:46326", + "issue:45588|issue:45663", + "issue:45684|issue:45698", + "issue:43854|issue:45216", + "issue:45292|issue:45335", + "issue:43645|issue:44164", + "issue:43646|issue:43708", + "issue:43828|issue:43957" + ], + "prepared_review_unit_hash": "fa648452b1ce020897d1f1356ae4d7ae281ace42b515367be977d381644b4292", + "component_budget": { + "node_count": 505, + "item_count": 505, + "soft_pair_count": 1854, + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 31, + "serialized_chars": 37088, + "estimated_input_tokens": 9272, + "estimated_eval_tokens": 18800 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly a grab-bag of unrelated Transformers bugs across different models and subsystems. The only clearly duplicate-like family is the resize_token_embeddings/output_embeddings reports, with the generic issue serving as the best representative. No PRs are present.", + "confidence": 0.31, + "canonical_issue_reason": "Issue 45292 is the most generic and reusable description of the resize_token_embeddings bug, and it best represents the close duplicate-like Gemma4/output_embeddings reports.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 45292 has the broadest framing of the underlying symptom and would be the best standalone tracker for the resize_token_embeddings family.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44402", + "right": "issue:44488", + "accept": false, + "reason": "Different problems: token vocab mismatch vs model loading failure for a specific checkpoint." + }, + { + "left": "issue:39692", + "right": "issue:44960", + "accept": false, + "reason": "Unrelated model/docs issue versus a GLM5 loading/title-only issue." + }, + { + "left": "issue:41720", + "right": "issue:44155", + "accept": false, + "reason": "Different code paths and symptoms: Qwen3 device mapping CUDA assert vs AudioFlamingo batched inference leak." + }, + { + "left": "issue:43425", + "right": "issue:43957", + "accept": false, + "reason": "Torch version compatibility and meta-device loading failures are separate issues." + }, + { + "left": "issue:45663", + "right": "issue:45759", + "accept": false, + "reason": "Different model families and bugs: Gemma4 FSDP2 attention state vs Qwen multimodal config unwrapping." + }, + { + "left": "issue:42947", + "right": "issue:45704", + "accept": false, + "reason": "Gradient checkpointing/LoRA behavior is unrelated to T5 apex RMSNorm memory leak." + }, + { + "left": "issue:43404", + "right": "issue:43454", + "accept": false, + "reason": "Both mention lm_head tying, but they are different model implementations and likely separate fixes." + }, + { + "left": "issue:44291", + "right": "issue:44756", + "accept": false, + "reason": "init_empty_weights argument handling and Strix Halo mmap/OOM are unrelated." + }, + { + "left": "issue:43873", + "right": "issue:44387", + "accept": false, + "reason": "Both involve quantization/memory, but the concrete failure modes differ too much to treat as one bug." + }, + { + "left": "issue:44079", + "right": "issue:46829", + "accept": false, + "reason": "ModelOutput key assignment and VideoPrism output fields are different issues." + }, + { + "left": "issue:45663", + "right": "issue:45859", + "accept": false, + "reason": "Different bugs in different models: Gemma4 FSDP2 vs Qwen3_5MoE CPU offload incompatibility." + }, + { + "left": "issue:44292", + "right": "issue:44912", + "accept": false, + "reason": "Different quantization formats and failure modes: NVFP4 load error vs MXFP4 fallback to bf16." + }, + { + "left": "issue:36010", + "right": "issue:39404", + "accept": false, + "reason": "ImportError for GenerationMixin is unrelated to Whisper return_language pipeline regression." + }, + { + "left": "issue:44155", + "right": "issue:45823", + "accept": false, + "reason": "AudioFlamingo batched inference leak and Gemma4 device_map mismatch are separate problems." + }, + { + "left": "issue:44387", + "right": "issue:44756", + "accept": false, + "reason": "Both can end in OOM, but one is int4 quantization memory growth and the other is Strix Halo mmap behavior." + }, + { + "left": "issue:42947", + "right": "issue:43381", + "accept": false, + "reason": "Both involve gradient checkpointing, but one is ineffective under PEFT LoRA and the other is invalid in eval mode." + }, + { + "left": "issue:43957", + "right": "issue:44756", + "accept": false, + "reason": "Meta-device loading and mmap-related OOM on Strix Halo are unrelated." + }, + { + "left": "issue:43335", + "right": "issue:45663", + "accept": false, + "reason": "SwitchTransformers sparse layer construction and Gemma4 FSDP2 attention state are unrelated." + }, + { + "left": "issue:44863", + "right": "issue:46468", + "accept": false, + "reason": "NemotronH checkpoint loading and a broken official Nemotron example are not the same concrete bug." + }, + { + "left": "issue:45276", + "right": "issue:45292", + "accept": true, + "reason": "These are effectively the same resize_token_embeddings/output_embeddings bug, with 45276 being a model-specific variant of the generic issue." + }, + { + "left": "issue:44479", + "right": "issue:44877", + "accept": false, + "reason": "Different model/regression contexts: video input regression vs strict config blocking granite_speech loading." + }, + { + "left": "issue:43381", + "right": "issue:44928", + "accept": false, + "reason": "Gradient checkpointing in eval mode is unrelated to the Qwen3.5 RLHF NaN/SDPA fallback issue." + }, + { + "left": "issue:44164", + "right": "issue:44756", + "accept": false, + "reason": "save/from_pretrained extra_state handling and Strix Halo mmap/OOM are unrelated." + }, + { + "left": "issue:42796", + "right": "issue:46326", + "accept": false, + "reason": "Tokenizer BPE-marker handling and continue_final_message behavior are different tokenizer bugs." + }, + { + "left": "issue:45588", + "right": "issue:45663", + "accept": false, + "reason": "flash_attention sink-less crash and Gemma4 FSDP2 shared_kv state are unrelated." + }, + { + "left": "issue:45684", + "right": "issue:45698", + "accept": false, + "reason": "Permissions propagation during save_pretrained and loading the wrong custom module are separate save/load bugs." + }, + { + "left": "issue:43854", + "right": "issue:45216", + "accept": false, + "reason": "Model load failure in tests and Qwen3.5 save_pretrained regression do not match." + }, + { + "left": "issue:45292", + "right": "issue:45335", + "accept": false, + "reason": "They are related in theme, but 45335 is a T5Gemma-specific decoder.embed_tokens symptom and not clearly the same concrete bug as the generic output_embeddings report." + }, + { + "left": "issue:43645", + "right": "issue:44164", + "accept": false, + "reason": "Custom model initialization in Jupyter and extra_state save/load handling are separate problems." + }, + { + "left": "issue:43646", + "right": "issue:43708", + "accept": false, + "reason": "Transformers 5.0 custom model init breakage and Trainer max_steps recalculation are unrelated." + }, + { + "left": "issue:43828", + "right": "issue:43957", + "accept": false, + "reason": "Autocast dtype mismatch in Phi-tiny-MoE and meta-device loading failures are different issues." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet and stays conservative. The only clearly related family identified is the resize_token_embeddings/output_embeddings cluster, and the accepted 45276\u201345292 soft-edge verdict is supported by the titles and high similarity. No PR-pair mergeability issues arise because there are no PRs in 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": 26, + "review_unit_index": 9, + "review_unit_count": 61, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-505", + "nodes": [ + "issue:36032", + "issue:39404", + "issue:43299", + "issue:43425", + "issue:43452", + "issue:43611", + "issue:43645", + "issue:43716", + "issue:43828", + "issue:43856", + "issue:43901", + "issue:43906", + "issue:43931", + "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:44898", + "issue:44912", + "issue:44991", + "issue:45081", + "issue:45084", + "issue:45479", + "issue:45701", + "issue:45759", + "issue:45812", + "issue:45823", + "issue:46050", + "issue:46315", + "issue:46326", + "issue:46491", + "issue:46665", + "issue:46821", + "issue:46869", + "issue:46894", + "issue:46897", + "issue:46899", + "issue:46901" + ], + "soft_pairs": [ + "issue:44402|issue:44451", + "issue:46894|issue:46901", + "issue:44589|issue:44756", + "issue:43425|issue:43828", + "issue:46315|issue:46491", + "issue:45812|issue:46315", + "issue:39404|issue:45701", + "issue:44530|issue:45084", + "issue:43611|issue:43645", + "issue:43425|issue:46869", + "issue:39404|issue:46665", + "issue:44683|issue:45823", + "issue:43425|issue:43716", + "issue:43425|issue:45823", + "issue:44479|issue:44805", + "issue:44589|issue:45084", + "issue:43906|issue:44530", + "issue:43901|issue:44509", + "issue:43994|issue:44079", + "issue:36032|issue:43452", + "issue:44265|issue:44898", + "issue:39404|issue:44912", + "issue:46050|issue:46326", + "issue:44164|issue:45759", + "issue:43906|issue:46326", + "issue:44671|issue:44991", + "issue:44991|issue:45081", + "issue:45823|issue:46899", + "issue:43299|issue:43931", + "issue:45479|issue:46894", + "issue:43856|issue:44756", + "issue:44164|issue:46821", + "issue:45479|issue:46897" + ], + "prepared_review_unit_hash": "a5cb1f0edeeab89741f5362db5ab851149cc2a6f802923f38e3530ead78d7117", + "component_budget": { + "node_count": 505, + "item_count": 505, + "soft_pair_count": 1854, + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 33, + "serialized_chars": 37758, + "estimated_input_tokens": 9440, + "estimated_eval_tokens": 19136 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The candidates are mostly broad, thematically similar regressions, but they do not look like the same concrete bug. I would keep them separate; the only close pair is the two training-loss double-shift reports, but they still appear model-specific rather than one shared code-path fix.", + "confidence": 0.82, + "canonical_issue_reason": "issue:46901 is the most concrete, technically actionable issue in the set: it states the exact label-shift mistake (`labels[..., 1:]` vs `labels`).", + "canonical_pr_reason": null, + "best_issue_reason": "issue:46901 is the best representative issue because it is precise, reproducible, and directly points to the faulty loss computation; the other nearby reports are broader or more model-specific.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44402", + "right": "issue:44451", + "accept": false, + "reason": "Both are tokenizer-loading problems, but one is a vocab-size metadata bug and the other is a specific model load failure; no evidence they share the same root cause." + }, + { + "left": "issue:46894", + "right": "issue:46901", + "accept": false, + "reason": "They describe the same symptom family, but on different model families with no proof of a shared code path; treat as related reports, not one duplicate bug." + }, + { + "left": "issue:44589", + "right": "issue:44756", + "accept": false, + "reason": "Completely different failures: missing Float8 storage vs mmap/OOM behavior on Strix Halo." + }, + { + "left": "issue:43425", + "right": "issue:43828", + "accept": false, + "reason": "One is a broad Torch 2.10 compatibility complaint, the other is a specific autocast dtype mismatch in Phi-tiny-MoE." + }, + { + "left": "issue:46315", + "right": "issue:46491", + "accept": false, + "reason": "Both touch tokenization, but one is a slowdown regression and the other is a round-trip correctness bug; different underlying issues." + }, + { + "left": "issue:45812", + "right": "issue:46315", + "accept": false, + "reason": "Granite wrong token IDs and slow `convert_tokens_to_ids` are distinct tokenization regressions." + }, + { + "left": "issue:39404", + "right": "issue:45701", + "accept": false, + "reason": "Whisper pipeline `return_language` and a generic tokenization change are unrelated behaviors." + }, + { + "left": "issue:44530", + "right": "issue:45084", + "accept": false, + "reason": "Different problems: Qwen3.5 cache group-type crash vs a template compilation TypeError." + }, + { + "left": "issue:43611", + "right": "issue:43645", + "accept": false, + "reason": "Both are v5 load/init regressions, but one is about `base_model_prefix` loading and the other about custom models in Jupyter; not the same bug." + }, + { + "left": "issue:43425", + "right": "issue:46869", + "accept": false, + "reason": "Torch version incompatibility and multi-GPU `device_map=\"auto\"` failure are unrelated." + }, + { + "left": "issue:39404", + "right": "issue:46665", + "accept": false, + "reason": "Whisper pipeline regression is unrelated to inconsistent `from_pretrained` behavior inside Python classes." + }, + { + "left": "issue:44683", + "right": "issue:45823", + "accept": false, + "reason": "Compiled flex_attention on torch >= 2.9 and Gemma4 device-mapping mismatch are different code-path failures." + }, + { + "left": "issue:43425", + "right": "issue:43716", + "accept": false, + "reason": "A Torch compatibility complaint is not the same as an image preprocessor/model dtype mismatch." + }, + { + "left": "issue:43425", + "right": "issue:45823", + "accept": false, + "reason": "These are different regressions in different subsystems; shared dependency version alone is not enough." + }, + { + "left": "issue:44479", + "right": "issue:44805", + "accept": false, + "reason": "Both are runtime errors, but one is a Qwen video-input regression and the other is a mask shape mismatch; no shared root cause is evident." + }, + { + "left": "issue:44589", + "right": "issue:45084", + "accept": false, + "reason": "Unrelated failures: missing Float8 storage vs Mistral tokenizer regex patch compilation." + }, + { + "left": "issue:43906", + "right": "issue:44530", + "accept": false, + "reason": "A reproduction of an earlier issue and a Qwen3.5 cache crash are not the same bug." + }, + { + "left": "issue:43901", + "right": "issue:44509", + "accept": false, + "reason": "Both concern docs, but they document different removed/changed pipeline behaviors; not a duplicate." + }, + { + "left": "issue:43994", + "right": "issue:44079", + "accept": false, + "reason": "SigLIP2 nonsensical outputs and `ModelOutput` key assignment are unrelated." + }, + { + "left": "issue:36032", + "right": "issue:43452", + "accept": false, + "reason": "Tokenizer method-name conflict and gguf loading breakage are different failures." + }, + { + "left": "issue:44265", + "right": "issue:44898", + "accept": false, + "reason": "`torch.export.export` failing under `torch_compilable_check` is not the same as Perceiver non-default-resolution failure." + }, + { + "left": "issue:39404", + "right": "issue:44912", + "accept": false, + "reason": "Whisper pipeline behavior and MXFP4 quantization fallback are unrelated." + }, + { + "left": "issue:46050", + "right": "issue:46326", + "accept": false, + "reason": "Quantized-model save/load corruption and a tokenizer `continue_final_message` bug do not share a concrete code path." + }, + { + "left": "issue:44164", + "right": "issue:45759", + "accept": false, + "reason": "Extra state handling in save/from_pretrained is unrelated to `text_config` unwrapping for multimodal configs." + }, + { + "left": "issue:43906", + "right": "issue:46326", + "accept": false, + "reason": "The earlier issue reproduction and the mistral_common tokenizer bug are distinct." + }, + { + "left": "issue:44671", + "right": "issue:44991", + "accept": false, + "reason": "CamemBERT masked-LM prediction drift and an Estonian RoBERTa tokenizer load failure are different issues." + }, + { + "left": "issue:44991", + "right": "issue:45081", + "accept": false, + "reason": "Both involve tokenizer loading, but one is a model-specific tokenizer failure and the other is a Mistral regex-patching crash." + }, + { + "left": "issue:45823", + "right": "issue:46899", + "accept": false, + "reason": "Same model family, but one is a device-map mismatch and the other is a 4-bit quantization dtype-cast bug." + }, + { + "left": "issue:43299", + "right": "issue:43931", + "accept": false, + "reason": "Both involve Qwen3-VL loading, but the reported failure modes differ: dev version breakage vs weight-shape mismatch." + }, + { + "left": "issue:45479", + "right": "issue:46894", + "accept": false, + "reason": "A degenerate single-label classification loss bug is unrelated to the ASR training-loss double-shift report." + }, + { + "left": "issue:43856", + "right": "issue:44756", + "accept": false, + "reason": "Qwen3 MoE training memory usage and Strix Halo mmap OOM are different problems." + }, + { + "left": "issue:44164", + "right": "issue:46821", + "accept": false, + "reason": "Saving/loading extra_state and a TimesFM `window_size` AttributeError are unrelated." + }, + { + "left": "issue:45479", + "right": "issue:46897", + "accept": false, + "reason": "Sequence-classification zero-loss behavior and Florence2 training-loss double-shift are different bugs, even if both concern loss computation." + } + ] + }, + "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": 26, + "review_unit_index": 10, + "review_unit_count": 61, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-505", + "nodes": [ + "issue:34567", + "issue:36032", + "issue:39404", + "issue:40444", + "issue:40990", + "issue:42175", + "issue:42796", + "issue:42915", + "issue:43317", + "issue:43404", + "issue:43421", + "issue:43452", + "issue:43482", + "issue:43541", + "issue:43582", + "issue:43596", + "issue:43611", + "issue:43632", + "issue:43643", + "issue:43644", + "issue:43645", + "issue:43646", + "issue:43716", + "issue:43746", + "issue:43883", + "issue:44206", + "issue:44220", + "issue:44451", + "issue:44462", + "issue:44488", + "issue:44530", + "issue:44661", + "issue:44671", + "issue:44683", + "issue:44737", + "issue:44811", + "issue:44987", + "issue:44991", + "issue:45081", + "issue:45084", + "issue:45507", + "issue:45584", + "issue:45859", + "issue:46178", + "issue:46285", + "issue:46326", + "issue:46402", + "issue:46489" + ], + "soft_pairs": [ + "issue:43596|issue:44683", + "issue:44488|issue:44987", + "issue:44488|issue:45081", + "issue:43611|issue:43644", + "issue:42796|issue:43404", + "issue:44462|issue:46489", + "issue:36032|issue:42915", + "issue:46285|issue:46402", + "issue:43716|issue:45084", + "issue:39404|issue:44671", + "issue:44737|issue:45859", + "issue:43643|issue:44530", + "issue:44661|issue:44683", + "issue:42915|issue:44530", + "issue:40990|issue:43421", + "issue:44451|issue:44987", + "issue:43541|issue:45084", + "issue:34567|issue:40444", + "issue:43644|issue:43646", + "issue:43582|issue:43632", + "issue:40444|issue:46326", + "issue:42175|issue:46178", + "issue:44206|issue:44220", + "issue:36032|issue:43645", + "issue:44811|issue:45584", + "issue:43452|issue:43482", + "issue:44987|issue:44991", + "issue:45084|issue:45507", + "issue:43317|issue:43482", + "issue:43883|issue:44488", + "issue:43746|issue:46326" + ], + "prepared_review_unit_hash": "bcf95be2e34c28ba27587ebc250e9ff99173dfc7dd6c599c70b46b16100cc974", + "component_budget": { + "node_count": 505, + "item_count": 505, + "soft_pair_count": 1854, + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 31, + "serialized_chars": 37015, + "estimated_input_tokens": 9254, + "estimated_eval_tokens": 18764 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "These are mostly unrelated closed issues spanning tokenizer loading, model init, multimodal, and backend-specific regressions. Only one soft pair looks plausibly duplicate; the rest are too semantically different to merge safely.", + "confidence": 0.58, + "canonical_issue_reason": "Issue 44462 is the most representative and concrete candidate: it describes a broad tokenizer-loading regression, is fairly specific, and has follow-up references pointing to the same area.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 44462 is the strongest standalone representative of the cluster because it is actionable, generalized, and directly about a core loading path rather than a model-specific symptom.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43596", + "right": "issue:44683", + "accept": false, + "reason": "Different failures in different code paths: DeepSpeed ZeRO init vs flex_attention compilation on newer torch." + }, + { + "left": "issue:44488", + "right": "issue:44987", + "accept": false, + "reason": "Both concern model loading, but they point to different repositories/models and likely different root causes." + }, + { + "left": "issue:44488", + "right": "issue:45081", + "accept": false, + "reason": "Repo-specific tokenizer loading failure vs a Mistral regex patch crash; not the same bug." + }, + { + "left": "issue:43611", + "right": "issue:43644", + "accept": false, + "reason": "Both are v5 regressions, but one is base_model_prefix loading and the other is non-persistent buffer initialization; too different." + }, + { + "left": "issue:42796", + "right": "issue:43404", + "accept": false, + "reason": "Tokenizer BPE-marker handling and tied lm_head weights are unrelated issues in Mistral3." + }, + { + "left": "issue:44462", + "right": "issue:46489", + "accept": false, + "reason": "Both are tokenizer-related, but one is about AutoTokenizer ignoring tokenizer.json and the other is a DeepSeek-Coder output regression." + }, + { + "left": "issue:36032", + "right": "issue:42915", + "accept": false, + "reason": "T5 tokenizer method conflict vs Qwen3Moe FP8 config failure; unrelated subsystems." + }, + { + "left": "issue:46285", + "right": "issue:46402", + "accept": false, + "reason": "CLIP model structure change vs Sam3 text-encoder weights not loading; different models and symptoms." + }, + { + "left": "issue:43716", + "right": "issue:45084", + "accept": false, + "reason": "Image preprocessor dtype mismatch and template compilation failure are unrelated." + }, + { + "left": "issue:39404", + "right": "issue:44671", + "accept": false, + "reason": "Whisper pipeline language return regression vs CamemBERT masked-LM prediction bug; no shared bug." + }, + { + "left": "issue:44737", + "right": "issue:45859", + "accept": false, + "reason": "XLNet CPU placement issue and Qwen3_5Moe CPU-offload incompatibility are different implementations." + }, + { + "left": "issue:43643", + "right": "issue:44530", + "accept": false, + "reason": "AutoConfig trust_remote_code field loss vs PagedAttentionCache linear_attention crash; unrelated." + }, + { + "left": "issue:44661", + "right": "issue:44683", + "accept": false, + "reason": "Model registration logic and flex_attention compilation failure do not share a concrete bug." + }, + { + "left": "issue:42915", + "right": "issue:44530", + "accept": false, + "reason": "Different Qwen-family failures with different error modes and likely different fixes." + }, + { + "left": "issue:40990", + "right": "issue:43421", + "accept": false, + "reason": "Perplexity regression on a model/dataset benchmark vs runtime special-token post-processor sync bug." + }, + { + "left": "issue:44451", + "right": "issue:44987", + "accept": false, + "reason": "Both are loading failures, but the affected models and likely root causes differ." + }, + { + "left": "issue:43541", + "right": "issue:45084", + "accept": false, + "reason": "Torch dynamo grouped_mm tracing failure vs template-node compilation failure; distinct compiler issues." + }, + { + "left": "issue:34567", + "right": "issue:40444", + "accept": false, + "reason": "Trainer token-count tracking and IterableDataset multimodal finetuning failure are not the same bug." + }, + { + "left": "issue:43644", + "right": "issue:43646", + "accept": true, + "reason": "Both describe the same v5 model-initialization regression: custom model initialization is broken, and non-persistent buffers being filled with junk is a specific manifestation of it." + }, + { + "left": "issue:43582", + "right": "issue:43632", + "accept": false, + "reason": "Apple Silicon allocator warmup TypeError and _is_hf_initialized flag breakage are unrelated." + }, + { + "left": "issue:40444", + "right": "issue:46326", + "accept": false, + "reason": "IterableDataset multimodal finetuning failure vs mistral_common continue_final_message bug; different code paths." + }, + { + "left": "issue:42175", + "right": "issue:46178", + "accept": false, + "reason": "Backend packaging/install issue and missing BACKENDS_MAPPING entry are different problems." + }, + { + "left": "issue:44206", + "right": "issue:44220", + "accept": false, + "reason": "Both audio-related, but one is an unsupported center argument and the other is a fbank feature extraction issue." + }, + { + "left": "issue:36032", + "right": "issue:43645", + "accept": false, + "reason": "Tokenizer method-name conflict and Jupyter custom model initialization breakage are not the same issue." + }, + { + "left": "issue:44811", + "right": "issue:45584", + "accept": false, + "reason": "Whisper decoding skip_special_tokens bug vs empty transcription after align_special_tokens; related area, but not the same failure." + }, + { + "left": "issue:43452", + "right": "issue:43482", + "accept": false, + "reason": "gguf_file loading breakage and Qwen2.5-GGUF v5 loading failure are different model-loading problems." + }, + { + "left": "issue:44987", + "right": "issue:44991", + "accept": false, + "reason": "Both are tokenizer/model loading regressions, but they affect different models and likely different fixes." + }, + { + "left": "issue:45084", + "right": "issue:45507", + "accept": false, + "reason": "Template compilation error vs GraniteMoEHybrid invalid method call; unrelated." + }, + { + "left": "issue:43317", + "right": "issue:43482", + "accept": false, + "reason": "Dequantized model offload failure vs Qwen2.5-GGUF loading regression; different bugs." + }, + { + "left": "issue:43883", + "right": "issue:44488", + "accept": false, + "reason": "Missing tied-weights attribute on a model class vs inability to load a specific tokenizer/model; unrelated." + }, + { + "left": "issue:43746", + "right": "issue:46326", + "accept": false, + "reason": "PEFT local-checkpoint loading failure vs mistral_common message-format bug; different surfaces and fixes." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The analysis is grounded in the packet and the soft-edge judgments are conservative. The one accepted pair, issue:43644 and issue:43646, looks plausibly mergeable as the same v5 custom model initialization regression with one being a specific symptom. The rest of the pairs are rejected with cautious, bug-specific reasons. The representative issue choice is also reasonable and not overstated." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 26, + "review_unit_index": 11, + "review_unit_count": 61, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-505", + "nodes": [ + "issue:28530", + "issue:34567", + "issue:36032", + "issue:41720", + "issue:42915", + "issue:43089", + "issue:43208", + "issue:43262", + "issue:43322", + "issue:43377", + "issue:43404", + "issue:43531", + "issue:43540", + "issue:43632", + "issue:43645", + "issue:43646", + "issue:43742", + "issue:43792", + "issue:43819", + "issue:43883", + "issue:43994", + "issue:44008", + "issue:44062", + "issue:44265", + "issue:44291", + "issue:44403", + "issue:44451", + "issue:44464", + "issue:44466", + "issue:44479", + "issue:44534", + "issue:44589", + "issue:44610", + "issue:44849", + "issue:44877", + "issue:44987", + "issue:44991", + "issue:45081", + "issue:45084", + "issue:45216", + "issue:45325", + "issue:45357", + "issue:45987", + "issue:46315", + "issue:46821", + "issue:46869", + "issue:46899", + "issue:47013" + ], + "soft_pairs": [ + "issue:43262|issue:43377", + "issue:43208|issue:47013", + "issue:43646|issue:44534", + "issue:43645|issue:44534", + "issue:44403|issue:45081", + "issue:43792|issue:46869", + "issue:42915|issue:45084", + "issue:43377|issue:43819", + "issue:44849|issue:45325", + "issue:43404|issue:44466", + "issue:44464|issue:45987", + "issue:44987|issue:45081", + "issue:28530|issue:34567", + "issue:43994|issue:44610", + "issue:42915|issue:44291", + "issue:44008|issue:46899", + "issue:44451|issue:45081", + "issue:43883|issue:44991", + "issue:36032|issue:43742", + "issue:44062|issue:44987", + "issue:44265|issue:44610", + "issue:43632|issue:46821", + "issue:43089|issue:45987", + "issue:43322|issue:43540", + "issue:42915|issue:44589", + "issue:44479|issue:45325", + "issue:36032|issue:41720", + "issue:43531|issue:44877", + "issue:45216|issue:45357", + "issue:44991|issue:46315" + ], + "prepared_review_unit_hash": "303e391b421c229000056ee4bdd5050b8e258923ca0bd3fde8a16cf971d263bf", + "component_budget": { + "node_count": 505, + "item_count": 505, + "soft_pair_count": 1854, + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 30, + "serialized_chars": 36927, + "estimated_input_tokens": 9232, + "estimated_eval_tokens": 18720 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The cluster is mostly heterogeneous, with one clear duplicate pair around Qwen3.5 save_pretrained checkpoint corruption. The rest are distinct bugs across different models, subsystems, and failure modes, so they should stay separate.", + "confidence": 0.71, + "canonical_issue_reason": "issue:45216 is the broader, earlier report of the Qwen3.5 save_pretrained regression; issue:45357 looks like a later, narrower manifestation focused on incorrect visual encoder keys.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45216 is the best representative of the only true duplicate pair because it captures the broader checkpoint-writing bug, while issue:45357 is a more specific follow-up.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43262", + "right": "issue:43377", + "accept": false, + "reason": "Both are audio-related, but one is a chat-template sampling-rate default issue and the other is missing padding-mask support causing batched/single mismatches." + }, + { + "left": "issue:43208", + "right": "issue:47013", + "accept": false, + "reason": "Both mention xLSTM, but one is a set of training blockers for small models and the other is a specific return_last_states unpack crash." + }, + { + "left": "issue:43646", + "right": "issue:44534", + "accept": false, + "reason": "Both are v5/custom-model regressions, but one is notebook/custom init breakage and the other is non-persistent buffer junk serialization." + }, + { + "left": "issue:43645", + "right": "issue:44534", + "accept": false, + "reason": "Same broad v5 area, but different concrete failures: Jupyter custom model init versus buffer state corruption." + }, + { + "left": "issue:44403", + "right": "issue:45081", + "accept": false, + "reason": "Unnecessary loading noise is unrelated to a tokenizer regex patch crash in Mistral." + }, + { + "left": "issue:43792", + "right": "issue:46869", + "accept": false, + "reason": "Different models and different code paths: Whisper load failure versus YOLO multi-GPU device-map failure." + }, + { + "left": "issue:42915", + "right": "issue:45084", + "accept": false, + "reason": "One is a FineGrainedFP8Config failure and the other is a template-compilation TypeError; they are not the same bug." + }, + { + "left": "issue:43377", + "right": "issue:43819", + "accept": false, + "reason": "Both are audio model issues, but one is batching/padding-mask behavior and the other is DAC latent reconstruction mismatch." + }, + { + "left": "issue:44849", + "right": "issue:45325", + "accept": false, + "reason": "Both are Qwen-family regressions, but one affects output_hidden_states and the other affects VL rope index position scaling." + }, + { + "left": "issue:43404", + "right": "issue:44466", + "accept": false, + "reason": "Both touch tied weights, but one is a model-class weight-tying bug and the other is device-dependent save_pretrained serialization." + }, + { + "left": "issue:44464", + "right": "issue:45987", + "accept": false, + "reason": "Different generation issues: compiled-forward chunked generation inconsistency versus StaticCache returning the wrong type." + }, + { + "left": "issue:44987", + "right": "issue:45081", + "accept": false, + "reason": "Tokenizer loading regression for one model is unrelated to the Mistral backend_tokenizer crash." + }, + { + "left": "issue:28530", + "right": "issue:34567", + "accept": false, + "reason": "Early stopping metric lookup and num_input_tokens_seen updates are unrelated training-state problems." + }, + { + "left": "issue:43994", + "right": "issue:44610", + "accept": false, + "reason": "Different multimodal preprocessing bugs: wrong AutoModel/pipeline outputs versus wrong processor image size." + }, + { + "left": "issue:42915", + "right": "issue:44291", + "accept": false, + "reason": "Both involve loading/init, but one is FP8 configuration failure and the other is an unexpected _is_hf_initialized argument in init_empty_weights." + }, + { + "left": "issue:44008", + "right": "issue:46899", + "accept": false, + "reason": "Different Gemma multimodal failures: variable-name collision versus unsafe dtype casting under 4-bit quantization." + }, + { + "left": "issue:44451", + "right": "issue:45081", + "accept": false, + "reason": "A generic model load failure is not the same as a Mistral tokenizer patching crash." + }, + { + "left": "issue:43883", + "right": "issue:44991", + "accept": false, + "reason": "Molmo tied-weights attribute error and EMBEDDIA tokenizer loading failure are unrelated." + }, + { + "left": "issue:36032", + "right": "issue:43742", + "accept": false, + "reason": "Tokenizer special-token construction and MobileLLM key loading are different failure modes." + }, + { + "left": "issue:44062", + "right": "issue:44987", + "accept": false, + "reason": "AddedToken kwarg duplication is unrelated to the physical-intelligence/fast loading regression." + }, + { + "left": "issue:44265", + "right": "issue:44610", + "accept": false, + "reason": "torch.export export failure and OmDet-Turbo processor size mismatch do not share the same underlying bug." + }, + { + "left": "issue:43632", + "right": "issue:46821", + "accept": false, + "reason": "_is_hf_initialized/custom-model breakage is unrelated to the TimesFM window_size attribute error." + }, + { + "left": "issue:43089", + "right": "issue:45987", + "accept": false, + "reason": "Generation overhead from syncs/dispatch is a performance issue, not the same as a cache-length type contract bug." + }, + { + "left": "issue:43322", + "right": "issue:43540", + "accept": false, + "reason": "Different multimodal bugs: Llava Next loading segfault versus Qwen3OmniMoe video-input processing ValueError." + }, + { + "left": "issue:42915", + "right": "issue:44589", + "accept": false, + "reason": "FineGrainedFP8Config failure and missing Float8 storage lookup are both fp8-adjacent but different concrete errors." + }, + { + "left": "issue:44479", + "right": "issue:45325", + "accept": false, + "reason": "One is a video-input regression across several Qwen models; the other is a still-image position-id scaling bug." + }, + { + "left": "issue:36032", + "right": "issue:41720", + "accept": false, + "reason": "Tokenizer add_special_tokens conflict is unrelated to Qwen3 auto device-mapping cudaErrorAssert." + }, + { + "left": "issue:43531", + "right": "issue:44877", + "accept": false, + "reason": "Qwen3-MoE sliding_window behavior and granite_speech strict-config loading are unrelated." + }, + { + "left": "issue:45216", + "right": "issue:45357", + "accept": true, + "reason": "Same underlying Qwen3.5 save_pretrained checkpoint corruption; 45357 is a narrower report about incorrect visual encoder keys." + }, + { + "left": "issue:44991", + "right": "issue:46315", + "accept": false, + "reason": "Tokenizer loading failure and slow convert_tokens_to_ids in slow tokenizers are different bugs." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet: the titles support a mostly heterogeneous cluster with one plausible duplicate pair, issue:45216 and issue:45357, both about Qwen3.5 save_pretrained checkpoint/key corruption. The rejected soft pairs are phrased conservatively and stay close to the specific failure modes in the titles, without overstating the evidence." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 26, + "review_unit_index": 12, + "review_unit_count": 61, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-505", + "nodes": [ + "issue:33666", + "issue:35707", + "issue:40444", + "issue:40990", + "issue:41720", + "issue:42175", + "issue:42796", + "issue:43296", + "issue:43317", + "issue:43377", + "issue:43381", + "issue:43421", + "issue:43425", + "issue:43482", + "issue:43525", + "issue:43531", + "issue:43611", + "issue:43632", + "issue:43646", + "issue:43716", + "issue:43746", + "issue:43792", + "issue:43856", + "issue:44079", + "issue:44220", + "issue:44291", + "issue:44403", + "issue:44462", + "issue:44534", + "issue:44568", + "issue:44589", + "issue:44683", + "issue:44756", + "issue:44811", + "issue:44991", + "issue:45081", + "issue:45325", + "issue:45491", + "issue:45538", + "issue:45542", + "issue:45663", + "issue:45812", + "issue:45923", + "issue:46491", + "issue:46710", + "issue:46821", + "issue:46899" + ], + "soft_pairs": [ + "issue:43425|issue:44403", + "issue:43716|issue:46899", + "issue:40444|issue:45325", + "issue:43611|issue:44534", + "issue:43296|issue:43531", + "issue:44220|issue:44683", + "issue:45812|issue:46491", + "issue:42796|issue:43746", + "issue:44403|issue:44991", + "issue:43381|issue:45923", + "issue:43632|issue:44756", + "issue:33666|issue:40444", + "issue:40990|issue:44811", + "issue:43792|issue:44589", + "issue:44291|issue:44589", + "issue:42796|issue:46710", + "issue:44079|issue:45663", + "issue:35707|issue:40444", + "issue:43482|issue:43525", + "issue:43632|issue:44534", + "issue:43531|issue:44291", + "issue:43421|issue:44568", + "issue:43482|issue:45538", + "issue:42175|issue:45542", + "issue:41720|issue:45081", + "issue:44079|issue:44534", + "issue:43317|issue:43856", + "issue:44462|issue:45812", + "issue:43296|issue:46821", + "issue:43611|issue:43646", + "issue:43377|issue:45491" + ], + "prepared_review_unit_hash": "b1cf07fe662836f58dfb6255d49094051ea743edffe0e31061f4b124ef94cdac", + "component_budget": { + "node_count": 505, + "item_count": 505, + "soft_pair_count": 1854, + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 31, + "serialized_chars": 36386, + "estimated_input_tokens": 9097, + "estimated_eval_tokens": 18450 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The soft pairs are mostly false positives: they share broad themes like tokenizers, v5 regressions, or multimodal training, but each looks like a different underlying bug. No pair is clearly the same concrete fix target, so I would not merge any of them as duplicates.", + "confidence": 0.93, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43425", + "right": "issue:44403", + "accept": false, + "reason": "One is a Torch 2.10 compatibility problem; the other is noisy model-loading behavior. Same broad area, different bug." + }, + { + "left": "issue:43716", + "right": "issue:46899", + "accept": false, + "reason": "Both mention dtype issues in multimodal paths, but they involve different models and different failure points/casts." + }, + { + "left": "issue:40444", + "right": "issue:45325", + "accept": false, + "reason": "Different Qwen2.5-VL bugs: iterable dataset multi-image finetuning vs still-image rope/position-id scaling." + }, + { + "left": "issue:43611", + "right": "issue:44534", + "accept": false, + "reason": "Both are Transformers v5 regressions, but one is model loading/base_model_prefix and the other is buffer initialization garbage." + }, + { + "left": "issue:43296", + "right": "issue:43531", + "accept": false, + "reason": "Different subsystems: vLLM loading failure for PaddleOCR-VL vs Qwen3-MoE sliding_window behavior." + }, + { + "left": "issue:44220", + "right": "issue:44683", + "accept": false, + "reason": "Audio feature extraction bug vs compiled flex_attention failure; unrelated code paths." + }, + { + "left": "issue:45812", + "right": "issue:46491", + "accept": false, + "reason": "Both are tokenizer regressions, but one is AutoTokenizer model-ID selection and the other is a leading-space round-trip issue in CodeLlama." + }, + { + "left": "issue:42796", + "right": "issue:43746", + "accept": false, + "reason": "Tokenizer BPE marker handling vs GraniteSpeech PEFT checkpoint loading; unrelated." + }, + { + "left": "issue:44403", + "right": "issue:44991", + "accept": false, + "reason": "Generic loading noise is not the same as a specific tokenizer-loading failure for EMBEDDIA/est-roberta." + }, + { + "left": "issue:43381", + "right": "issue:45923", + "accept": false, + "reason": "Eval-mode gradient checkpointing error vs missing supports_gradient_checkpointing flag on a trust_remote_code model." + }, + { + "left": "issue:43632", + "right": "issue:44756", + "accept": false, + "reason": "Different failures: _is_hf_initialized regression vs mmap OOM workaround." + }, + { + "left": "issue:33666", + "right": "issue:40444", + "accept": false, + "reason": "Different Qwen2-VL vs Qwen2.5-VL training issues with no concrete shared bug." + }, + { + "left": "issue:40990", + "right": "issue:44811", + "accept": false, + "reason": "Perplexity anomaly on gpt-oss-20b is unrelated to Whisper batch_decode skip_special_tokens handling." + }, + { + "left": "issue:43792", + "right": "issue:44589", + "accept": false, + "reason": "Whisper runtime failure vs missing Float8 storage type; different causes." + }, + { + "left": "issue:44291", + "right": "issue:44589", + "accept": false, + "reason": "init_empty_weights/_is_hf_initialized regression is distinct from Float8 storage lookup failure." + }, + { + "left": "issue:42796", + "right": "issue:46710", + "accept": false, + "reason": "Tokenizer BPE marker bug vs downstream accuracy regression after upgrade; not the same concrete defect." + }, + { + "left": "issue:44079", + "right": "issue:45663", + "accept": false, + "reason": "ModelOutput key assignment bug is unrelated to Gemma4 FSDP2 shared_kv_states KeyError." + }, + { + "left": "issue:35707", + "right": "issue:40444", + "accept": false, + "reason": "Progressive generation with inputs_embeds/past_key_values is unrelated to multimodal iterable-dataset finetuning." + }, + { + "left": "issue:43482", + "right": "issue:43525", + "accept": false, + "reason": "GGUF loading failure in Transformers v5 and missing pad_token_id on Llama4Config are different bugs." + }, + { + "left": "issue:43632", + "right": "issue:44534", + "accept": false, + "reason": "Both are v5 regressions, but one concerns initialization flags and the other non-persistent buffer values." + }, + { + "left": "issue:43531", + "right": "issue:44291", + "accept": false, + "reason": "Qwen3-MoE sliding_window behavior is unrelated to init_empty_weights argument handling." + }, + { + "left": "issue:43421", + "right": "issue:44568", + "accept": false, + "reason": "Runtime post-processor sync for special tokens is different from add_special_tokens not inserting BOS/EOS." + }, + { + "left": "issue:43482", + "right": "issue:45538", + "accept": false, + "reason": "GGUF loading failure and CLIPTokenizer max_length default are unrelated." + }, + { + "left": "issue:42175", + "right": "issue:45542", + "accept": false, + "reason": "Both involve TensorFlow optional deps, but one is missing backend install metadata and the other is a runtime undefined backend error." + }, + { + "left": "issue:41720", + "right": "issue:45081", + "accept": false, + "reason": "Qwen3 auto device mapping cudaErrorAssert is unrelated to Mistral tokenizer regex patch crashing." + }, + { + "left": "issue:44079", + "right": "issue:44534", + "accept": false, + "reason": "ModelOutput key handling and non-persistent buffer junk are distinct Transformer v5 regressions." + }, + { + "left": "issue:43317", + "right": "issue:43856", + "accept": false, + "reason": "Dequantized model loading on gpu+cpu offload is unrelated to Qwen3 MoE training memory use." + }, + { + "left": "issue:44462", + "right": "issue:45812", + "accept": false, + "reason": "Both are AutoTokenizer regressions, but one is ignoring tokenizer.json while the other is wrong IDs on Granite models; not clearly the same bug." + }, + { + "left": "issue:43296", + "right": "issue:46821", + "accept": false, + "reason": "PaddleOCR-VL load failure is unrelated to TimesFM window_size AttributeError." + }, + { + "left": "issue:43611", + "right": "issue:43646", + "accept": false, + "reason": "Both mention Transformers 5.0.0, but model loading with base_model_prefix and custom model initialization are not the same defect." + }, + { + "left": "issue:43377", + "right": "issue:45491", + "accept": false, + "reason": "Missing padding-mask support in MIMI encoder is a different bug from Gemma3 NaN embeddings with mixed-length sliding-window batches." + } + ] + }, + "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": 26, + "review_unit_index": 13, + "review_unit_count": 61, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-505", + "nodes": [ + "issue:34634", + "issue:35707", + "issue:39692", + "issue:40444", + "issue:40990", + "issue:41553", + "issue:41720", + "issue:42175", + "issue:42796", + "issue:42915", + "issue:43262", + "issue:43296", + "issue:43454", + "issue:43482", + "issue:43493", + "issue:43550", + "issue:43582", + "issue:43606", + "issue:43644", + "issue:43701", + "issue:43746", + "issue:43818", + "issue:43819", + "issue:43828", + "issue:43873", + "issue:43994", + "issue:44079", + "issue:44186", + "issue:44265", + "issue:44291", + "issue:44466", + "issue:44479", + "issue:44509", + "issue:44829", + "issue:44877", + "issue:45072", + "issue:45325", + "issue:45357", + "issue:45381", + "issue:45479", + "issue:45584", + "issue:45701", + "issue:45823", + "issue:45834", + "issue:46402", + "issue:46821", + "issue:46909" + ], + "soft_pairs": [ + "issue:44509|issue:46909", + "issue:41553|issue:43582", + "issue:42175|issue:43296", + "issue:43644|issue:44079", + "issue:44829|issue:45479", + "issue:43818|issue:46402", + "issue:43818|issue:45357", + "issue:40444|issue:45381", + "issue:44265|issue:45072", + "issue:43493|issue:43994", + "issue:45325|issue:45834", + "issue:45325|issue:45381", + "issue:43454|issue:44466", + "issue:42915|issue:43482", + "issue:44291|issue:44877", + "issue:43873|issue:45701", + "issue:43701|issue:45325", + "issue:42796|issue:45381", + "issue:40990|issue:45584", + "issue:43606|issue:44265", + "issue:43262|issue:43746", + "issue:44479|issue:45823", + "issue:43828|issue:46821", + "issue:35707|issue:41720", + "issue:34634|issue:35707", + "issue:39692|issue:40444", + "issue:44186|issue:44265", + "issue:43454|issue:43550", + "issue:41720|issue:45823", + "issue:43606|issue:43819" + ], + "prepared_review_unit_hash": "0918d6736931199622805e0ad6f82733baad936f3ce8c6a7ef03552dd00f8112", + "component_budget": { + "node_count": 505, + "item_count": 505, + "soft_pair_count": 1854, + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 30, + "serialized_chars": 36270, + "estimated_input_tokens": 9068, + "estimated_eval_tokens": 18392 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "I did not find any soft pair that looks like the same concrete bug/change. Most pairs share only a broad subsystem or a model family name, but the failure modes are distinct enough to keep separate.", + "confidence": 0.9, + "canonical_issue_reason": "issue:43873 is the broadest still-open report and has the most follow-up activity/inbound references, so it is the best umbrella issue among this set.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43873 is a reasonable global best issue because it is active, open, and broad enough to collect related offloading/quantization reports.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44509", + "right": "issue:46909", + "accept": false, + "reason": "Both are documentation problems, but one is about removed pipeline tasks while the other is a broken DIA example; different docs bugs." + }, + { + "left": "issue:41553", + "right": "issue:43582", + "accept": false, + "reason": "Tokenizer loading error for Voxtral and an Apple Silicon allocator TypeError are unrelated failures." + }, + { + "left": "issue:42175", + "right": "issue:43296", + "accept": false, + "reason": "Package extra/backend dependency issue and a PaddleOCR-VL loading failure are different code paths." + }, + { + "left": "issue:43644", + "right": "issue:44079", + "accept": false, + "reason": "Non-persistent buffer serialization and ModelOutput key assignment are separate implementation bugs." + }, + { + "left": "issue:44829", + "right": "issue:45479", + "accept": false, + "reason": "Both mention bad training behavior, but one is specific to flash_attention_3 and the other to num_labels=1 classification loss handling." + }, + { + "left": "issue:43818", + "right": "issue:46402", + "accept": false, + "reason": "Different models and different bugs: Video-LLaVA architecture/weights vs Sam3 text encoder loading." + }, + { + "left": "issue:43818", + "right": "issue:45357", + "accept": false, + "reason": "Video-LLaVA temporal-attention/weights issue is unrelated to Qwen3.5 visual encoder key serialization." + }, + { + "left": "issue:40444", + "right": "issue:45381", + "accept": false, + "reason": "Multiple-images iterable dataset failure and video vision_position_ids regression are both Qwen2.5-VL related but not the same concrete bug." + }, + { + "left": "issue:44265", + "right": "issue:45072", + "accept": false, + "reason": "torch.export with torch_compilable_check and bfloat16 dtype mismatches are different failures." + }, + { + "left": "issue:43493", + "right": "issue:43994", + "accept": false, + "reason": "Both are about SigLIP2, but one reports implementation discrepancy and the other nonsensical inference results; not clearly the same bug." + }, + { + "left": "issue:45325", + "right": "issue:45834", + "accept": false, + "reason": "Wrong rope/position scaling in Qwen2.5-VL is unrelated to Kosmos2.5 OCR index errors." + }, + { + "left": "issue:45325", + "right": "issue:45381", + "accept": false, + "reason": "These both involve Qwen2.5-VL vision positions, but one is still-image rope scaling and the other is video vision_position_ids; too different to merge." + }, + { + "left": "issue:43454", + "right": "issue:44466", + "accept": false, + "reason": "Both mention tied lm_head weights, but one is failure to tie weights and the other is device-dependent serialization inconsistency." + }, + { + "left": "issue:42915", + "right": "issue:43482", + "accept": false, + "reason": "Qwen3Moe FP8 config failure and Qwen2.5-GGUF loading on v5 are unrelated model-loading issues." + }, + { + "left": "issue:44291", + "right": "issue:44877", + "accept": false, + "reason": "Unexpected init_empty_weights argument and strict config rejection are different load-time problems." + }, + { + "left": "issue:43873", + "right": "issue:45701", + "accept": false, + "reason": "Offloading/quantization behavior and tokenization changes are unrelated." + }, + { + "left": "issue:43701", + "right": "issue:45325", + "accept": false, + "reason": "Checkpoint key mismatch has nothing to do with Qwen2.5-VL rope indexing." + }, + { + "left": "issue:42796", + "right": "issue:45381", + "accept": false, + "reason": "Tokenizer BPE marker handling and video vision position bugs are unrelated." + }, + { + "left": "issue:40990", + "right": "issue:45584", + "accept": false, + "reason": "High perplexity on GPT-OSS and Whisper empty-transcription generation failure are different issues." + }, + { + "left": "issue:43606", + "right": "issue:44265", + "accept": false, + "reason": "CPU offload device mismatch for Bark and export failure for torch_compilable_check are unrelated." + }, + { + "left": "issue:43262", + "right": "issue:43746", + "accept": false, + "reason": "Audio chat-template sampling-rate default and GraniteSpeech PEFT checkpoint loading are separate bugs." + }, + { + "left": "issue:44479", + "right": "issue:45823", + "accept": false, + "reason": "Video-input regression across Qwen models and Gemma4 PLE device mismatch are different code paths." + }, + { + "left": "issue:43828", + "right": "issue:46821", + "accept": false, + "reason": "Autocast dtype mismatch in Phi-tiny-MoE and TimesFM window_size AttributeError are unrelated." + }, + { + "left": "issue:35707", + "right": "issue:41720", + "accept": false, + "reason": "inputs_embeds/past_key_values generation and Qwen3 auto device mapping cudaErrorAssert are different failures." + }, + { + "left": "issue:34634", + "right": "issue:35707", + "accept": false, + "reason": "Bark voice_preset and progressive generation with past_key_values do not describe the same bug." + }, + { + "left": "issue:39692", + "right": "issue:40444", + "accept": false, + "reason": "SigLIP2 doc example errors and Qwen2.5-VL multi-image finetuning failure are unrelated." + }, + { + "left": "issue:44186", + "right": "issue:44265", + "accept": false, + "reason": "LayoutLMv2 tokenization crashes and torch.export export failures are different issues." + }, + { + "left": "issue:43454", + "right": "issue:43550", + "accept": false, + "reason": "AyaVision lm_head tying/serialization and Bamba torch.compile SDPA failure are unrelated." + }, + { + "left": "issue:41720", + "right": "issue:45823", + "accept": false, + "reason": "Qwen3 auto device mapping crash and Gemma4 PLE device mismatch are separate model/runtime bugs." + }, + { + "left": "issue:43606", + "right": "issue:43819", + "accept": false, + "reason": "Bark CPU offload device mismatch and DAC.from_latents/STE mismatch 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": 26, + "review_unit_index": 14, + "review_unit_count": 61, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-505", + "nodes": [ + "issue:26413", + "issue:28530", + "issue:36331", + "issue:40990", + "issue:41720", + "issue:42617", + "issue:42796", + "issue:42915", + "issue:42947", + "issue:43065", + "issue:43257", + "issue:43317", + "issue:43322", + "issue:43377", + "issue:43450", + "issue:43454", + "issue:43572", + "issue:43720", + "issue:43749", + "issue:43873", + "issue:43881", + "issue:44186", + "issue:44387", + "issue:44466", + "issue:44488", + "issue:44534", + "issue:44568", + "issue:44610", + "issue:44756", + "issue:44877", + "issue:44898", + "issue:44977", + "issue:44991", + "issue:45072", + "issue:45357", + "issue:45701", + "issue:45704", + "issue:45759", + "issue:45823", + "issue:45920", + "issue:45923", + "issue:46050", + "issue:46209", + "issue:46305", + "issue:46315", + "issue:46402", + "issue:46821" + ], + "soft_pairs": [ + "issue:43377|issue:44610", + "issue:43257|issue:45923", + "issue:44568|issue:45920", + "issue:42796|issue:46209", + "issue:43454|issue:45072", + "issue:46305|issue:46315", + "issue:45823|issue:46821", + "issue:43881|issue:44488", + "issue:40990|issue:42947", + "issue:42617|issue:42915", + "issue:43377|issue:44898", + "issue:43454|issue:43720", + "issue:43749|issue:44534", + "issue:43317|issue:43873", + "issue:43572|issue:45759", + "issue:43322|issue:46050", + "issue:41720|issue:43322", + "issue:36331|issue:42617", + "issue:44387|issue:45704", + "issue:41720|issue:44977", + "issue:36331|issue:42915", + "issue:40990|issue:45701", + "issue:44466|issue:45357", + "issue:43450|issue:45701", + "issue:44877|issue:44991", + "issue:43065|issue:46402", + "issue:41720|issue:42617", + "issue:26413|issue:28530", + "issue:44186|issue:45072", + "issue:43873|issue:44756" + ], + "prepared_review_unit_hash": "58ace76568ecf16b07eace4af5990b1adc580b6784f0e310803cc15290e9c0fd", + "component_budget": { + "node_count": 505, + "item_count": 505, + "soft_pair_count": 1854, + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 30, + "serialized_chars": 36063, + "estimated_input_tokens": 9016, + "estimated_eval_tokens": 18288 + }, + "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 share broad subsystem wording, but each points to a different concrete bug, model, or code path.", + "confidence": 0.95, + "canonical_issue_reason": "No exact duplicate exists in this set; if one issue must serve as the cluster anchor, issue 43873 is the closest because it is active, discussion-heavy, and broadly representative of the offloading/quantization theme seen in the nearest neighbors.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 43873 is the strongest representative candidate by activity and breadth, but it should not be treated as a true duplicate of the others.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43377", + "right": "issue:44610", + "accept": false, + "reason": "Different model/components: MIMI encoder padding-mask batching bug vs OmDet-Turbo processor/model input-size mismatch." + }, + { + "left": "issue:43257", + "right": "issue:45923", + "accept": false, + "reason": "Both are Qwen-related, but one is MoE weight conversion/loading and the other is a missing gradient-checkpointing flag." + }, + { + "left": "issue:44568", + "right": "issue:45920", + "accept": false, + "reason": "Both are tokenizer regressions, but they affect different tokenizers and failure modes (special tokens vs wrong IDs)." + }, + { + "left": "issue:42796", + "right": "issue:46209", + "accept": false, + "reason": "Tokenizer BPE-marker handling is unrelated to FP8 inference failure on B200." + }, + { + "left": "issue:43454", + "right": "issue:45072", + "accept": false, + "reason": "One is a tied-weights serialization/generation bug; the other is a bfloat16 dtype mismatch in inference." + }, + { + "left": "issue:46305", + "right": "issue:46315", + "accept": false, + "reason": "Off-by-one sliding-window masking is a different bug from convert_tokens_to_ids performance regression." + }, + { + "left": "issue:45823", + "right": "issue:46821", + "accept": false, + "reason": "Device-mapping forward mismatch and a window_size AttributeError are different failure paths." + }, + { + "left": "issue:43881", + "right": "issue:44488", + "accept": false, + "reason": "Separate model-loading failures with different models and likely different causes." + }, + { + "left": "issue:40990", + "right": "issue:42947", + "accept": false, + "reason": "High perplexity on one model/dataset is not the same as gradient-checkpointing ineffectiveness with LoRA." + }, + { + "left": "issue:42617", + "right": "issue:42915", + "accept": false, + "reason": "3d_parallel execution failure and Qwen3 MoE FP8 config failure are unrelated." + }, + { + "left": "issue:43377", + "right": "issue:44898", + "accept": false, + "reason": "Both involve batched-vs-single behavior, but one is missing padding-mask support while the other is an image-size/interpolation failure." + }, + { + "left": "issue:43454", + "right": "issue:43720", + "accept": false, + "reason": "Different bugs: lm_head weight tying/serialization vs packed-weight unpacking during accelerate loading." + }, + { + "left": "issue:43749", + "right": "issue:44534", + "accept": false, + "reason": "FSDP CPU-efficient loading breakage and non-persistent buffer junk are separate regressions." + }, + { + "left": "issue:43317", + "right": "issue:43873", + "accept": false, + "reason": "Both mention offload/quantization, but one is dequantized model loading with device_map=auto and the other is a broader quantization offloading problem." + }, + { + "left": "issue:43572", + "right": "issue:45759", + "accept": false, + "reason": "Missing pad_token_idx in a config is unrelated to AutoModelForCausalLM not unwrapping text_config for composite multimodal configs." + }, + { + "left": "issue:43322", + "right": "issue:46050", + "accept": false, + "reason": "Segmentation fault while loading Llava Next is not the same as sinq quantized model save/load corruption." + }, + { + "left": "issue:41720", + "right": "issue:43322", + "accept": false, + "reason": "Qwen3 auto device-mapping cuda assert and Llava Next loading segfault are different model/load paths." + }, + { + "left": "issue:36331", + "right": "issue:42617", + "accept": false, + "reason": "Trainer compute_loss signature change is unrelated to the 3d_parallel runtime failure." + }, + { + "left": "issue:44387", + "right": "issue:45704", + "accept": false, + "reason": "Both are memory-related, but int4 CUDA reserved-memory growth is not the same as T5 using apex.FusedRMSNorm." + }, + { + "left": "issue:41720", + "right": "issue:44977", + "accept": false, + "reason": "Both are Qwen-family reports, but the failures are different: device mapping cuda assert vs flash-attention generation issues." + }, + { + "left": "issue:36331", + "right": "issue:42915", + "accept": false, + "reason": "Unexpected compute_loss argument and Qwen3 MoE FP8 conversion failure are unrelated bugs." + }, + { + "left": "issue:40990", + "right": "issue:45701", + "accept": false, + "reason": "Extremely high perplexity and tokenization-version-dependent tokenization changes are different symptoms and code paths." + }, + { + "left": "issue:44466", + "right": "issue:45357", + "accept": false, + "reason": "Both touch serialization, but one concerns tied lm_head weights while the other concerns incorrect visual encoder keys." + }, + { + "left": "issue:43450", + "right": "issue:45701", + "accept": false, + "reason": "Video processor batch shape bug is unrelated to a tokenizer version changing tokenization." + }, + { + "left": "issue:44877", + "right": "issue:44991", + "accept": false, + "reason": "Strict config rejection for granite_speech and tokenizer loading failure for est-roberta are different loaders and causes." + }, + { + "left": "issue:43065", + "right": "issue:46402", + "accept": false, + "reason": "Sam3PixelDecoder dummy Conv2d and Sam3Model text encoder weights not loading are different parts of the model." + }, + { + "left": "issue:41720", + "right": "issue:42617", + "accept": false, + "reason": "Different failures: Qwen3 device-mapping cuda assert vs 3d_parallel execution failure." + }, + { + "left": "issue:26413", + "right": "issue:28530", + "accept": false, + "reason": "Checkpoint-resume epoch_iterator error is unrelated to early-stopping metric resolution." + }, + { + "left": "issue:44186", + "right": "issue:45072", + "accept": false, + "reason": "Tokenizer crash on NER/batched padding is not the same bug as dtype mismatch in bfloat16 inference." + }, + { + "left": "issue:43873", + "right": "issue:44756", + "accept": false, + "reason": "Both involve memory/offload concerns, but one is quantization offloading behavior and the other is mmap/OOM on Strix Halo." + } + ] + }, + "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": 26, + "review_unit_index": 15, + "review_unit_count": 61, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-505", + "nodes": [ + "issue:28530", + "issue:33666", + "issue:36010", + "issue:36331", + "issue:41720", + "issue:42617", + "issue:42915", + "issue:43295", + "issue:43317", + "issue:43404", + "issue:43425", + "issue:43526", + "issue:43531", + "issue:43550", + "issue:43572", + "issue:43577", + "issue:43606", + "issue:43646", + "issue:43708", + "issue:43720", + "issue:43881", + "issue:43883", + "issue:44008", + "issue:44186", + "issue:44361", + "issue:44423", + "issue:44464", + "issue:44610", + "issue:44671", + "issue:44734", + "issue:44743", + "issue:44898", + "issue:44991", + "issue:45161", + "issue:45357", + "issue:45406", + "issue:45632", + "issue:45684", + "issue:45685", + "issue:45701", + "issue:45823", + "issue:46291", + "issue:46612", + "issue:46665", + "issue:46821", + "issue:46858", + "issue:46869" + ], + "soft_pairs": [ + "issue:43526|issue:43883", + "issue:43881|issue:44991", + "issue:46612|issue:46858", + "issue:43577|issue:43708", + "issue:43550|issue:43720", + "issue:43295|issue:45701", + "issue:45685|issue:46821", + "issue:43646|issue:46665", + "issue:44186|issue:44610", + "issue:43550|issue:43606", + "issue:36010|issue:42617", + "issue:43404|issue:45357", + "issue:44610|issue:44898", + "issue:42915|issue:43317", + "issue:44734|issue:46858", + "issue:44423|issue:45406", + "issue:46821|issue:46869", + "issue:44008|issue:45823", + "issue:41720|issue:45161", + "issue:28530|issue:33666", + "issue:44464|issue:46858", + "issue:45632|issue:45684", + "issue:43531|issue:43572", + "issue:44671|issue:44743", + "issue:42915|issue:46291", + "issue:36331|issue:41720", + "issue:44186|issue:44361", + "issue:43425|issue:46821" + ], + "prepared_review_unit_hash": "f711175a2d63a874aeab4ccde3976ae6328aaa88444510913c89f4520332eed0", + "component_budget": { + "node_count": 505, + "item_count": 505, + "soft_pair_count": 1854, + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 28, + "serialized_chars": 35677, + "estimated_input_tokens": 8920, + "estimated_eval_tokens": 18096 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is not a duplicate set; it\u2019s a grab bag of unrelated Transformers issues. All soft-similarity pairs are false positives, mostly because they share broad subsystems (tokenizers, cache, loading, serving, quantization) but describe different concrete failures and code paths.", + "confidence": 0.98, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43526", + "right": "issue:43883", + "accept": false, + "reason": "Different failures: BeitImageProcessorFast label reduction vs MolmoForCausalLM missing tied-weights metadata." + }, + { + "left": "issue:43881", + "right": "issue:44991", + "accept": false, + "reason": "Unrelated model-specific loading problems: GLM-4V load failure vs tokenizer loading for a RoBERTa checkpoint." + }, + { + "left": "issue:46612", + "right": "issue:46858", + "accept": false, + "reason": "Both involve cache behavior, but one is missing reorder logic for several generation models and the other is a static-cache seq_length tensor/arithmetic bug under compile on non-CUDA backends." + }, + { + "left": "issue:43577", + "right": "issue:43708", + "accept": false, + "reason": "Distinct areas: dtype propagation in Blip2 loading vs Trainer checkpoint resume/max_steps calculation." + }, + { + "left": "issue:43550", + "right": "issue:43720", + "accept": false, + "reason": "Different code paths: torch.compile+SDPA in Bamba vs Accelerate loading of packed BitNet weights." + }, + { + "left": "issue:43295", + "right": "issue:45701", + "accept": false, + "reason": "One is a regression in processor/tokenizer access and image passing; the other is a broader tokenization change report with different symptoms." + }, + { + "left": "issue:45685", + "right": "issue:46821", + "accept": false, + "reason": "MPS histogram/int bug in MoE routing is unrelated to TimesFM's missing window_size attribute." + }, + { + "left": "issue:43646", + "right": "issue:46665", + "accept": false, + "reason": "Custom model init breakage in Transformers 5.0 is not the same as inconsistent from_pretrained behavior inside Python classes." + }, + { + "left": "issue:44186", + "right": "issue:44610", + "accept": false, + "reason": "Both are preprocessing issues, but LayoutLMv2 NER/padding crashes are unrelated to OmDet-Turbo's input-size mismatch." + }, + { + "left": "issue:43550", + "right": "issue:43606", + "accept": false, + "reason": "Different bugs: SDPA/torch.compile failure vs CPU-offload device mismatch in bark-small." + }, + { + "left": "issue:36010", + "right": "issue:42617", + "accept": false, + "reason": "GenerationMixin import breakage is unrelated to a 3d_parallel.py runtime failure." + }, + { + "left": "issue:43404", + "right": "issue:45357", + "accept": false, + "reason": "Different multimodal bugs: untied lm_head weights in Mistral3 vs incorrect visual encoder keys in Qwen3.5 save_pretrained." + }, + { + "left": "issue:44610", + "right": "issue:44898", + "accept": false, + "reason": "Both mention image sizing, but OmDet-Turbo processor output size and Perceiver positional interpolation are different failures." + }, + { + "left": "issue:42915", + "right": "issue:43317", + "accept": false, + "reason": "Quantization/FP8 failure in Qwen3Moe is separate from dequantized model offload loading on device_map=auto." + }, + { + "left": "issue:44734", + "right": "issue:46858", + "accept": false, + "reason": "Serve KV-cache continuation indexing bug is different from the static-cache seq_length tensor issue in prefill/compile." + }, + { + "left": "issue:44423", + "right": "issue:45406", + "accept": false, + "reason": "Both affect transformers serve with multimodal models, but the concrete errors and processor failures are different (`str.to` vs missing `_tokenizer`)." + }, + { + "left": "issue:46821", + "right": "issue:46869", + "accept": false, + "reason": "TimesFM's window_size AttributeError is unrelated to YOLO multi-GPU device_map loading." + }, + { + "left": "issue:44008", + "right": "issue:45823", + "accept": false, + "reason": "Different Gemma family bugs: variable-name collision in forward vs PLE device mismatch with auto device mapping." + }, + { + "left": "issue:41720", + "right": "issue:45161", + "accept": false, + "reason": "Qwen3 auto device-map cudaErrorAssert and GPT-OSS MoE tensor-parallel failure are different model/runtime issues." + }, + { + "left": "issue:28530", + "right": "issue:33666", + "accept": false, + "reason": "Early stopping metric selection has nothing to do with Qwen2-VL multi-GPU training." + }, + { + "left": "issue:44464", + "right": "issue:46858", + "accept": false, + "reason": "Compiled forward chunked generation inconsistency is not the same bug as static-cache seq_length returning a device tensor." + }, + { + "left": "issue:45632", + "right": "issue:45684", + "accept": false, + "reason": "Cache-path collision for local trust_remote_code models is unrelated to read-only permission propagation during save_pretrained." + }, + { + "left": "issue:43531", + "right": "issue:43572", + "accept": false, + "reason": "Qwen3-MoE sliding_window behavior and StableLmConfig pad_token_idx missing are separate config/model issues." + }, + { + "left": "issue:44671", + "right": "issue:44743", + "accept": false, + "reason": "CamemBERT masked-LM prediction regression and Qwen3.5 recurrent-state reset are unrelated." + }, + { + "left": "issue:46821", + "right": "issue:46869", + "accept": false, + "reason": "TimesFM window_size attribute error does not match YOLO multi-GPU device_map auto failure." + }, + { + "left": "issue:44008", + "right": "issue:45823", + "accept": false, + "reason": "Same broader family, but different concrete failure modes and forward-path bugs." + }, + { + "left": "issue:463?", + "right": "issue:463?", + "accept": false, + "reason": "No additional soft edge provided." + }, + { + "left": "issue:43425", + "right": "issue:46821", + "accept": false, + "reason": "Torch 2.10 compatibility concerns are unrelated to TimesFM's window_size attribute issue." + } + ] + }, + "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": 26, + "review_unit_index": 16, + "review_unit_count": 61, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-505", + "nodes": [ + "issue:28530", + "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:43526", + "issue:43531", + "issue:43540", + "issue:43550", + "issue:43606", + "issue:43653", + "issue:43720", + "issue:43746", + "issue:43782", + "issue:43873", + "issue:43881", + "issue:44077", + "issue:44387", + "issue:44451", + "issue:44492", + "issue:44509", + "issue:44534", + "issue:44589", + "issue:44625", + "issue:44913", + "issue:44991", + "issue:45059", + "issue:45357", + "issue:45406", + "issue:45464", + "issue:45701", + "issue:45704", + "issue:45759", + "issue:45859", + "issue:46402", + "issue:46665", + "issue:46821", + "issue:46858" + ], + "soft_pairs": [ + "issue:43316|issue:44913", + "issue:28530|issue:30064", + "issue:42915|issue:43322", + "issue:33357|issue:35707", + "issue:43540|issue:43782", + "issue:29159|issue:36032", + "issue:43873|issue:45704", + "issue:44387|issue:45701", + "issue:30064|issue:34634", + "issue:44534|issue:46665", + "issue:43550|issue:43653", + "issue:41720|issue:43746", + "issue:43531|issue:45759", + "issue:43526|issue:43881", + "issue:43606|issue:43720", + "issue:44492|issue:44509", + "issue:42490|issue:43329", + "issue:45406|issue:45464", + "issue:43388|issue:45059", + "issue:34567|issue:35707", + "issue:43606|issue:43653", + "issue:34634|issue:41720", + "issue:43531|issue:46821", + "issue:43540|issue:46858", + "issue:45759|issue:45859", + "issue:43531|issue:44589", + "issue:44451|issue:44991", + "issue:45357|issue:46402", + "issue:44077|issue:44625" + ], + "prepared_review_unit_hash": "fb67cf9670f4c62cef4332a84b0c4830607ad4239bdd9452a24a449f1691bc90", + "component_budget": { + "node_count": 505, + "item_count": 505, + "soft_pair_count": 1854, + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 29, + "serialized_chars": 36369, + "estimated_input_tokens": 9093, + "estimated_eval_tokens": 18442 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "No soft pair looks like the same underlying bug/change. The items span unrelated subsystems (configs, tokenizers, vision, multimodal loading, offloading, docs), so this is not a real duplicate cluster.", + "confidence": 0.94, + "canonical_issue_reason": "issue:43873 is the least-specific, most hub-like report here (open, 6 inbound refs), so it is the best available anchor even though the set is not a true duplicate cluster.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43873 is the best global issue candidate only as a routing anchor: it has the strongest ongoing signal and broadest symptom framing, but it still does not represent the rest of the items well.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43316", + "right": "issue:44913", + "accept": false, + "reason": "Different config bugs: Gemma3TextConfig API mismatch vs GPTNeoX rotary_pct not persisting on reload." + }, + { + "left": "issue:28530", + "right": "issue:30064", + "accept": false, + "reason": "Early stopping metric selection is unrelated to image processor void segmentation map handling." + }, + { + "left": "issue:42915", + "right": "issue:43322", + "accept": false, + "reason": "Different failures and code paths: Qwen3MoE FP8 config vs Llava Next loading segfault." + }, + { + "left": "issue:33357", + "right": "issue:35707", + "accept": false, + "reason": "MacOS CLIP bus error is unrelated to progressive generation with inputs_embeds/past_key_values." + }, + { + "left": "issue:43540", + "right": "issue:43782", + "accept": false, + "reason": "Video input parsing error in Qwen3OmniMoe is not the same as Qwen3VL weight_only loading failure." + }, + { + "left": "issue:29159", + "right": "issue:36032", + "accept": false, + "reason": "Tokenizer slow/fast inconsistency is a different bug from T5 add_special_tokens name conflict." + }, + { + "left": "issue:43873", + "right": "issue:45704", + "accept": false, + "reason": "Quantization/offloading regression is unrelated to T5 apex.FusedRMSNorm memory leak." + }, + { + "left": "issue:44387", + "right": "issue:45701", + "accept": false, + "reason": "Int4 CUDA reserved-memory/OOM regression is not the same as tokenization behavior changes." + }, + { + "left": "issue:30064", + "right": "issue:34634", + "accept": false, + "reason": "Void segmentation-map processing and BarkProcessor voice_preset are unrelated." + }, + { + "left": "issue:44534", + "right": "issue:46665", + "accept": false, + "reason": "Non-persistent buffer initialization bug is different from from_pretrained behavior inside Python classes." + }, + { + "left": "issue:43550", + "right": "issue:43653", + "accept": false, + "reason": "torch.compile SDPA failure for Bamba is unrelated to BigBirdTokenizer special-token registration." + }, + { + "left": "issue:41720", + "right": "issue:43746", + "accept": false, + "reason": "Qwen3 auto device mapping CUDA assert is a different bug from GraniteSpeech PEFT local checkpoint loading." + }, + { + "left": "issue:43531", + "right": "issue:45759", + "accept": false, + "reason": "Sliding-window Qwen3-MoE bug is unrelated to multimodal config unwrapping in AutoModelForCausalLM.from_config." + }, + { + "left": "issue:43526", + "right": "issue:43881", + "accept": false, + "reason": "BeitImageProcessorFast label reduction bug is unrelated to glm-4v-9b loading failure." + }, + { + "left": "issue:43606", + "right": "issue:43720", + "accept": false, + "reason": "CPU offload device mismatch in Bark is unrelated to BitNet packed-weight unpacking during accelerate loading." + }, + { + "left": "issue:44492", + "right": "issue:44509", + "accept": false, + "reason": "A cache-strategy typo is not the same as docs mentioning removed pipeline tasks." + }, + { + "left": "issue:42490", + "right": "issue:43329", + "accept": false, + "reason": "Trailing-slash local-path import warning is unrelated to undefined video-branch helpers in multimodal processors." + }, + { + "left": "issue:45406", + "right": "issue:45464", + "accept": false, + "reason": "Gemma4Processor missing _tokenizer and Qwen3.5 streaming chat/completions failures are different issues." + }, + { + "left": "issue:43388", + "right": "issue:45059", + "accept": false, + "reason": "gather_for_metrics truncation of tuple labels is unrelated to SAM3 text/bbox behavior." + }, + { + "left": "issue:34567", + "right": "issue:35707", + "accept": false, + "reason": "TrainerState token counter not updating is not the same as progressive generation cache behavior." + }, + { + "left": "issue:43606", + "right": "issue:43653", + "accept": false, + "reason": "Bark CPU offload mismatch and BigBirdTokenizer decode issues do not share an underlying bug." + }, + { + "left": "issue:34634", + "right": "issue:41720", + "accept": false, + "reason": "Bark voice_preset and Qwen3 auto device mapping failures are unrelated subsystems and symptoms." + }, + { + "left": "issue:43531", + "right": "issue:46821", + "accept": false, + "reason": "Qwen3-MoE sliding-window bug is unrelated to TimesFM window_size attribute errors." + }, + { + "left": "issue:43540", + "right": "issue:46858", + "accept": false, + "reason": "Qwen3OmniMoe video ValueError is unrelated to static-cache seq-length tensor arithmetic." + }, + { + "left": "issue:45759", + "right": "issue:45859", + "accept": false, + "reason": "Config unwrapping for composite Qwen multimodal models is unrelated to CPU-offload incompatibility in rotary embedding forward." + }, + { + "left": "issue:43531", + "right": "issue:44589", + "accept": false, + "reason": "Sliding-window handling and Float8 storage lookup failures are different code paths." + }, + { + "left": "issue:44451", + "right": "issue:44991", + "accept": false, + "reason": "ScandiBERT loading failure is not the same as est-roberta tokenizer changes under v5." + }, + { + "left": "issue:45357", + "right": "issue:46402", + "accept": false, + "reason": "Qwen3.5 visual encoder key saving regression is unrelated to Sam3 text encoder weights not loading." + }, + { + "left": "issue:44077", + "right": "issue:44625", + "accept": false, + "reason": "patchtsmixer post_init policy and Qwen3.5 num_labels propagation are unrelated API problems." + }, + { + "left": "issue:44589", + "right": "issue:46821", + "accept": false, + "reason": "Float8 storage resolution failure is unrelated to TimesFM window_size AttributeError." + } + ] + }, + "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": 26, + "review_unit_index": 17, + "review_unit_count": 61, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-505", + "nodes": [ + "issue:30064", + "issue:36010", + "issue:39692", + "issue:41720", + "issue:42886", + "issue:43381", + "issue:43421", + "issue:43644", + "issue:43653", + "issue:43746", + "issue:43782", + "issue:43844", + "issue:43873", + "issue:43927", + "issue:43994", + "issue:44062", + "issue:44186", + "issue:44368", + "issue:44451", + "issue:44514", + "issue:44545", + "issue:44610", + "issue:44625", + "issue:44671", + "issue:44843", + "issue:44898", + "issue:44918", + "issue:44936", + "issue:44987", + "issue:44995", + "issue:45042", + "issue:45072", + "issue:45081", + "issue:45356", + "issue:45357", + "issue:45381", + "issue:45479", + "issue:45704", + "issue:45759", + "issue:45907", + "issue:45923", + "issue:46032", + "issue:46050", + "issue:46463", + "issue:46655", + "issue:46665", + "issue:46693", + "issue:46858" + ], + "soft_pairs": [ + "issue:43644|issue:46665", + "issue:44062|issue:44451", + "issue:44898|issue:45072", + "issue:44987|issue:45042", + "issue:44625|issue:45479", + "issue:43381|issue:43844", + "issue:39692|issue:44610", + "issue:44368|issue:45357", + "issue:41720|issue:45704", + "issue:43746|issue:46050", + "issue:46655|issue:46693", + "issue:43381|issue:45704", + "issue:43421|issue:43927", + "issue:44918|issue:44936", + "issue:45704|issue:45923", + "issue:42886|issue:44843", + "issue:45759|issue:45923", + "issue:43746|issue:43873", + "issue:44514|issue:44545", + "issue:43994|issue:44671", + "issue:46032|issue:46463", + "issue:39692|issue:43873", + "issue:43421|issue:43653", + "issue:43782|issue:45381", + "issue:30064|issue:36010", + "issue:45081|issue:45356", + "issue:43421|issue:44186", + "issue:44995|issue:46858", + "issue:45356|issue:45907" + ], + "prepared_review_unit_hash": "331e5ca6938ce89341c0dd167b71f655bfabe660adffd6f52c2b256d031bceaa", + "component_budget": { + "node_count": 505, + "item_count": 505, + "soft_pair_count": 1854, + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 29, + "serialized_chars": 36886, + "estimated_input_tokens": 9222, + "estimated_eval_tokens": 18700 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The cluster is mostly a set of unrelated Transformers issues with only one clear duplicate pair: the two Qwen2_5_VLProcessor batched padding=False crash reports. Most other soft-similarity pairs share only a broad subsystem or keyword, not the same concrete bug.", + "confidence": 0.87, + "canonical_issue_reason": "issue:44545 is the best canonical issue here: it is an exact duplicate of issue:44514, has the same concrete failure mode, and carries slightly more discussion/inbound reference signal.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:44545 is the strongest standalone issue candidate in the cluster because it is the clearest exact duplicate, is highly specific, and appears most actionable for deduplication/triage.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43644", + "right": "issue:46665", + "accept": false, + "reason": "Both mention model loading behavior, but one is about non-persistent buffers in Transformers 5.0.0 and the other is about from_pretrained behavior inside Python classes; different bugs." + }, + { + "left": "issue:44062", + "right": "issue:44451", + "accept": false, + "reason": "AddedToken constructor argument duplication and failing to load vesteinn/ScandiBERT are different failure modes with no shared concrete code path." + }, + { + "left": "issue:44898", + "right": "issue:45072", + "accept": false, + "reason": "Perceiver resolution/interpolate_pos_encoding and bf16 dtype mismatches in SwitchTransformers/TimmWrapperModel are unrelated." + }, + { + "left": "issue:44987", + "right": "issue:45042", + "accept": false, + "reason": "Loading a specific model package and an image-processor torchvision dependency regression are different issues." + }, + { + "left": "issue:44625", + "right": "issue:45479", + "accept": false, + "reason": "Both involve num_labels, but one is config propagation and the other is a degenerate loss bug in single-label classification; not the same bug." + }, + { + "left": "issue:43381", + "right": "issue:43844", + "accept": false, + "reason": "Eval-mode gradient checkpointing restriction and ZeRO-3 gradient explosion are unrelated training-path problems." + }, + { + "left": "issue:39692", + "right": "issue:44610", + "accept": false, + "reason": "Both are model/processor mismatch reports, but they affect different models and different concrete processors." + }, + { + "left": "issue:44368", + "right": "issue:45357", + "accept": false, + "reason": "A tied-embeddings warning during LoRA fine-tuning is unrelated to incorrect visual-encoder keys saved by Qwen3.5." + }, + { + "left": "issue:41720", + "right": "issue:45704", + "accept": false, + "reason": "CUDA assert during Qwen3 auto device mapping and a T5 RMSNorm memory leak are different bugs." + }, + { + "left": "issue:43746", + "right": "issue:46050", + "accept": false, + "reason": "PEFT local-checkpoint loading and sinq quantized save/load breakage are different code paths." + }, + { + "left": "issue:46655", + "right": "issue:46693", + "accept": false, + "reason": "Incorrect video padding in SmolVLM and FA2 long-context performance regression are not the same issue." + }, + { + "left": "issue:43381", + "right": "issue:45704", + "accept": false, + "reason": "Gradient checkpointing eval-mode behavior and an apex RMSNorm leak are unrelated." + }, + { + "left": "issue:43421", + "right": "issue:43927", + "accept": false, + "reason": "Runtime post-processor updates for special tokens and DiaConfig losing custom token IDs after save/load are distinct tokenizer bugs." + }, + { + "left": "issue:44918", + "right": "issue:44936", + "accept": false, + "reason": "TRL embedding-unpack failure and trainer.evaluate() failing after train() are different trainer issues." + }, + { + "left": "issue:45704", + "right": "issue:45923", + "accept": false, + "reason": "T5 apex memory leak and a missing supports_gradient_checkpointing flag are unrelated." + }, + { + "left": "issue:42886", + "right": "issue:44843", + "accept": false, + "reason": "Both are offline-mode related, but one is cache loading and the other is an unconditional model_info() call in _patch_mistral_regex; different root causes." + }, + { + "left": "issue:45759", + "right": "issue:45923", + "accept": false, + "reason": "AutoModelForCausalLM config unwrapping and a missing gradient-checkpointing flag are unrelated." + }, + { + "left": "issue:43746", + "right": "issue:43873", + "accept": false, + "reason": "Local checkpoint loading with PEFT adapters and quantization/offloading behavior are different issues." + }, + { + "left": "issue:44514", + "right": "issue:44545", + "accept": true, + "reason": "Exact same title and symptom: Qwen2_5_VLProcessor.apply_chat_template crashes on batched input when padding=False." + }, + { + "left": "issue:43994", + "right": "issue:44671", + "accept": false, + "reason": "Both concern model outputs, but one is SigLIP2 and the other is CamemBERT masked-LM behavior; not the same bug." + }, + { + "left": "issue:46032", + "right": "issue:46463", + "accept": false, + "reason": "Mamba2Mixer incorrect results with use_cache and a T5 decoder-input validation error message are different problems." + }, + { + "left": "issue:39692", + "right": "issue:43873", + "accept": false, + "reason": "SigLIP2 documentation/model mismatch and quantization offloading are unrelated." + }, + { + "left": "issue:43421", + "right": "issue:43653", + "accept": false, + "reason": "Both touch tokenizer special-token handling, but one is runtime post-processor updates and the other is a BigBirdTokenizer mask-token registration bug." + }, + { + "left": "issue:43782", + "right": "issue:45381", + "accept": false, + "reason": "A from_pretrained weight_only error for Qwen3VL and wrong vision_position_ids in Qwen2.5-VL are different multimodal bugs." + }, + { + "left": "issue:30064", + "right": "issue:36010", + "accept": false, + "reason": "Image segmentation map processing and a GenerationMixin import error have no common code path." + }, + { + "left": "issue:45081", + "right": "issue:45356", + "accept": false, + "reason": "Both involve tokenizer regex patching, but one is a crash in _patch_mistral_regex and the other is a codec/regression warning issue; not the same concrete bug." + }, + { + "left": "issue:43421", + "right": "issue:44186", + "accept": false, + "reason": "Special-token post-processor updates and LayoutLMv2 NER padding/truncation crashes are distinct tokenizer problems." + }, + { + "left": "issue:44995", + "right": "issue:46858", + "accept": false, + "reason": "Stale indexer cache on second forward pass and device-tensor seq-length arithmetic in static cache/torch.compile are different cache bugs." + }, + { + "left": "issue:45356", + "right": "issue:45907", + "accept": false, + "reason": "Tokenizer regression in Kimi-K2.5 and list_repo_files missing revision forwarding are unrelated." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet and stays conservative. The only positive duplicate callout, issue:44514 \u2194 issue:44545, is well supported by the identical title/symptom and the higher inbound-reference signal on issue:44545. The remaining soft-edge rejections do not overclaim shared bugs, and the cluster-level summary appropriately characterizes the rest as mostly unrelated." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 26, + "review_unit_index": 18, + "review_unit_count": 61, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-505", + "nodes": [ + "issue:30064", + "issue:34634", + "issue:36010", + "issue:36331", + "issue:39692", + "issue:41720", + "issue:43295", + "issue:43335", + "issue:43531", + "issue:43540", + "issue:43550", + "issue:43596", + "issue:43611", + "issue:43618", + "issue:43643", + "issue:43653", + "issue:43697", + "issue:43720", + "issue:43756", + "issue:43819", + "issue:43927", + "issue:44008", + "issue:44062", + "issue:44186", + "issue:44220", + "issue:44442", + "issue:44448", + "issue:44488", + "issue:44661", + "issue:44671", + "issue:44821", + "issue:44843", + "issue:44877", + "issue:44898", + "issue:44991", + "issue:45072", + "issue:45081", + "issue:45084", + "issue:45381", + "issue:45834", + "issue:45987", + "issue:46294", + "issue:46665", + "issue:46693", + "issue:46829", + "issue:46832", + "issue:46858", + "issue:46932" + ], + "soft_pairs": [ + "issue:43720|issue:43819", + "issue:46829|issue:46832", + "issue:43653|issue:43720", + "issue:45987|issue:46829", + "issue:43295|issue:45834", + "issue:44062|issue:44488", + "issue:43550|issue:43927", + "issue:43550|issue:45072", + "issue:43540|issue:46294", + "issue:45381|issue:46693", + "issue:43531|issue:43643", + "issue:43596|issue:44220", + "issue:43611|issue:46665", + "issue:44186|issue:44898", + "issue:43653|issue:43819", + "issue:44008|issue:46858", + "issue:41720|issue:44991", + "issue:45987|issue:46932", + "issue:30064|issue:36331", + "issue:44220|issue:44661", + "issue:43618|issue:43697", + "issue:43643|issue:44877", + "issue:43335|issue:43756", + "issue:44843|issue:45081", + "issue:43550|issue:43819", + "issue:39692|issue:45834", + "issue:44821|issue:45084", + "issue:44448|issue:44671", + "issue:34634|issue:36010", + "issue:44186|issue:44442", + "issue:43540|issue:44008" + ], + "prepared_review_unit_hash": "6e0f8afc1c14e3c99e0befd2e30080671003a75f316495f0ae8bc5c75af696fd", + "component_budget": { + "node_count": 505, + "item_count": 505, + "soft_pair_count": 1854, + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 31, + "serialized_chars": 37229, + "estimated_input_tokens": 9308, + "estimated_eval_tokens": 18872 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "All proposed soft pairs are false positives: they share broad wording like loading, tokenizer, config, or vision/video, but each points to a different concrete bug or code path. This cluster is too heterogeneous to have a single canonical duplicate.", + "confidence": 0.96, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43720", + "right": "issue:43819", + "accept": false, + "reason": "Different subsystems and failures: BitNet accelerate weight unpacking vs DAC latent reconstruction/STE behavior." + }, + { + "left": "issue:46829", + "right": "issue:46832", + "accept": false, + "reason": "Both are vision-model issues, but one is an output-field mismatch and the other is an RT-DETR feature-level shape/type error." + }, + { + "left": "issue:43653", + "right": "issue:43720", + "accept": false, + "reason": "Tokenizer special-token registration bug vs quantized linear loading/unpacking bug; unrelated code paths." + }, + { + "left": "issue:45987", + "right": "issue:46829", + "accept": false, + "reason": "Cache sequence-length return type issue vs VideoPrism output tensor assignment; not the same underlying bug." + }, + { + "left": "issue:43295", + "right": "issue:45834", + "accept": false, + "reason": "Processor/tokenizer regression in v4.57.5 vs Kosmos2.5 OCR long-input indexing; different models and failure modes." + }, + { + "left": "issue:44062", + "right": "issue:44488", + "accept": false, + "reason": "AddedToken keyword collision vs loading a specific tokenizer/model; the reported errors do not share a concrete code path." + }, + { + "left": "issue:43550", + "right": "issue:43927", + "accept": false, + "reason": "torch.compile/SDPA issue in Bamba vs DiaConfig losing custom token IDs on save/load." + }, + { + "left": "issue:43550", + "right": "issue:45072", + "accept": false, + "reason": "Both involve runtime/dtype behavior, but the concrete bugs are different models and different failure mechanisms." + }, + { + "left": "issue:43540", + "right": "issue:46294", + "accept": false, + "reason": "Both are video-related, but one is a Qwen3OmniMoe video-processing ValueError and the other is a chat-template token/feature mismatch." + }, + { + "left": "issue:45381", + "right": "issue:46693", + "accept": false, + "reason": "Qwen2.5-VL video position IDs bug vs FA2 long-context performance regression; unrelated." + }, + { + "left": "issue:43531", + "right": "issue:43643", + "accept": false, + "reason": "Sliding-window handling in Qwen3-MoE vs AutoConfig trust_remote_code field loss; different layers of the stack." + }, + { + "left": "issue:43596", + "right": "issue:44220", + "accept": false, + "reason": "DeepSpeed ZeRO-3/BertModel init failure vs audio fbank feature extraction bug." + }, + { + "left": "issue:43611", + "right": "issue:46665", + "accept": false, + "reason": "Both touch loading, but one is base_model_prefix compatibility in Transformers 5 and the other is from_pretrained behavior inside Python classes." + }, + { + "left": "issue:44186", + "right": "issue:44898", + "accept": false, + "reason": "Tokenizer crash on NER/padding vs Perceiver image classification resolution failure; different model families and code paths." + }, + { + "left": "issue:43653", + "right": "issue:43819", + "accept": false, + "reason": "BigBird tokenizer special-token registration vs DAC.from_latents mismatch; no shared underlying defect." + }, + { + "left": "issue:44008", + "right": "issue:46858", + "accept": false, + "reason": "Gemma3n forward variable collision vs cache.get_seq_length returning a device tensor; unrelated bugs." + }, + { + "left": "issue:41720", + "right": "issue:44991", + "accept": false, + "reason": "Qwen3 auto device mapping CUDA assert vs tokenizer loading failure for a specific model." + }, + { + "left": "issue:45987", + "right": "issue:46932", + "accept": false, + "reason": "StaticCache return-type contract bug vs Gemma4 quantized embedder dtype casting; different components." + }, + { + "left": "issue:30064", + "right": "issue:36331", + "accept": false, + "reason": "ImageProcessor segmentation-map processing vs CustomTrainer.compute_loss signature change; no overlap in code path." + }, + { + "left": "issue:44220", + "right": "issue:44661", + "accept": false, + "reason": "Audio feature extraction helper vs auto-model registration failure in TOKENIZER_MAPPING_NAMES." + }, + { + "left": "issue:43618", + "right": "issue:43697", + "accept": false, + "reason": "CLIPOutput attentions field regression vs RTDetrV2 output differences; both are v5 regressions but not the same bug." + }, + { + "left": "issue:43643", + "right": "issue:44877", + "accept": false, + "reason": "AutoConfig trust_remote_code missing fields vs strict config blocking granite_speech loading; different symptoms and fixes." + }, + { + "left": "issue:43335", + "right": "issue:43756", + "accept": false, + "reason": "SwitchTransformers sparse-layer construction bug vs Smollm3 RoPE-layer dropping bug." + }, + { + "left": "issue:44843", + "right": "issue:45081", + "accept": false, + "reason": "Both involve _patch_mistral_regex, but one is an offline model_info() network call and the other is a backend_tokenizer AttributeError." + }, + { + "left": "issue:43550", + "right": "issue:43819", + "accept": false, + "reason": "Same as above pair: different model/runtime bug vs DAC latent reconstruction bug." + }, + { + "left": "issue:39692", + "right": "issue:45834", + "accept": false, + "reason": "Documentation/example errors for SigLIP2 vs Kosmos2.5 OCR long-input indexing; not the same issue." + }, + { + "left": "issue:44821", + "right": "issue:45084", + "accept": false, + "reason": "AutoImageProcessor URL loading vs a template compilation TypeError; unrelated functionality." + }, + { + "left": "issue:44448", + "right": "issue:44671", + "accept": false, + "reason": "Different model regressions in v5: Pegasus output change vs CamemBERT masked-LM prediction change." + }, + { + "left": "issue:34634", + "right": "issue:36010", + "accept": false, + "reason": "BarkProcessor voice_preset bug vs GenerationMixin import failure; unrelated." + }, + { + "left": "issue:44186", + "right": "issue:44442", + "accept": false, + "reason": "LayoutLMv2 tokenizer padding/NER crash vs AutoTokenizer failing to load a FastSpeech2Conformer tokenizer." + }, + { + "left": "issue:43540", + "right": "issue:44008", + "accept": false, + "reason": "Qwen3OmniMoe video-input handling vs Gemma3n variable-name collision in forward(); different models and bugs." + }, + { + "left": "issue:45381", + "right": "issue:46693", + "accept": false, + "reason": "Qwen video-position-id bug vs FA2 long-context performance regression; no shared code-path." + } + ] + }, + "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": 26, + "review_unit_index": 19, + "review_unit_count": 61, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-505", + "nodes": [ + "issue:28218", + "issue:29159", + "issue:43322", + "issue:43388", + "issue:43421", + "issue:43540", + "issue:43577", + "issue:43582", + "issue:43606", + "issue:43644", + "issue:43645", + "issue:43716", + "issue:43720", + "issue:43749", + "issue:43782", + "issue:43818", + "issue:43825", + "issue:43827", + "issue:43931", + "issue:43994", + "issue:44008", + "issue:44291", + "issue:44479", + "issue:44561", + "issue:44610", + "issue:44671", + "issue:44805", + "issue:44811", + "issue:44936", + "issue:44977", + "issue:44991", + "issue:45020", + "issue:45072", + "issue:45183", + "issue:45381", + "issue:45479", + "issue:45584", + "issue:45685", + "issue:45701", + "issue:45834", + "issue:46032", + "issue:46291", + "issue:46294", + "issue:46463", + "issue:46512", + "issue:46665", + "issue:46846" + ], + "soft_pairs": [ + "issue:45584|issue:45834", + "issue:45183|issue:45479", + "issue:43322|issue:45685", + "issue:43644|issue:43749", + "issue:43606|issue:45072", + "issue:43782|issue:44977", + "issue:43421|issue:44811", + "issue:43582|issue:46512", + "issue:44479|issue:45834", + "issue:43577|issue:43818", + "issue:43388|issue:46032", + "issue:44561|issue:45020", + "issue:43825|issue:43994", + "issue:44991|issue:46291", + "issue:46294|issue:46463", + "issue:43716|issue:44008", + "issue:43931|issue:46846", + "issue:44805|issue:44936", + "issue:43540|issue:45381", + "issue:44610|issue:45072", + "issue:43749|issue:44291", + "issue:28218|issue:29159", + "issue:43720|issue:45072", + "issue:43645|issue:46665", + "issue:43825|issue:43827", + "issue:44671|issue:45701", + "issue:43577|issue:43582" + ], + "prepared_review_unit_hash": "f8a68d1e059e85628e4cd4b1f63b0417852e15d58f5048bce1d1e60c3b281cfe", + "component_budget": { + "node_count": 505, + "item_count": 505, + "soft_pair_count": 1854, + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 27, + "serialized_chars": 35589, + "estimated_input_tokens": 8898, + "estimated_eval_tokens": 18052 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly a grab bag of unrelated tokenizer, loading, video, and dtype issues. The only likely duplicate pair is the tokenizer added-token spacing report and the slow/fast tokenizer inconsistency report.", + "confidence": 0.86, + "canonical_issue_reason": "issue:28218 is the most specific statement of the tokenizer spacing bug and likely captures the underlying symptom behind the broader tokenizer inconsistency report.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:28218 is the best representative for the only clear duplicate subcluster in this set; the remaining issues are not close enough to share a single root cause.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45584", + "right": "issue:45834", + "accept": false, + "reason": "Both are input/decoding failures, but they involve different models and different code paths (Whisper empty transcription vs Kosmos2.5 OCR length/indexing)." + }, + { + "left": "issue:45183", + "right": "issue:45479", + "accept": false, + "reason": "One is an input validation error-message problem; the other is a degenerate loss bug in single-label classification. Not the same bug." + }, + { + "left": "issue:43322", + "right": "issue:45685", + "accept": false, + "reason": "Segfault loading Llava Next and an MPS histogram limitation are unrelated failures." + }, + { + "left": "issue:43644", + "right": "issue:43749", + "accept": false, + "reason": "Both concern loading, but one is junk-filled non-persistent buffers and the other is broken FSDP CPU RAM efficient loading; different root causes." + }, + { + "left": "issue:43606", + "right": "issue:45072", + "accept": false, + "reason": "CPU offload device mismatch in bark-small is unrelated to bfloat16 dtype mismatches in other models." + }, + { + "left": "issue:43782", + "right": "issue:44977", + "accept": false, + "reason": "Qwen3VL weight_only loading error and Qwen3.5 flash-attention generation failure are different concrete failures." + }, + { + "left": "issue:43421", + "right": "issue:44811", + "accept": false, + "reason": "Both touch tokenizers, but runtime special-token post-processor updates and Whisper batch_decode skip_special_tokens are different behaviors." + }, + { + "left": "issue:43582", + "right": "issue:46512", + "accept": false, + "reason": "Apple Silicon warmup TypeError and stale config.dtype after from_config are different dtype/config issues." + }, + { + "left": "issue:44479", + "right": "issue:45834", + "accept": false, + "reason": "Different regressions: multi-model video-input processing versus long OCR indexing on Kosmos2.5." + }, + { + "left": "issue:43577", + "right": "issue:43818", + "accept": false, + "reason": "Blip2 dtype retention and Video-LLaVA missing temporal attention are unrelated model-loading/model-architecture bugs." + }, + { + "left": "issue:43388", + "right": "issue:46032", + "accept": false, + "reason": "One is metrics gathering dropping tuple labels; the other is Mamba2 use_cache correctness. Not the same issue." + }, + { + "left": "issue:44561", + "right": "issue:45020", + "accept": false, + "reason": "Both mention v5/remote_code breakages, but one is specifically is_torch_fx_available removal and the other is a broad umbrella report." + }, + { + "left": "issue:43825", + "right": "issue:43994", + "accept": false, + "reason": "An error-message wording issue and a model output correctness issue are not the same bug." + }, + { + "left": "issue:44991", + "right": "issue:46291", + "accept": false, + "reason": "Tokenizer loading failure for one model and kernels dependency/version import failure are unrelated." + }, + { + "left": "issue:46294", + "right": "issue:46463", + "accept": false, + "reason": "Qwen3.5 video chat template mismatch and T5 decoder-input validation messaging are different models and code paths." + }, + { + "left": "issue:43716", + "right": "issue:44008", + "accept": false, + "reason": "dtype mismatch in Mistral-3 preprocessing/model loading is unrelated to Gemma3n's variable-name collision." + }, + { + "left": "issue:43931", + "right": "issue:46846", + "accept": false, + "reason": "Qwen3-VL weight-shape mismatch on load is not the same as missing tensor-parallel plan entries causing OOM/shape errors." + }, + { + "left": "issue:44805", + "right": "issue:44936", + "accept": false, + "reason": "A tensor-mask shape IndexError and trainer.evaluate() failing after train() are different failure modes." + }, + { + "left": "issue:43540", + "right": "issue:45381", + "accept": false, + "reason": "Both concern video inputs, but Qwen3OmniMoe processing ValueError and qwen2.5-vl vision_position_ids regression are distinct." + }, + { + "left": "issue:44610", + "right": "issue:45072", + "accept": false, + "reason": "Processor image-size mismatch and dtype mismatch are different bugs." + }, + { + "left": "issue:43749", + "right": "issue:44291", + "accept": false, + "reason": "Broken FSDP CPU RAM efficient loading and an init_empty_weights unexpected-argument TypeError are different loading-path issues." + }, + { + "left": "issue:28218", + "right": "issue:29159", + "accept": true, + "reason": "Both describe the same tokenizer symptom: added-token handling differs between tokenizer implementations and introduces an extra/incorrect space." + }, + { + "left": "issue:43720", + "right": "issue:45072", + "accept": false, + "reason": "BitNet packed-weight unpacking during accelerate loading is unrelated to bfloat16 dtype mismatches." + }, + { + "left": "issue:43645", + "right": "issue:46665", + "accept": false, + "reason": "Custom model initialization in notebooks and from_pretrained behavior inside Python classes are different workflow bugs." + }, + { + "left": "issue:43825", + "right": "issue:43827", + "accept": false, + "reason": "One is an incorrect pipeline error message; the other is stale documentation. Same theme, not the same underlying bug." + }, + { + "left": "issue:44671", + "right": "issue:45701", + "accept": false, + "reason": "CamemBERT masked-LM prediction regression and a version-dependent tokenization change are not clearly the same code-path defect." + }, + { + "left": "issue:43577", + "right": "issue:43582", + "accept": false, + "reason": "A model dtype problem and an Apple Silicon warmup TypeError are unrelated." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet and stays conservative. The only clearly supported duplicate pair is issue:28218 / issue:29159, and the other soft-edge rejections do not overstate sameness of bug. The canonical/best issue choice is reasonable given the evidence." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 26, + "review_unit_index": 20, + "review_unit_count": 61, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-505", + "nodes": [ + "issue:34634", + "issue:36010", + "issue:36331", + "issue:42175", + "issue:43257", + "issue:43262", + "issue:43329", + "issue:43452", + "issue:43526", + "issue:43531", + "issue:43550", + "issue:43582", + "issue:43638", + "issue:43645", + "issue:43653", + "issue:43697", + "issue:43742", + "issue:43782", + "issue:43819", + "issue:43906", + "issue:43927", + "issue:44060", + "issue:44368", + "issue:44554", + "issue:44589", + "issue:44857", + "issue:44912", + "issue:44936", + "issue:44991", + "issue:45003", + "issue:45020", + "issue:45042", + "issue:45081", + "issue:45084", + "issue:45325", + "issue:45372", + "issue:45406", + "issue:45538", + "issue:45600", + "issue:46175", + "issue:46209", + "issue:46285", + "issue:46291", + "issue:46326", + "issue:46697", + "issue:46728", + "issue:46932" + ], + "soft_pairs": [ + "issue:34634|issue:36331", + "issue:42175|issue:45081", + "issue:46175|issue:46932", + "issue:46697|issue:46728", + "issue:45003|issue:46285", + "issue:43782|issue:45325", + "issue:43653|issue:44912", + "issue:43262|issue:43819", + "issue:43582|issue:44857", + "issue:45372|issue:46326", + "issue:45081|issue:45406", + "issue:43742|issue:45020", + "issue:44589|issue:45042", + "issue:43550|issue:44912", + "issue:43452|issue:44991", + "issue:43257|issue:43329", + "issue:44060|issue:44368", + "issue:36010|issue:43645", + "issue:43526|issue:43697", + "issue:43906|issue:46209", + "issue:36010|issue:36331", + "issue:43329|issue:45600", + "issue:43526|issue:44554", + "issue:45084|issue:45538", + "issue:43819|issue:43927", + "issue:43531|issue:45042", + "issue:43638|issue:44936", + "issue:42175|issue:46291" + ], + "prepared_review_unit_hash": "a50d7c4635a2c41f1989a1d3a1e08e448151015bc7f8c255a01df1e42644822c", + "component_budget": { + "node_count": 505, + "item_count": 505, + "soft_pair_count": 1854, + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 28, + "serialized_chars": 35990, + "estimated_input_tokens": 8998, + "estimated_eval_tokens": 18252 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "These items are mostly unrelated model-loading, processor, trainer, and backend regressions. No soft pair looks like the same concrete bug/change, so none should be merged as duplicates. If a single issue must anchor the set, the broadest umbrella is the remote_code loading regression.", + "confidence": 0.79, + "canonical_issue_reason": "issue:45020 is the broadest, most umbrella-like regression in the list and can plausibly subsume several model-loading breakages better than the highly specific issues.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45020 is the best representative because it describes a wide loader regression rather than a model-specific symptom, making it the most plausible cluster anchor.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:34634", + "right": "issue:36331", + "accept": false, + "reason": "Different subsystems and failure modes: Bark voice preset handling vs CustomTrainer loss signature mismatch." + }, + { + "left": "issue:42175", + "right": "issue:45081", + "accept": false, + "reason": "Unrelated packaging/backend dependency issue vs Mistral tokenizer regex crash." + }, + { + "left": "issue:46175", + "right": "issue:46932", + "accept": false, + "reason": "Both are Gemma4-related, but one is a large vision patch embedding memory issue and the other is dtype casting for quantized embedders; not the same bug." + }, + { + "left": "issue:46697", + "right": "issue:46728", + "accept": false, + "reason": "Different image processors and different padding/counting logic; no shared concrete code path." + }, + { + "left": "issue:45003", + "right": "issue:46285", + "accept": false, + "reason": "sys.modules access bug in modeling_utils is not the same as CLIP text-model structure divergence." + }, + { + "left": "issue:43782", + "right": "issue:45325", + "accept": false, + "reason": "Different Qwen VL generations and different failure mechanisms (weight_only loading vs rope index scaling)." + }, + { + "left": "issue:43653", + "right": "issue:44912", + "accept": false, + "reason": "Tokenizer special-token registration bug vs MXFP4 quantization fallback; unrelated." + }, + { + "left": "issue:43262", + "right": "issue:43819", + "accept": false, + "reason": "Audio chat-template sampling-rate default bug vs DAC latent/STE mismatch; different code paths." + }, + { + "left": "issue:43582", + "right": "issue:44857", + "accept": false, + "reason": "Apple Silicon allocator TypeError is unrelated to CUDA AMP loss crash." + }, + { + "left": "issue:45372", + "right": "issue:46326", + "accept": false, + "reason": "Gemma4 processor import issue from mistral_common is different from continue_final_message behavior in mistral_common tokenizers." + }, + { + "left": "issue:45081", + "right": "issue:45406", + "accept": false, + "reason": "Mistral tokenizer regex patch crash and Gemma4Processor missing _tokenizer are separate processor/tokenizer bugs." + }, + { + "left": "issue:43742", + "right": "issue:45020", + "accept": false, + "reason": "A specific MobileLLM key error is not the same as the broader remote_code regression, even if both are loading-related." + }, + { + "left": "issue:44589", + "right": "issue:45042", + "accept": false, + "reason": "Float8 storage lookup error and PIL/torchvision backend requirement are unrelated." + }, + { + "left": "issue:43550", + "right": "issue:44912", + "accept": false, + "reason": "torch.compile+SDPA failure in Bamba is unrelated to MXFP4 quantization fallback." + }, + { + "left": "issue:43452", + "right": "issue:44991", + "accept": false, + "reason": "Different tokenizer-loading breakages with different triggers; not the same concrete bug." + }, + { + "left": "issue:43257", + "right": "issue:43329", + "accept": false, + "reason": "Qwen3 MoE weight conversion under accelerate/deepspeed is unrelated to an undefined video-token helper path." + }, + { + "left": "issue:44060", + "right": "issue:44368", + "accept": false, + "reason": "Both mention tied-weights warnings, but one is a model wiring bug and the other is a config warning during LoRA fine-tuning; not the same underlying issue." + }, + { + "left": "issue:36010", + "right": "issue:43645", + "accept": false, + "reason": "GenerationMixin import-path breakage is not the same as Jupyter custom-model initialization failures." + }, + { + "left": "issue:43526", + "right": "issue:43697", + "accept": false, + "reason": "BEiT label reduction bug and RTDetrV2 output-difference regression are different model/processor issues." + }, + { + "left": "issue:43906", + "right": "issue:46209", + "accept": false, + "reason": "A reproduction of an old issue is not the same as FP8 inference failure on B200." + }, + { + "left": "issue:36010", + "right": "issue:36331", + "accept": false, + "reason": "Import path breakage for GenerationMixin and Trainer compute_loss signature mismatch are unrelated." + }, + { + "left": "issue:43329", + "right": "issue:45600", + "accept": false, + "reason": "Undefined multimodal-token helpers and stale auto_mappings references are different maintenance bugs." + }, + { + "left": "issue:43526", + "right": "issue:44554", + "accept": false, + "reason": "Image label reduction and MPS attention correctness are unrelated." + }, + { + "left": "issue:45084", + "right": "issue:45538", + "accept": false, + "reason": "Template compilation TypeError and CLIPTokenizer max_length constant are different problems." + }, + { + "left": "issue:43819", + "right": "issue:43927", + "accept": false, + "reason": "DAC latent reconstruction mismatch and DiaConfig token-ID persistence are unrelated." + }, + { + "left": "issue:43531", + "right": "issue:45042", + "accept": false, + "reason": "Qwen3-MoE sliding-window behavior is unrelated to PIL backend dependency requirements." + }, + { + "left": "issue:43638", + "right": "issue:44936", + "accept": false, + "reason": "Zero3/non-pretrained BERT index error and evaluate-after-train failure are not the same code-path bug." + }, + { + "left": "issue:42175", + "right": "issue:46291", + "accept": false, + "reason": "Packaging extras missing TensorFlow is unrelated to kernels version/revision import failures." + } + ] + }, + "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": 26, + "review_unit_index": 21, + "review_unit_count": 61, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-505", + "nodes": [ + "issue:28282", + "issue:29159", + "issue:29870", + "issue:34567", + "issue:38175", + "issue:39692", + "issue:40990", + "issue:42490", + "issue:42915", + "issue:43257", + "issue:43262", + "issue:43329", + "issue:43450", + "issue:43454", + "issue:43540", + "issue:43577", + "issue:43606", + "issue:43643", + "issue:43650", + "issue:43696", + "issue:43720", + "issue:43742", + "issue:43746", + "issue:43749", + "issue:43854", + "issue:43906", + "issue:43927", + "issue:43931", + "issue:44291", + "issue:44387", + "issue:44479", + "issue:44568", + "issue:44610", + "issue:44617", + "issue:44756", + "issue:44811", + "issue:44821", + "issue:44912", + "issue:44977", + "issue:45042", + "issue:45356", + "issue:45459", + "issue:45759", + "issue:45803", + "issue:45859", + "issue:46421", + "issue:46424", + "issue:46701" + ], + "soft_pairs": [ + "issue:43746|issue:44811", + "issue:42490|issue:43257", + "issue:43854|issue:44977", + "issue:43643|issue:45759", + "issue:39692|issue:44912", + "issue:43720|issue:44610", + "issue:38175|issue:44610", + "issue:43257|issue:43696", + "issue:43262|issue:44811", + "issue:44387|issue:44617", + "issue:43257|issue:43931", + "issue:43329|issue:43450", + "issue:43329|issue:44479", + "issue:42915|issue:43742", + "issue:34567|issue:40990", + "issue:45459|issue:45803", + "issue:46421|issue:46424", + "issue:44291|issue:45042", + "issue:43749|issue:44756", + "issue:29159|issue:34567", + "issue:43650|issue:43906", + "issue:42915|issue:44821", + "issue:43606|issue:45859", + "issue:43540|issue:46701", + "issue:43454|issue:43577", + "issue:28282|issue:29870", + "issue:43927|issue:44912", + "issue:44568|issue:45356" + ], + "prepared_review_unit_hash": "1d6bcb0c560a453e33607d27cecb4c73a9b335f3fd7ee7da87f3d439308ff89c", + "component_budget": { + "node_count": 505, + "item_count": 505, + "soft_pair_count": 1854, + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 28, + "serialized_chars": 36313, + "estimated_input_tokens": 9079, + "estimated_eval_tokens": 18414 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The items are broadly related to Transformers/model-loading topics, but the soft pairs do not describe the same underlying bug or change. They span distinct failures in tokenizers, multimodal processors, quantization, offload, cache behavior, and training-state tracking.", + "confidence": 0.96, + "canonical_issue_reason": "No clear canonical issue: the set is heterogeneous, and the candidate pairs point to different root causes rather than one duplicate defect.", + "canonical_pr_reason": null, + "best_issue_reason": "No single issue is a strong representative because the cluster does not collapse into one bug; any pick would be arbitrary.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43746", + "right": "issue:44811", + "accept": false, + "reason": "Local PEFT checkpoint loading vs Whisper batch_decode special-token handling; different bugs." + }, + { + "left": "issue:42490", + "right": "issue:43257", + "accept": false, + "reason": "Trailing-slash import warning vs Qwen3 MoE weight-conversion failure; unrelated." + }, + { + "left": "issue:43854", + "right": "issue:44977", + "accept": false, + "reason": "GLM-4.7-Flash load/test issue vs Qwen3.5 flash-attention generation regression; different code paths." + }, + { + "left": "issue:43643", + "right": "issue:45759", + "accept": false, + "reason": "Missing fields with trust_remote_code vs text_config unwrapping for composite multimodal configs; not the same defect." + }, + { + "left": "issue:39692", + "right": "issue:44912", + "accept": false, + "reason": "Documentation/example errors vs MXFP4 quantization fallback bug; different symptoms and causes." + }, + { + "left": "issue:43720", + "right": "issue:44610", + "accept": false, + "reason": "Packed-weight accelerate loading bug vs processor image-size mismatch; unrelated." + }, + { + "left": "issue:38175", + "right": "issue:44610", + "accept": false, + "reason": "Unexpected zero probabilities vs image preprocessing size mismatch; different issues." + }, + { + "left": "issue:43257", + "right": "issue:43696", + "accept": false, + "reason": "MoE weight conversion/loading bug vs CUDA OOM on GPT-oss; not a duplicate." + }, + { + "left": "issue:43262", + "right": "issue:44811", + "accept": false, + "reason": "Audio chat-template sample-rate default vs Whisper decode special-token behavior; different components." + }, + { + "left": "issue:44387", + "right": "issue:44617", + "accept": false, + "reason": "General int4 memory regression vs Sam3Video OOM; different models and root causes." + }, + { + "left": "issue:43257", + "right": "issue:43931", + "accept": false, + "reason": "Qwen3 MoE conversion issue vs Qwen3-VL weight-shape mismatch; unrelated load failures." + }, + { + "left": "issue:43329", + "right": "issue:43450", + "accept": false, + "reason": "Undefined multimodal-token helper vars vs batched video shape mismatch; different bugs." + }, + { + "left": "issue:43329", + "right": "issue:44479", + "accept": false, + "reason": "Specific helper-function undefined-name bug vs broader video-input regression; too different to merge." + }, + { + "left": "issue:42915", + "right": "issue:43742", + "accept": false, + "reason": "FineGrainedFP8Config failure in Qwen3 MoE vs key error loading MobileLLM-125M; unrelated." + }, + { + "left": "issue:34567", + "right": "issue:40990", + "accept": false, + "reason": "TrainerState token counter bug vs high perplexity on GPT-oss; no shared root cause." + }, + { + "left": "issue:45459", + "right": "issue:45803", + "accept": false, + "reason": "Protobuf import error masking tokenizer failures vs bare except swallowing KeyboardInterrupt; both exception-related but distinct." + }, + { + "left": "issue:46421", + "right": "issue:46424", + "accept": false, + "reason": "Lazy cache init causing recompilation vs max_cache_length only honored for quantized cache; different cache bugs." + }, + { + "left": "issue:44291", + "right": "issue:45042", + "accept": false, + "reason": "init_empty_weights argument mismatch vs PIL backend requiring torchvision; unrelated." + }, + { + "left": "issue:43749", + "right": "issue:44756", + "accept": false, + "reason": "FSDP CPU RAM efficient loading bug vs mmap-on-Strix-Halo OOM; different mechanisms." + }, + { + "left": "issue:29159", + "right": "issue:34567", + "accept": false, + "reason": "Slow/fast tokenizer inconsistency vs TrainerState input-token counting; unrelated." + }, + { + "left": "issue:43650", + "right": "issue:43906", + "accept": false, + "reason": "Placeholder content issue vs isolated reproduction of another issue; not the same bug." + }, + { + "left": "issue:42915", + "right": "issue:44821", + "accept": false, + "reason": "Qwen3 MoE FP8 loading failure vs AutoImageProcessor URL loading; different areas." + }, + { + "left": "issue:43606", + "right": "issue:45859", + "accept": false, + "reason": "Both mention CPU offload, but one is a device mismatch and the other a rotary-embedding incompatibility; not the same change." + }, + { + "left": "issue:43540", + "right": "issue:46701", + "accept": false, + "reason": "Qwen3OmniMoe video-input ValueError vs DiffusionGemma image-text-to-text pipeline failure; different pipelines." + }, + { + "left": "issue:43454", + "right": "issue:43577", + "accept": false, + "reason": "AyaVision weight-tying bug vs BLIP2 dtype propagation bug; unrelated." + }, + { + "left": "issue:28282", + "right": "issue:29870", + "accept": false, + "reason": "Missing PyTorch ImportError vs Bart generate UnboundLocalError; different failure layers." + }, + { + "left": "issue:43927", + "right": "issue:44912", + "accept": false, + "reason": "DiaConfig save/load token-ID loss vs MXFP4 quantization fallback; different bugs." + }, + { + "left": "issue:44568", + "right": "issue:45356", + "accept": false, + "reason": "mDeBERTa special-token regression vs Kimi-K2.5 tokenizer codec/regex regression; unrelated tokenizer issues." + } + ] + }, + "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": 26, + "review_unit_index": 22, + "review_unit_count": 61, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-505", + "nodes": [ + "issue:24643", + "issue:27957", + "issue:28282", + "issue:29127", + "issue:29159", + "issue:33357", + "issue:34567", + "issue:34689", + "issue:36331", + "issue:41720", + "issue:42175", + "issue:42915", + "issue:43404", + "issue:43425", + "issue:43452", + "issue:43525", + "issue:43540", + "issue:43577", + "issue:43582", + "issue:43697", + "issue:43742", + "issue:43749", + "issue:43824", + "issue:44062", + "issue:44291", + "issue:44479", + "issue:44488", + "issue:44560", + "issue:44849", + "issue:44857", + "issue:44936", + "issue:44987", + "issue:44991", + "issue:45127", + "issue:45137", + "issue:45357", + "issue:45381", + "issue:45405", + "issue:45538", + "issue:45701", + "issue:45907", + "issue:46291", + "issue:46512", + "issue:46693", + "issue:46821", + "issue:46832", + "issue:46869" + ], + "soft_pairs": [ + "issue:43525|issue:46291", + "issue:44987|issue:45538", + "issue:28282|issue:34689", + "issue:33357|issue:34567", + "issue:43425|issue:46291", + "issue:44857|issue:46832", + "issue:43697|issue:45357", + "issue:43404|issue:45127", + "issue:45137|issue:45405", + "issue:42175|issue:43824", + "issue:45538|issue:46869", + "issue:42915|issue:45538", + "issue:28282|issue:29127", + "issue:44479|issue:46693", + "issue:44062|issue:44991", + "issue:43742|issue:44291", + "issue:44936|issue:46821", + "issue:43452|issue:43582", + "issue:42915|issue:43749", + "issue:24643|issue:29159", + "issue:43577|issue:46512", + "issue:44849|issue:45701", + "issue:27957|issue:28282", + "issue:44987|issue:45907", + "issue:41720|issue:43540", + "issue:43540|issue:44560", + "issue:44479|issue:45381", + "issue:43540|issue:44479", + "issue:42175|issue:44488", + "issue:34567|issue:36331" + ], + "prepared_review_unit_hash": "67b34d0821c2a553656c901f2677f76d7f7e84d07d8cb2e3ca0e589049b2ddfa", + "component_budget": { + "node_count": 505, + "item_count": 505, + "soft_pair_count": 1854, + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 30, + "serialized_chars": 35978, + "estimated_input_tokens": 8995, + "estimated_eval_tokens": 18246 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The cluster is mostly heterogeneous; most soft pairs are false positives. Only two pairs look like true duplicates: the AddedToken tokenizer-load failure and the Qwen2.5-VL video position-id regression.", + "confidence": 0.62, + "canonical_issue_reason": "No single duplicate theme dominates the set; issue 43749 is the best anchor because it is open, heavily discussed, and has inbound references, making it the strongest triage target.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 43749 is the strongest overall issue to anchor the cluster: concrete regression, active discussion, and external references.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43525", + "right": "issue:46291", + "accept": false, + "reason": "Different subsystems: missing config attribute for Llama4 vs kernels revision/version requirement in pipelines import." + }, + { + "left": "issue:44987", + "right": "issue:45538", + "accept": false, + "reason": "Different bugs: model loading regression vs CLIPTokenizer max_length constant." + }, + { + "left": "issue:28282", + "right": "issue:34689", + "accept": false, + "reason": "Both are loading failures, but one is missing PyTorch and the other is a specific Llama 3.2 Vision regression." + }, + { + "left": "issue:33357", + "right": "issue:34567", + "accept": false, + "reason": "Unrelated: Mac bus error on CLIP model vs TrainerState token counter update issue." + }, + { + "left": "issue:43425", + "right": "issue:46291", + "accept": false, + "reason": "Torch compatibility problem vs kernels revision-forwarding issue; not the same code path." + }, + { + "left": "issue:44857", + "right": "issue:46832", + "accept": false, + "reason": "Both touch detection models, but the crashes are different and not the same underlying bug." + }, + { + "left": "issue:43697", + "right": "issue:45357", + "accept": false, + "reason": "Different regressions: RT-DETR output mismatch vs save_pretrained key serialization." + }, + { + "left": "issue:43404", + "right": "issue:45127", + "accept": false, + "reason": "Both involve weight tying, but one is a model implementation bug and the other is LoRA merge behavior with extended vocab." + }, + { + "left": "issue:45137", + "right": "issue:45405", + "accept": false, + "reason": "DeepSpeed ZeRO3 deque error is unrelated to the PEFT version bump." + }, + { + "left": "issue:42175", + "right": "issue:43824", + "accept": false, + "reason": "Packaging/backend dependency issue vs a missing model class import." + }, + { + "left": "issue:45538", + "right": "issue:46869", + "accept": false, + "reason": "Tokenizer max-length constant is unrelated to YOLO multi-GPU device-map failure." + }, + { + "left": "issue:42915", + "right": "issue:45538", + "accept": false, + "reason": "Qwen3Moe FP8 config failure vs CLIPTokenizer max_length bug; different components." + }, + { + "left": "issue:28282", + "right": "issue:29127", + "accept": false, + "reason": "Model import error vs LayoutLMv3 validation/error-message clarity issue." + }, + { + "left": "issue:44479", + "right": "issue:46693", + "accept": false, + "reason": "Different regressions: multimodal video input bug vs FA2 long-context performance regression." + }, + { + "left": "issue:44062", + "right": "issue:44991", + "accept": true, + "reason": "Likely the same tokenizer deserialization bug: AddedToken special-argument handling breaks loading for a specific tokenizer/model." + }, + { + "left": "issue:43742", + "right": "issue:44291", + "accept": false, + "reason": "Both are load-time errors, but one is a key error for a model and the other is an init_empty_weights argument mismatch." + }, + { + "left": "issue:44936", + "right": "issue:46821", + "accept": false, + "reason": "Trainer train/evaluate state issue vs TimesFM window_size AttributeError; unrelated." + }, + { + "left": "issue:43452", + "right": "issue:43582", + "accept": false, + "reason": "gguf_file loading breakage is unrelated to Apple Silicon caching_allocator_warmup TypeError." + }, + { + "left": "issue:42915", + "right": "issue:43749", + "accept": false, + "reason": "Different domains: Qwen3Moe FP8 config vs FSDP CPU RAM efficient loading." + }, + { + "left": "issue:24643", + "right": "issue:29159", + "accept": false, + "reason": "DeepSpeed weight-dimension error vs tokenizer slow/fast inconsistency; unrelated." + }, + { + "left": "issue:43577", + "right": "issue:46512", + "accept": false, + "reason": "Both concern dtype propagation, but one is model submodule dtype on load and the other is stale config.dtype from from_config." + }, + { + "left": "issue:44849", + "right": "issue:45701", + "accept": false, + "reason": "Qwen3.5 hidden-states bug vs tokenizer behavior change across versions; not the same bug." + }, + { + "left": "issue:27957", + "right": "issue:28282", + "accept": false, + "reason": "Flash Attention 2 support issue vs missing PyTorch environment dependency." + }, + { + "left": "issue:44987", + "right": "issue:45907", + "accept": false, + "reason": "Model loading regression vs list_repo_files revision forwarding; different APIs." + }, + { + "left": "issue:41720", + "right": "issue:43540", + "accept": false, + "reason": "Auto device-map CUDA assert vs video-input processing error in Qwen3OmniMoe; different failure modes." + }, + { + "left": "issue:43540", + "right": "issue:44560", + "accept": false, + "reason": "Both are video-input bugs, but they affect different models and likely different code paths." + }, + { + "left": "issue:44479", + "right": "issue:45381", + "accept": true, + "reason": "Same underlying Qwen2.5-VL video regression: incorrect vision_position_ids in the video path." + }, + { + "left": "issue:42175", + "right": "issue:44488", + "accept": false, + "reason": "TensorFlow backend packaging issue vs loading a specific model; unrelated." + }, + { + "left": "issue:34567", + "right": "issue:36331", + "accept": false, + "reason": "TrainerState token accounting bug vs CustomTrainer compute_loss signature mismatch." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet: issue 43749 is a reasonable anchor given its open state, discussion, and inbound references, and the soft-edge judgments stay mostly conservative. The two accepted pairs are plausible duplicates from the issue titles, and the rejected pairs are not overstated." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 26, + "review_unit_index": 23, + "review_unit_count": 61, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-505", + "nodes": [ + "issue:24643", + "issue:29870", + "issue:38175", + "issue:43257", + "issue:43452", + "issue:43531", + "issue:43540", + "issue:43577", + "issue:43582", + "issue:43697", + "issue:43716", + "issue:43742", + "issue:43756", + "issue:43906", + "issue:44062", + "issue:44291", + "issue:44488", + "issue:44568", + "issue:44589", + "issue:44625", + "issue:44737", + "issue:44792", + "issue:44821", + "issue:44877", + "issue:44898", + "issue:44964", + "issue:45042", + "issue:45072", + "issue:45081", + "issue:45325", + "issue:45405", + "issue:45488", + "issue:45588", + "issue:45600", + "issue:45663", + "issue:45672", + "issue:45674", + "issue:45823", + "issue:46399", + "issue:46463", + "issue:46468", + "issue:46491", + "issue:46693", + "issue:46701", + "issue:46821", + "issue:46829", + "issue:46846", + "issue:46899" + ], + "soft_pairs": [ + "issue:38175|issue:44898", + "issue:44792|issue:44964", + "issue:44568|issue:45488", + "issue:44062|issue:45081", + "issue:44291|issue:44821", + "issue:43257|issue:43756", + "issue:43582|issue:46821", + "issue:45042|issue:45081", + "issue:43742|issue:44589", + "issue:43756|issue:46399", + "issue:43756|issue:45325", + "issue:43716|issue:46701", + "issue:43577|issue:45072", + "issue:44568|issue:46491", + "issue:44737|issue:46463", + "issue:45405|issue:45600", + "issue:43531|issue:44821", + "issue:45672|issue:45674", + "issue:45823|issue:46463", + "issue:43906|issue:46693", + "issue:24643|issue:29870", + "issue:43452|issue:44488", + "issue:43906|issue:46468", + "issue:44625|issue:44877", + "issue:43540|issue:46693", + "issue:45663|issue:46846", + "issue:43697|issue:46829", + "issue:44737|issue:45588", + "issue:46463|issue:46899" + ], + "prepared_review_unit_hash": "0f1e931427547e4bf150ec8b338ff816e756a8f37e6ff6be15b15efc70dc3335", + "component_budget": { + "node_count": 505, + "item_count": 505, + "soft_pair_count": 1854, + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 29, + "serialized_chars": 36848, + "estimated_input_tokens": 9212, + "estimated_eval_tokens": 18680 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is a heterogeneous mix of unrelated bug reports across tokenizers, multimodal models, quantization, distributed training, and config loading. None of the soft pairs look like the same underlying bug or change, so all soft edges are rejected.", + "confidence": 0.94, + "canonical_issue_reason": "If a single issue must anchor the cluster, issue 45488 is the strongest candidate because it is the most detailed and broadly actionable report, but there is no real duplicate center in this set.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 45488 is the best representative only as a high-signal, well-described issue; it should not be treated as a duplicate hub for the other items.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:38175", + "right": "issue:44898", + "accept": false, + "reason": "Both are vision-related regressions, but they involve different models, different symptoms, and different code paths." + }, + { + "left": "issue:44792", + "right": "issue:44964", + "accept": false, + "reason": "Different multimodal models with unrelated failures: a Janus test regression versus Phi-4 model loading." + }, + { + "left": "issue:44568", + "right": "issue:45488", + "accept": false, + "reason": "Both are tokenizer regressions, but they affect different tokenizers and different breakages (special tokens handling vs pre-tokenizer override)." + }, + { + "left": "issue:44062", + "right": "issue:45081", + "accept": false, + "reason": "Separate tokenizer bugs: AddedToken argument duplication versus a Mistral regex patch crash." + }, + { + "left": "issue:44291", + "right": "issue:44821", + "accept": false, + "reason": "One is about init_empty_weights unexpected args; the other is AutoImageProcessor URL loading. No shared bug." + }, + { + "left": "issue:43257", + "right": "issue:43756", + "accept": false, + "reason": "Different Qwen/Smollm model issues with unrelated root causes (MoE weight conversion vs RoPE layer dropping)." + }, + { + "left": "issue:43582", + "right": "issue:46821", + "accept": false, + "reason": "Completely different runtime failures in unrelated components: Apple Silicon allocator warmup versus TimesFM window_size handling." + }, + { + "left": "issue:45042", + "right": "issue:45081", + "accept": false, + "reason": "Image processor torchvision dependency regression is unrelated to the Mistral tokenizer regex crash." + }, + { + "left": "issue:43742", + "right": "issue:44589", + "accept": false, + "reason": "Different load-time errors with different mechanisms: config/key loading versus Float8 storage lookup." + }, + { + "left": "issue:43756", + "right": "issue:46399", + "accept": false, + "reason": "Both touch Qwen/attention internals, but one is a RoPE configuration bug and the other is a kernel registration crash." + }, + { + "left": "issue:43756", + "right": "issue:45325", + "accept": false, + "reason": "Different Qwen model families and different rotary-position bugs; not the same underlying issue." + }, + { + "left": "issue:43716", + "right": "issue:46701", + "accept": false, + "reason": "Unrelated multimodal bugs: image preprocessor dtype mismatch versus a pipeline/output wiring issue." + }, + { + "left": "issue:43577", + "right": "issue:45072", + "accept": false, + "reason": "Both mention dtype mismatches, but they occur in different models and different execution paths." + }, + { + "left": "issue:44568", + "right": "issue:46491", + "accept": false, + "reason": "Two distinct tokenizer regressions affecting different model families and different round-trip behaviors." + }, + { + "left": "issue:44737", + "right": "issue:46463", + "accept": false, + "reason": "XLNet CPU placement bug is unrelated to T5 decoder input validation messaging." + }, + { + "left": "issue:45405", + "right": "issue:45600", + "accept": false, + "reason": "A dependency version bump problem is not the same as broken auto_mappings references to removed configs." + }, + { + "left": "issue:43531", + "right": "issue:44821", + "accept": false, + "reason": "Sliding-window behavior in Qwen3-MoE is unrelated to loading AutoImageProcessor from a URL." + }, + { + "left": "issue:45672", + "right": "issue:45674", + "accept": false, + "reason": "Both involve 4-bit inference, but the root causes differ: unsafe dtype cast in Gemma4 audio/vision modules versus BitsAndBytes skip-module exclusion logic." + }, + { + "left": "issue:45823", + "right": "issue:46463", + "accept": false, + "reason": "Gemma4 device placement during forward is unrelated to the T5 decoder-input error message bug." + }, + { + "left": "issue:43906", + "right": "issue:46693", + "accept": false, + "reason": "The titles indicate different problems: an isolated repro of an older issue versus a new FA2 long-context performance regression." + }, + { + "left": "issue:24643", + "right": "issue:29870", + "accept": false, + "reason": "DeepSpeed training weight-shape failure and BART generate decoder_attention_mask bug are distinct issues." + }, + { + "left": "issue:43452", + "right": "issue:44488", + "accept": false, + "reason": "A gguf_file-related from_pretrained breakage is not the same as a specific model failing to load." + }, + { + "left": "issue:43906", + "right": "issue:46468", + "accept": false, + "reason": "An isolated repro of a prior issue and a broken Nemotron example are different failures with no clear shared code path." + }, + { + "left": "issue:44625", + "right": "issue:44877", + "accept": false, + "reason": "Qwen3.5 label propagation and granite_speech strict-config loading are unrelated config bugs." + }, + { + "left": "issue:43540", + "right": "issue:46693", + "accept": false, + "reason": "Video-input processing in Qwen3OmniMoe is unrelated to FA2 long-context performance regression." + }, + { + "left": "issue:45663", + "right": "issue:46846", + "accept": false, + "reason": "Different distributed-training failures in different models: Gemma4 FSDP2 KeyError versus Qwen3.5 TP plan omissions." + }, + { + "left": "issue:43697", + "right": "issue:46829", + "accept": false, + "reason": "Different model outputs/outputs structure bugs in different vision models, not the same defect." + }, + { + "left": "issue:44737", + "right": "issue:45588", + "accept": false, + "reason": "XLNet CPU allocation bug and flash_attention s_aux=None crash are unrelated." + }, + { + "left": "issue:46463", + "right": "issue:46899", + "accept": false, + "reason": "T5 decoder validation messaging has nothing to do with Gemma4 multimodal 4-bit quantization dtype casting." + } + ] + }, + "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": 26, + "review_unit_index": 24, + "review_unit_count": 61, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-505", + "nodes": [ + "issue:28282", + "issue:32090", + "issue:38175", + "issue:41762", + "issue:43295", + "issue:43452", + "issue:43525", + "issue:43526", + "issue:43531", + "issue:43643", + "issue:43653", + "issue:43688", + "issue:43697", + "issue:43720", + "issue:43742", + "issue:43749", + "issue:43761", + "issue:43819", + "issue:43994", + "issue:44361", + "issue:44442", + "issue:44466", + "issue:44493", + "issue:44811", + "issue:44843", + "issue:44857", + "issue:44898", + "issue:44913", + "issue:44936", + "issue:45005", + "issue:45200", + "issue:45216", + "issue:45356", + "issue:45357", + "issue:45372", + "issue:45375", + "issue:45479", + "issue:45920", + "issue:46050", + "issue:46121", + "issue:46170", + "issue:46463", + "issue:46701", + "issue:46829", + "issue:46832", + "issue:46932", + "issue:47000" + ], + "soft_pairs": [ + "issue:45357|issue:45920", + "issue:28282|issue:32090", + "issue:44913|issue:46121", + "issue:43526|issue:45479", + "issue:43697|issue:45216", + "issue:43720|issue:44898", + "issue:44466|issue:45005", + "issue:43643|issue:45375", + "issue:45200|issue:46463", + "issue:44493|issue:46050", + "issue:41762|issue:44936", + "issue:43452|issue:44843", + "issue:43761|issue:46829", + "issue:43688|issue:43749", + "issue:44361|issue:44442", + "issue:38175|issue:43295", + "issue:43653|issue:44361", + "issue:43819|issue:44811", + "issue:43994|issue:44898", + "issue:45372|issue:46170", + "issue:44857|issue:47000", + "issue:46832|issue:46932", + "issue:43525|issue:46701", + "issue:43531|issue:43742", + "issue:43295|issue:45356" + ], + "prepared_review_unit_hash": "d27eac30ff96a244a85f0cbd37db6fcf44b24a3e191c7dbaf3cd8f88fe57a71e", + "component_budget": { + "node_count": 505, + "item_count": 505, + "soft_pair_count": 1854, + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 25, + "serialized_chars": 35141, + "estimated_input_tokens": 8786, + "estimated_eval_tokens": 17828 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "Mostly heterogeneous issue cluster with one likely duplicate pair around tied-weights serialization in v5. Most other pairs are different model/component failures and should stay separate.", + "confidence": 0.84, + "canonical_issue_reason": "issue:45005 is the broadest and clearest representative of the tied-weights v5 regression, and it plausibly subsumes the narrower lm_head.weight serialization report in issue:44466.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45005 best captures the underlying bug class and is the most suitable anchor for deduping the tied-weights serialization reports.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45357", + "right": "issue:45920", + "accept": false, + "reason": "Different regressions: Qwen3.5 visual encoder key serialization vs broad tokenizer token-id corruption across many models." + }, + { + "left": "issue:28282", + "right": "issue:32090", + "accept": false, + "reason": "Unrelated failures: missing PyTorch import for AutoModel vs Trainer GPU broadcast TypeError." + }, + { + "left": "issue:44913", + "right": "issue:46121", + "accept": false, + "reason": "Both rope-related, but one is config reload defaulting and the other is a JSON-list type error in validation; not the same bug." + }, + { + "left": "issue:43526", + "right": "issue:45479", + "accept": false, + "reason": "Different model families and failure modes: BeitImageProcessorFast label reduction vs degenerate classification loss." + }, + { + "left": "issue:43697", + "right": "issue:45216", + "accept": false, + "reason": "Different concrete bugs: RTDetrV2 output mismatch vs Qwen3.5 save_pretrained checkpoint corruption." + }, + { + "left": "issue:43720", + "right": "issue:44898", + "accept": false, + "reason": "Different subsystems and symptoms: BitNet packed-weight loading vs Perceiver positional interpolation failure." + }, + { + "left": "issue:44466", + "right": "issue:45005", + "accept": true, + "reason": "Both describe the same tied-weights serialization/save_pretrained regression in v5; 44466 looks like a narrower instance of the broader 45005 issue." + }, + { + "left": "issue:43643", + "right": "issue:45375", + "accept": false, + "reason": "Different config-field loss bugs: missing fields from trust_remote_code AutoConfig vs missing deepstack_visual_indexes in Qwen3_5MoeVisionConfig." + }, + { + "left": "issue:45200", + "right": "issue:46463", + "accept": false, + "reason": "Different bugs: Gemma 4 mm_token_type_ids defaulting issue vs T5 decoder input validation error messaging." + }, + { + "left": "issue:44493", + "right": "issue:46050", + "accept": false, + "reason": "Different serialization problems: unexpected position-id keys vs sinq quantization save/load breakage." + }, + { + "left": "issue:41762", + "right": "issue:44936", + "accept": false, + "reason": "Different failures: Gemma3 ZeRO-3 loading index error vs trainer.evaluate after trainer.train." + }, + { + "left": "issue:43452", + "right": "issue:44843", + "accept": false, + "reason": "Different causes: gguf_file handling in AutoTokenizer/AutoModelForCausalLM vs offline-model_info call in _patch_mistral_regex." + }, + { + "left": "issue:43761", + "right": "issue:46829", + "accept": false, + "reason": "Different output bugs in different models: CLIPVisionModel hidden_states regression vs VideoPrism hidden_state field mix-up." + }, + { + "left": "issue:43688", + "right": "issue:43749", + "accept": false, + "reason": "Different training/loading issues: auxiliary-loss normalization vs FSDP CPU RAM efficient loading breakage." + }, + { + "left": "issue:44361", + "right": "issue:44442", + "accept": false, + "reason": "Both tokenizer-related, but different tokenizers and failure types; not the same underlying bug." + }, + { + "left": "issue:38175", + "right": "issue:43295", + "accept": false, + "reason": "Different regressions: siglip2 zero probabilities vs processor.tokenizer/image-to-tokenizer regression." + }, + { + "left": "issue:43653", + "right": "issue:44361", + "accept": false, + "reason": "Different tokenizer bugs: BigBird special-token registration vs MLukeTokenizer attribute error on tasks." + }, + { + "left": "issue:43819", + "right": "issue:44811", + "accept": false, + "reason": "Different code paths: DAC latent conversion mismatch vs Whisper batch_decode skip_special_tokens handling." + }, + { + "left": "issue:43994", + "right": "issue:44898", + "accept": false, + "reason": "Different model inference issues: siglip2 AutoModel/pipeline nonsense outputs vs Perceiver interpolation failure." + }, + { + "left": "issue:45372", + "right": "issue:46170", + "accept": false, + "reason": "Unrelated import/dependency problems: mistral_common ReasoningEffort import vs Windows ROCm torch.distributed dependency." + }, + { + "left": "issue:44857", + "right": "issue:47000", + "accept": false, + "reason": "Different RT-DETR losses with different failure modes: AMP/CUDA crash vs infeasible cost matrix." + }, + { + "left": "issue:46832", + "right": "issue:46932", + "accept": false, + "reason": "Different models and bugs: RT-DETR backbone feature-level mismatch vs Gemma4 quantized embedder dtype casting." + }, + { + "left": "issue:43525", + "right": "issue:46701", + "accept": false, + "reason": "Unrelated issues: Llama4Config missing pad_token_id vs DiffusionGemma pipeline failure." + }, + { + "left": "issue:43531", + "right": "issue:43742", + "accept": false, + "reason": "Different model-specific errors: Qwen3-MoE sliding window behavior vs MobileLLM key error on load." + }, + { + "left": "issue:43295", + "right": "issue:45356", + "accept": false, + "reason": "Different regressions in different layers: processor/tokenizer image handling vs Kimi-K2.5 tokenizer codec/regex warning issues." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The assessment is grounded in the packet and stays conservative overall. The only accepted pair, issue:44466 \u2194 issue:45005, is plausibly the same tied-weights/save_pretrained regression at different specificity levels. The remaining soft-edge rejections are supported by clear title-level differences 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": 26, + "review_unit_index": 25, + "review_unit_count": 61, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-505", + "nodes": [ + "issue:29159", + "issue:32090", + "issue:33357", + "issue:33666", + "issue:36010", + "issue:41720", + "issue:42898", + "issue:43452", + "issue:43504", + "issue:43550", + "issue:43577", + "issue:43646", + "issue:43653", + "issue:43688", + "issue:43720", + "issue:43749", + "issue:43756", + "issue:43761", + "issue:43782", + "issue:43927", + "issue:44361", + "issue:44464", + "issue:44466", + "issue:44479", + "issue:44484", + "issue:44568", + "issue:44625", + "issue:44704", + "issue:44734", + "issue:44743", + "issue:44821", + "issue:44843", + "issue:44977", + "issue:44991", + "issue:45440", + "issue:45491", + "issue:45542", + "issue:45701", + "issue:45704", + "issue:46002", + "issue:46050", + "issue:46178", + "issue:46402", + "issue:46491", + "issue:46612", + "issue:46693", + "issue:46821", + "issue:47013" + ], + "soft_pairs": [ + "issue:29159|issue:33357", + "issue:45542|issue:46178", + "issue:43653|issue:44568", + "issue:43646|issue:43749", + "issue:43688|issue:45704", + "issue:43452|issue:43577", + "issue:43688|issue:45701", + "issue:44843|issue:44991", + "issue:43761|issue:47013", + "issue:42898|issue:46491", + "issue:43756|issue:46402", + "issue:44464|issue:44743", + "issue:43653|issue:43927", + "issue:44479|issue:44625", + "issue:45491|issue:46002", + "issue:44484|issue:44704", + "issue:44821|issue:46821", + "issue:33666|issue:36010", + "issue:43550|issue:44361", + "issue:44466|issue:46050", + "issue:32090|issue:33666", + "issue:43688|issue:45440", + "issue:44977|issue:46693", + "issue:44464|issue:44977", + "issue:44734|issue:46612", + "issue:41720|issue:44464", + "issue:43504|issue:43550", + "issue:43504|issue:43720", + "issue:43782|issue:46693" + ], + "prepared_review_unit_hash": "7ca687a5ddab8b50da23aed8d9dd494bbb1848f1df6be0c5adcd22811dd8abfd", + "component_budget": { + "node_count": 505, + "item_count": 505, + "soft_pair_count": 1854, + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 29, + "serialized_chars": 36521, + "estimated_input_tokens": 9131, + "estimated_eval_tokens": 18518 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is only loosely related by broad Transformer/tokenization/generation regression themes. The soft pairs mostly point to different code paths or different models/backends, so none should be merged as duplicates. The best representative issue is the slow-vs-fast tokenizer inconsistency report.", + "confidence": 0.64, + "canonical_issue_reason": "issue:29159 is the clearest, most central tokenizer-regression report here: it describes a concrete inconsistency between slow and fast tokenizer behavior, which is the broadest shared theme among the tokenizer-related items.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:29159 is the best single issue to anchor this cluster because it is concrete, widely applicable across tokenizer implementations, and closer to a canonical duplicate target than the more model-specific or version-specific reports.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:29159", + "right": "issue:33357", + "accept": false, + "reason": "Tokenizer behavior inconsistency vs MacOS bus error loading a community CLIP model; unrelated failures." + }, + { + "left": "issue:45542", + "right": "issue:46178", + "accept": false, + "reason": "Both mention backend/import problems, but one is TensorBoard-without-TensorFlow and the other is tensorflow_text backend mapping; different bugs." + }, + { + "left": "issue:43653", + "right": "issue:44568", + "accept": false, + "reason": "Both are tokenizer special-token regressions, but BigBird mask-token registration and mDeBERTa BOS/EOS insertion are different concrete issues." + }, + { + "left": "issue:43646", + "right": "issue:43749", + "accept": false, + "reason": "Custom model initialization regression vs FSDP CPU RAM efficient loading breakage; same broad area, not the same bug." + }, + { + "left": "issue:43688", + "right": "issue:45704", + "accept": false, + "reason": "Auxiliary-loss normalization bug vs T5/apex RMSNorm memory leak; unrelated." + }, + { + "left": "issue:43452", + "right": "issue:43577", + "accept": false, + "reason": "gguf_file loading failure vs Blip2 dtype propagation issue; different model-loading paths and symptoms." + }, + { + "left": "issue:43688", + "right": "issue:45701", + "accept": false, + "reason": "Auxiliary-loss normalization vs generic tokenization changes across versions; not the same underlying problem." + }, + { + "left": "issue:44843", + "right": "issue:44991", + "accept": false, + "reason": "Offline model_info call in AutoTokenizer patch vs tokenizer loading failure for a specific repo; different root causes." + }, + { + "left": "issue:43761", + "right": "issue:47013", + "accept": false, + "reason": "CLIP hidden_states regression vs xLSTM return-last-states unpacking crash; both are output-shape issues but on different models/code paths." + }, + { + "left": "issue:42898", + "right": "issue:46491", + "accept": false, + "reason": "cleanup-tokenization-space behavior change vs CodeLlama leading-space strip regression; both tokenization-related but not the same bug." + }, + { + "left": "issue:43756", + "right": "issue:46402", + "accept": false, + "reason": "Smollm3 RoPE-layer mismatch vs Sam3 text encoder weights not loading; unrelated model-specific loading/config issues." + }, + { + "left": "issue:44464", + "right": "issue:44743", + "accept": false, + "reason": "Chunked generation with compiled forward vs recurrent state reset when seq_len>1; both generation-related, but different failure modes and code paths." + }, + { + "left": "issue:43653", + "right": "issue:43927", + "accept": false, + "reason": "BigBird special-token registration bug vs DiaConfig losing custom token IDs after save/load; both config/token metadata issues, not the same defect." + }, + { + "left": "issue:44479", + "right": "issue:44625", + "accept": false, + "reason": "Video-input regression in Qwen VL models vs num_labels propagation in Qwen3.5; different model components." + }, + { + "left": "issue:45491", + "right": "issue:46002", + "accept": false, + "reason": "Gemma3 mixed-length NaN embeddings vs NaNs in classification heads after checkout/init; both NaN reports but unrelated contexts." + }, + { + "left": "issue:44484", + "right": "issue:44704", + "accept": false, + "reason": "Question about save_pretrained shard size vs AutoProcessor kwargs forwarding bug; not duplicates." + }, + { + "left": "issue:44821", + "right": "issue:46821", + "accept": false, + "reason": "AutoImageProcessor URL loading issue vs TimesFM window_size AttributeError; unrelated." + }, + { + "left": "issue:33666", + "right": "issue:36010", + "accept": false, + "reason": "Qwen2-VL multi-GPU training issue vs GenerationMixin import error; unrelated." + }, + { + "left": "issue:43550", + "right": "issue:44361", + "accept": false, + "reason": "torch.compile/SDPA failure in Bamba vs MLukeTokenizer AttributeError; different subsystems." + }, + { + "left": "issue:44466", + "right": "issue:46050", + "accept": false, + "reason": "Inconsistent lm_head serialization by device vs quantized-model save/load breakage; same general persistence area but different bugs." + }, + { + "left": "issue:32090", + "right": "issue:33666", + "accept": false, + "reason": "Trainer broadcast TypeError vs Qwen2-VL multi-GPU training; both training-related, not the same issue." + }, + { + "left": "issue:43688", + "right": "issue:45440", + "accept": false, + "reason": "Auxiliary-loss normalization vs DeepSeekV3MoE divergence from remote implementation; different model logic." + }, + { + "left": "issue:44977", + "right": "issue:46693", + "accept": false, + "reason": "Qwen3.5 flash-attention generation correctness regression vs FA2 long-context performance regression; related area, different problems." + }, + { + "left": "issue:44464", + "right": "issue:44977", + "accept": false, + "reason": "Compiled-forward chunked generation inconsistency vs flash-attention generation failure; not the same code-path issue." + }, + { + "left": "issue:44734", + "right": "issue:46612", + "accept": false, + "reason": "KV-cache continuation indexing crash in serve vs beam-search cache reorder skipped for certain models; both cache-related but distinct bugs." + }, + { + "left": "issue:41720", + "right": "issue:44464", + "accept": false, + "reason": "Auto device mapping CUDA assert vs chunked generation/compiled forward inconsistency; unrelated." + }, + { + "left": "issue:43504", + "right": "issue:43550", + "accept": false, + "reason": "Legacy field prevents Beit semantic-segmentation load vs Bamba torch.compile SDPA failure; unrelated." + }, + { + "left": "issue:43504", + "right": "issue:43720", + "accept": false, + "reason": "Legacy-field load failure vs BitNet packed-weight unpacking failure; different model-loading defects." + }, + { + "left": "issue:43782", + "right": "issue:46693", + "accept": false, + "reason": "Qwen3VL weight_only loading error vs FA2 long-context performance regression; different failure modes." + } + ] + }, + "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": 26, + "review_unit_index": 26, + "review_unit_count": 61, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-505", + "nodes": [ + "issue:28218", + "issue:33357", + "issue:33666", + "issue:36331", + "issue:39692", + "issue:42796", + "issue:43262", + "issue:43278", + "issue:43295", + "issue:43452", + "issue:43493", + "issue:43504", + "issue:43525", + "issue:43572", + "issue:43577", + "issue:43606", + "issue:43611", + "issue:43643", + "issue:43646", + "issue:43688", + "issue:43720", + "issue:43825", + "issue:43844", + "issue:43927", + "issue:43994", + "issue:44361", + "issue:44451", + "issue:44466", + "issue:44514", + "issue:44554", + "issue:44561", + "issue:44568", + "issue:44610", + "issue:44625", + "issue:44964", + "issue:44977", + "issue:45290", + "issue:45405", + "issue:45584", + "issue:45823", + "issue:46050", + "issue:46192", + "issue:46209", + "issue:46365", + "issue:46489", + "issue:46710", + "issue:46846" + ], + "soft_pairs": [ + "issue:43577|issue:44554", + "issue:43611|issue:46846", + "issue:43278|issue:43688", + "issue:33666|issue:36331", + "issue:43262|issue:43295", + "issue:43720|issue:43927", + "issue:43606|issue:43927", + "issue:43504|issue:44610", + "issue:43643|issue:44561", + "issue:39692|issue:43493", + "issue:44514|issue:45290", + "issue:44964|issue:45405", + "issue:43720|issue:44361", + "issue:43493|issue:43825", + "issue:44554|issue:46365", + "issue:43525|issue:43572", + "issue:44977|issue:46209", + "issue:44625|issue:45584", + "issue:43452|issue:44451", + "issue:45823|issue:46050", + "issue:28218|issue:33357", + "issue:43994|issue:46192", + "issue:42796|issue:44977", + "issue:39692|issue:43295", + "issue:43493|issue:43844", + "issue:43278|issue:43844", + "issue:43646|issue:44466", + "issue:44568|issue:46489", + "issue:46209|issue:46710", + "issue:33357|issue:33666" + ], + "prepared_review_unit_hash": "cd5030d7f39e558e999f63331cc1f985bd45f0b417141ef420c4ee99d55f47b4", + "component_budget": { + "node_count": 505, + "item_count": 505, + "soft_pair_count": 1854, + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 30, + "serialized_chars": 36349, + "estimated_input_tokens": 9088, + "estimated_eval_tokens": 18432 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The set is heterogeneous and the proposed soft links are mostly title-level similarity, not the same concrete bug or change. I found no strong duplicate cluster and would not merge any of the soft pairs.", + "confidence": 0.89, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43577", + "right": "issue:44554", + "accept": false, + "reason": "Different problems: one is dtype staying fp32 in BLIP2 loading; the other is an MPS attention correctness issue with mismatched head dims." + }, + { + "left": "issue:43611", + "right": "issue:46846", + "accept": false, + "reason": "Both mention model loading, but one is a base_model_prefix regression and the other is a tensor-parallel planning/OOM issue in Qwen3.5." + }, + { + "left": "issue:43278", + "right": "issue:43688", + "accept": false, + "reason": "Both involve numerical behavior, but one is eval/training dtype drift and the other is auxiliary-loss normalization in mixture-of-experts models." + }, + { + "left": "issue:33666", + "right": "issue:36331", + "accept": false, + "reason": "Completely different areas: Qwen2-VL multi-GPU training vs a CustomTrainer compute_loss signature regression." + }, + { + "left": "issue:43262", + "right": "issue:43295", + "accept": false, + "reason": "Audio processor sampling-rate defaulting and processor.tokenizer/image-handling regression are separate processor bugs." + }, + { + "left": "issue:43720", + "right": "issue:43927", + "accept": false, + "reason": "Different failure modes: unpacked packed weights during accelerate loading vs DiaConfig losing custom token IDs after save/load." + }, + { + "left": "issue:43606", + "right": "issue:43927", + "accept": false, + "reason": "Unrelated code paths: CPU offload device mismatch in Bark vs config serialization/token-ID loss in Dia." + }, + { + "left": "issue:43504", + "right": "issue:44610", + "accept": false, + "reason": "Different models and bugs: Beit legacy-field loading vs OmDet-Turbo processor/model input-size mismatch." + }, + { + "left": "issue:43643", + "right": "issue:44561", + "accept": false, + "reason": "One is missing fields from AutoConfig with trust_remote_code; the other is a missing utility function breaking remote-code models." + }, + { + "left": "issue:39692", + "right": "issue:43493", + "accept": false, + "reason": "Both concern SigLIP2, but one is a documentation/example issue and the other is an implementation mismatch with the original JAX model." + }, + { + "left": "issue:44514", + "right": "issue:45290", + "accept": false, + "reason": "Both are chat-template related, but one crashes on batched padding=False input and the other on assistant tool-call messages with no content." + }, + { + "left": "issue:44964", + "right": "issue:45405", + "accept": false, + "reason": "Model loading failure vs dependency-version bump; the latter is about an unreleased PEFT minimum version, not the Phi-4 multimodal bug." + }, + { + "left": "issue:43720", + "right": "issue:44361", + "accept": false, + "reason": "Different tokenizer/model areas: BitNet accelerate loading of packed weights vs MLukeTokenizer AttributeError on tasks." + }, + { + "left": "issue:43493", + "right": "issue:43825", + "accept": false, + "reason": "SigLIP2 implementation discrepancy and pipeline error-message wording are unrelated issues." + }, + { + "left": "issue:44554", + "right": "issue:46365", + "accept": false, + "reason": "Attention correctness on MPS is unrelated to an import-time Trainer/datasets dependency problem." + }, + { + "left": "issue:43525", + "right": "issue:43572", + "accept": false, + "reason": "Both mention missing token-related fields, but Llama4 pad_token_id and StableLm pad_token_idx are different configs and regressions." + }, + { + "left": "issue:44977", + "right": "issue:46209", + "accept": false, + "reason": "Qwen3.5 flash-attention generation bug is unrelated to FP8 inference failure on B200 hardware." + }, + { + "left": "issue:44625", + "right": "issue:45584", + "accept": false, + "reason": "Different models and symptoms: Qwen3.5 num_labels propagation vs Whisper empty-transcription failure after token alignment." + }, + { + "left": "issue:43452", + "right": "issue:44451", + "accept": false, + "reason": "gguf_file loading regression is not the same as inability to load a specific pretrained model." + }, + { + "left": "issue:45823", + "right": "issue:46050", + "accept": false, + "reason": "Device mismatch during forward with device_map=auto is different from quantized-model save/load corruption." + }, + { + "left": "issue:28218", + "right": "issue:33357", + "accept": false, + "reason": "Tokenizer extra-space behavior and a Mac bus error on CLIP are unrelated; similarity is only that both affect tokenization/model usage." + }, + { + "left": "issue:43994", + "right": "issue:46192", + "accept": false, + "reason": "SigLIP2 wrong outputs with AutoModel/pipeline and return_tensors being ignored when text_kwargs is passed are separate API bugs." + }, + { + "left": "issue:42796", + "right": "issue:44977", + "accept": false, + "reason": "Ministral BPE-marker handling and Qwen3.5 flash-attention generation are different tokenizer vs inference-path issues." + }, + { + "left": "issue:39692", + "right": "issue:43295", + "accept": false, + "reason": "Documentation example errors for SigLIP2 and a regression in processor.tokenizer/image handling are not the same bug." + }, + { + "left": "issue:43493", + "right": "issue:43844", + "accept": false, + "reason": "SigLIP2 implementation mismatch and gradient blow-up under DeepSpeed ZeRO-3 are unrelated training dynamics bugs." + }, + { + "left": "issue:43278", + "right": "issue:43844", + "accept": false, + "reason": "A dtype mismatch in evaluation is not the same as abnormal gradient growth during ZeRO-3 training." + }, + { + "left": "issue:43646", + "right": "issue:44466", + "accept": false, + "reason": "Custom model initialization breakage in v5 and tied lm_head serialization depending on device are different regressions." + }, + { + "left": "issue:44568", + "right": "issue:46489", + "accept": false, + "reason": "Both concern tokenizer behavior, but one is missing BOS/EOS addition and the other is DeepSeek-Coder BPE output corruption after a prior fix." + }, + { + "left": "issue:46209", + "right": "issue:46710", + "accept": false, + "reason": "Hardware-specific FP8 inference failure is not the same as an accuracy regression on DeepSeek-R1 after upgrade." + }, + { + "left": "issue:33357", + "right": "issue:33666", + "accept": false, + "reason": "A macOS bus error in CLIP and Qwen2-VL multi-GPU training are unrelated issues." + } + ] + }, + "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": 26, + "review_unit_index": 27, + "review_unit_count": 61, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-505", + "nodes": [ + "issue:43257", + "issue:43278", + "issue:43299", + "issue:43334", + "issue:43504", + "issue:43572", + "issue:43575", + "issue:43606", + "issue:43653", + "issue:43701", + "issue:43824", + "issue:43844", + "issue:43873", + "issue:43927", + "issue:43937", + "issue:43994", + "issue:44112", + "issue:44291", + "issue:44361", + "issue:44464", + "issue:44488", + "issue:44493", + "issue:44568", + "issue:44589", + "issue:44792", + "issue:44843", + "issue:44849", + "issue:44964", + "issue:45072", + "issue:45229", + "issue:45245", + "issue:45357", + "issue:45405", + "issue:45520", + "issue:45759", + "issue:45812", + "issue:45823", + "issue:46050", + "issue:46209", + "issue:46285", + "issue:46399", + "issue:46424", + "issue:46701", + "issue:46710", + "issue:46821", + "issue:46832", + "issue:46846" + ], + "soft_pairs": [ + "issue:44488|issue:44843", + "issue:44849|issue:46209", + "issue:43278|issue:43701", + "issue:44568|issue:45812", + "issue:44792|issue:45405", + "issue:43927|issue:44361", + "issue:43575|issue:45405", + "issue:43824|issue:44589", + "issue:44849|issue:45357", + "issue:46821|issue:46832", + "issue:43504|issue:43606", + "issue:43257|issue:43299", + "issue:45245|issue:46710", + "issue:43334|issue:45759", + "issue:44493|issue:46285", + "issue:43334|issue:43572", + "issue:44589|issue:45520", + "issue:43572|issue:44291", + "issue:46399|issue:46846", + "issue:43606|issue:44112", + "issue:45072|issue:46832", + "issue:43504|issue:43653", + "issue:44464|issue:46424", + "issue:43937|issue:46701", + "issue:43873|issue:46050", + "issue:43844|issue:44964", + "issue:43994|issue:44493", + "issue:45229|issue:45823" + ], + "prepared_review_unit_hash": "50ca0adb3d94b05cbfbb1a77d60381aab80fcd7a0167d5280eb3b4c7bcf3c571", + "component_budget": { + "node_count": 505, + "item_count": 505, + "soft_pair_count": 1854, + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 28, + "serialized_chars": 35616, + "estimated_input_tokens": 8904, + "estimated_eval_tokens": 18064 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The set is highly heterogeneous: most items are distinct model-loading, tokenizer, device/offload, quantization, or CI regressions across different model families. None of the proposed soft edges look like the same underlying bug/change, so they should all be rejected.", + "confidence": 0.95, + "canonical_issue_reason": "No safe canonical issue: these issues span unrelated regressions (Qwen/VL loading, tokenizer special-token handling, offload/device mismatches, quantization persistence, config errors, CI failures), so a single issue would be misleading as a duplicate target.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 44843 is the broadest single user-facing regression in the set, but it only covers the AutoTokenizer/_patch_mistral_regex offline-path bug and does not subsume the rest.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44488", + "right": "issue:44843", + "accept": false, + "reason": "Both are model-loading related, but one is a specific cjvt/sleng-bert load failure and the other is an HF_HUB_OFFLINE/model_info regression; different code paths." + }, + { + "left": "issue:44849", + "right": "issue:46209", + "accept": false, + "reason": "Qwen3.5 hidden-state output bug vs FP8 inference on B200 are unrelated features and failure modes." + }, + { + "left": "issue:43278", + "right": "issue:43701", + "accept": false, + "reason": "BF16-to-FP32 eval dtype change is unrelated to a resume_from_checkpoint key mismatch." + }, + { + "left": "issue:44568", + "right": "issue:45812", + "accept": false, + "reason": "Both concern tokenizer behavior, but one is mDeBERTa special-token insertion and the other is a Granite model token-id regression; not the same bug." + }, + { + "left": "issue:44792", + "right": "issue:45405", + "accept": false, + "reason": "A janus image-generation test failure is unrelated to a PEFT version bump." + }, + { + "left": "issue:43927", + "right": "issue:44361", + "accept": false, + "reason": "DiaConfig custom token IDs after save/load and MLukeTokenizer task AttributeError involve different models and different failure paths." + }, + { + "left": "issue:43575", + "right": "issue:45405", + "accept": false, + "reason": "Qwen2 TP OOM has nothing to do with the MIN_PEFT_VERSION release issue." + }, + { + "left": "issue:43824", + "right": "issue:44589", + "accept": false, + "reason": "Missing Qwen2_5_VL import and Float8 storage lookup failure are unrelated import/runtime errors." + }, + { + "left": "issue:44849", + "right": "issue:45357", + "accept": false, + "reason": "Qwen3.5 hidden-states behavior and Qwen3.5 visual-encoder save_pretrained keys are different bugs in different subsystems." + }, + { + "left": "issue:46821", + "right": "issue:46832", + "accept": false, + "reason": "TimesFM window_size AttributeError and RT-DETR feature-level mismatch are unrelated model-specific issues." + }, + { + "left": "issue:43504", + "right": "issue:43606", + "accept": false, + "reason": "BEiT legacy-field loading and bark-small CPU offload device mismatch do not share the same underlying code path." + }, + { + "left": "issue:43257", + "right": "issue:43299", + "accept": false, + "reason": "Both mention Qwen3 MoE loading, but one is accelerate+deepspeed weight conversion and the other is Qwen3VL loading in dev; too different to merge." + }, + { + "left": "issue:45245", + "right": "issue:46710", + "accept": false, + "reason": "2^24 category limit and DeepSeek accuracy regression are unrelated." + }, + { + "left": "issue:43334", + "right": "issue:45759", + "accept": false, + "reason": "Both touch Qwen multimodal configs, but one is a missing pad_token_id attribute and the other is from_config not unwrapping text_config; different fixes." + }, + { + "left": "issue:44493", + "right": "issue:46285", + "accept": false, + "reason": "Unexpected position-id keys across many models vs CLIP text-model structure inconsistency are different regressions." + }, + { + "left": "issue:43334", + "right": "issue:43572", + "accept": false, + "reason": "Qwen3-VL pad_token_id and StableLm pad_token_idx are different missing-field errors in different configs." + }, + { + "left": "issue:44589", + "right": "issue:45520", + "accept": false, + "reason": "Float8 storage lookup failure and flash_attn KeyError on Python 3.13 are unrelated dependency/runtime issues." + }, + { + "left": "issue:43572", + "right": "issue:44291", + "accept": false, + "reason": "StableLm missing pad_token_idx and init_empty_weights unexpected _is_hf_initialized argument are separate config/API regressions." + }, + { + "left": "issue:46399", + "right": "issue:46846", + "accept": false, + "reason": "Qwen3.5 missing decorator for apply_rotary_pos_emb registration and missing linear_attn TP-plan entries are different Qwen3.5 bugs." + }, + { + "left": "issue:43606", + "right": "issue:44112", + "accept": false, + "reason": "CPU offload device mismatch and a stale GraniteSpeech CI override test are not the same bug." + }, + { + "left": "issue:45072", + "right": "issue:46832", + "accept": false, + "reason": "bfloat16 dtype mismatches in SwitchTransformers/TimmWrapperModel are unrelated to RT-DETR feature-level shape errors." + }, + { + "left": "issue:43504", + "right": "issue:43653", + "accept": false, + "reason": "BEiT legacy-field loading and BigBirdTokenizer mask-token registration are different issues despite both involving model/tokenizer setup." + }, + { + "left": "issue:44464", + "right": "issue:46424", + "accept": false, + "reason": "Chunked generation inconsistency with compiled forward is not the same as cache max length only applying to quantized cache." + }, + { + "left": "issue:43937", + "right": "issue:46701", + "accept": false, + "reason": "Invalid GenerationConfig for GLM-5 and DiffusionGemma pipeline failure are unrelated generation problems." + }, + { + "left": "issue:43873", + "right": "issue:46050", + "accept": false, + "reason": "Quantization/offload not working and quantized model save/load breakage are related broadly, but they describe different concrete failures and paths." + }, + { + "left": "issue:43844", + "right": "issue:44964", + "accept": false, + "reason": "ZeRO-3 gradient growth and Phi-4 multimodal loading failure are unrelated." + }, + { + "left": "issue:45229", + "right": "issue:45823", + "accept": false, + "reason": "Gemma4 multi-GPU OOM and Gemma4 PLE device mismatch both mention Gemma4, but one is memory pressure and the other is a forward device-placement bug." + }, + { + "left": "issue:45357", + "right": "issue:45812", + "accept": false, + "reason": "Qwen3.5 visual-encoder key regression and Granite tokenizer ID regression are different model families and different serialization/tokenization problems." + }, + { + "left": "issue:46050", + "right": "issue:46873", + "accept": false, + "reason": "No such pair in the input cluster." + } + ] + }, + "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": 26, + "review_unit_index": 28, + "review_unit_count": 61, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-505", + "nodes": [ + "issue:15354", + "issue:28218", + "issue:29159", + "issue:29942", + "issue:30333", + "issue:32090", + "issue:34567", + "issue:37428", + "issue:38175", + "issue:41720", + "issue:42490", + "issue:43065", + "issue:43295", + "issue:43450", + "issue:43493", + "issue:43540", + "issue:43572", + "issue:43696", + "issue:43701", + "issue:43873", + "issue:43957", + "issue:44186", + "issue:44423", + "issue:44554", + "issue:44610", + "issue:44734", + "issue:44869", + "issue:44977", + "issue:45072", + "issue:45084", + "issue:45198", + "issue:45200", + "issue:45229", + "issue:45406", + "issue:45446", + "issue:45520", + "issue:45588", + "issue:45759", + "issue:45823", + "issue:46050", + "issue:46121", + "issue:46192", + "issue:46369", + "issue:46701", + "issue:46710", + "issue:46752", + "issue:46821" + ], + "soft_pairs": [ + "issue:44869|issue:46752", + "issue:45200|issue:45229", + "issue:45588|issue:46050", + "issue:29159|issue:32090", + "issue:42490|issue:43696", + "issue:29942|issue:30333", + "issue:28218|issue:34567", + "issue:44977|issue:46701", + "issue:45084|issue:45520", + "issue:43065|issue:46369", + "issue:15354|issue:38175", + "issue:44554|issue:45446", + "issue:41720|issue:46710", + "issue:43493|issue:43701", + "issue:44423|issue:44734", + "issue:45406|issue:46121", + "issue:44977|issue:46710", + "issue:37428|issue:43957", + "issue:43295|issue:43450", + "issue:45072|issue:45198", + "issue:45588|issue:45759", + "issue:43540|issue:45200", + "issue:43295|issue:43873", + "issue:44186|issue:46192", + "issue:45823|issue:46192", + "issue:43572|issue:46821", + "issue:43295|issue:44610", + "issue:45200|issue:45823" + ], + "prepared_review_unit_hash": "e374a2c507739cae09605455730dcbd67f59963d4d3aa02b8c79d0a7b28a4e4f", + "component_budget": { + "node_count": 505, + "item_count": 505, + "soft_pair_count": 1854, + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 28, + "serialized_chars": 35531, + "estimated_input_tokens": 8883, + "estimated_eval_tokens": 18022 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is heterogeneous rather than duplicate-heavy: it mixes unrelated tokenizer, serving, quantization, device-mapping, and model-specific bugs. I rejected every soft pair as a true duplicate.", + "confidence": 0.97, + "canonical_issue_reason": "No real canonical duplicate set here; the items span unrelated bugs. If a representative must be chosen, issue:43295 is the cleanest regression report, but it does not subsume the others.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43295 is the strongest standalone representative because it has a concrete regression description and user-impacting repro, but the cluster is too heterogeneous to treat it as a true duplicate anchor.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44869", + "right": "issue:46752", + "accept": false, + "reason": "Different failure modes and code paths: Whisper word-timestamp decoding vs apply_chat_template on an empty list." + }, + { + "left": "issue:45200", + "right": "issue:45229", + "accept": false, + "reason": "Gemma 4 multimodal token-type defaults vs a multi-GPU OOM report; not the same bug." + }, + { + "left": "issue:45588", + "right": "issue:46050", + "accept": false, + "reason": "flash_attention.py s_aux=None crash and sinq save/load breakage are unrelated." + }, + { + "left": "issue:29159", + "right": "issue:32090", + "accept": false, + "reason": "Tokenizer slow/fast inconsistency vs Trainer _gpu_broadcast_one NoneType error." + }, + { + "left": "issue:42490", + "right": "issue:43696", + "accept": false, + "reason": "Local-path import warning and GPT-oss CUDA OOM are unrelated." + }, + { + "left": "issue:29942", + "right": "issue:30333", + "accept": false, + "reason": "Flash Attention 2 test failures vs MLFlow reporting after failed training jobs." + }, + { + "left": "issue:28218", + "right": "issue:34567", + "accept": false, + "reason": "Added-token spacing bug vs TrainerState num_input_tokens_seen not updating." + }, + { + "left": "issue:44977", + "right": "issue:46701", + "accept": false, + "reason": "Both involve generation, but Qwen3.5 flash-attention behavior and DiffusionGemma pipeline failure are different code paths." + }, + { + "left": "issue:45084", + "right": "issue:45520", + "accept": false, + "reason": "Compilation error for non-template nodes vs Python 3.13 flash_attn import KeyError." + }, + { + "left": "issue:43065", + "right": "issue:46369", + "accept": false, + "reason": "Same model family, but dummy Conv2d design and litetext memory overflow are not the same underlying bug." + }, + { + "left": "issue:15354", + "right": "issue:38175", + "accept": false, + "reason": "torch.jit.script GeneratorExp support and SigLIP2 zero-probability output are unrelated." + }, + { + "left": "issue:44554", + "right": "issue:45446", + "accept": false, + "reason": "MPS attention correctness issue and flex_attention import version check are different problems." + }, + { + "left": "issue:41720", + "right": "issue:46710", + "accept": false, + "reason": "Qwen3 device-mapping CUDA assert and DeepSeek accuracy regression are unrelated." + }, + { + "left": "issue:43493", + "right": "issue:43701", + "accept": false, + "reason": "SigLIP2 HF-vs-JAX discrepancy and resume_from_checkpoint key mismatch do not describe the same bug." + }, + { + "left": "issue:44423", + "right": "issue:44734", + "accept": false, + "reason": "Both are serve crashes, but multimodal continuous-batching and KV-cache continuation indexing are distinct issues." + }, + { + "left": "issue:45406", + "right": "issue:46121", + "accept": false, + "reason": "Gemma4Processor missing _tokenizer and rope-params list type handling are unrelated." + }, + { + "left": "issue:44977", + "right": "issue:46710", + "accept": false, + "reason": "Qwen3.5 flash-attention generation issue is not the same as a DeepSeek accuracy regression." + }, + { + "left": "issue:37428", + "right": "issue:43957", + "accept": false, + "reason": "Flash-attention import name breakage and meta-device model loading failures are separate bugs." + }, + { + "left": "issue:43295", + "right": "issue:43450", + "accept": false, + "reason": "Processor.tokenizer/images regression vs batched video-shape bug; both are processor-related but not the same defect." + }, + { + "left": "issue:45072", + "right": "issue:45198", + "accept": false, + "reason": "bfloat16 dtype mismatch CI issue and Wav2Vec2 save/tokenization failure are unrelated." + }, + { + "left": "issue:45588", + "right": "issue:45759", + "accept": false, + "reason": "flash_attention sinkless-model crash and multimodal text_config unwrapping are different code paths." + }, + { + "left": "issue:43540", + "right": "issue:45200", + "accept": false, + "reason": "Video-input processing error in Qwen3OmniMoe and Gemma 4 token-type defaults are unrelated." + }, + { + "left": "issue:43295", + "right": "issue:43873", + "accept": false, + "reason": "Processor regression and quantized offloading failure are different issues." + }, + { + "left": "issue:44186", + "right": "issue:46192", + "accept": false, + "reason": "LayoutLMv2Tokenizer crash on NER/padding and return_tensors handling are distinct tokenizer behaviors." + }, + { + "left": "issue:45823", + "right": "issue:46192", + "accept": false, + "reason": "Gemma4 device-mapping mismatch and return_tensors handling are unrelated." + }, + { + "left": "issue:43572", + "right": "issue:46821", + "accept": false, + "reason": "StableLmConfig pad_token_idx missing and TimesFM window_size AttributeError are different model/config bugs." + }, + { + "left": "issue:43295", + "right": "issue:44610", + "accept": false, + "reason": "Processor API regression and OmDet-Turbo input-size mismatch are separate bugs." + }, + { + "left": "issue:45200", + "right": "issue:45823", + "accept": false, + "reason": "Gemma4 token-type defaulting and Gemma4 PLE device mismatch are different failure modes." + } + ] + }, + "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": 26, + "review_unit_index": 29, + "review_unit_count": 61, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-505", + "nodes": [ + "issue:28218", + "issue:29870", + "issue:29942", + "issue:35141", + "issue:37428", + "issue:39692", + "issue:42490", + "issue:42915", + "issue:43232", + "issue:43262", + "issue:43278", + "issue:43295", + "issue:43334", + "issue:43450", + "issue:43452", + "issue:43493", + "issue:43525", + "issue:43550", + "issue:43572", + "issue:43575", + "issue:43653", + "issue:43824", + "issue:43844", + "issue:43856", + "issue:43950", + "issue:43975", + "issue:44112", + "issue:44466", + "issue:44521", + "issue:44554", + "issue:44737", + "issue:44849", + "issue:44987", + "issue:45127", + "issue:45200", + "issue:45290", + "issue:45356", + "issue:45381", + "issue:45520", + "issue:46178", + "issue:46192", + "issue:46209", + "issue:46285", + "issue:46512", + "issue:46894", + "issue:46899", + "issue:46932" + ], + "soft_pairs": [ + "issue:43550|issue:43844", + "issue:29870|issue:29942", + "issue:43450|issue:43575", + "issue:45520|issue:46178", + "issue:43572|issue:44987", + "issue:37428|issue:43824", + "issue:42490|issue:43493", + "issue:39692|issue:46192", + "issue:44554|issue:45356", + "issue:43232|issue:44849", + "issue:43295|issue:43493", + "issue:44521|issue:45290", + "issue:44466|issue:45127", + "issue:45381|issue:46209", + "issue:28218|issue:35141", + "issue:44737|issue:46894", + "issue:43278|issue:43856", + "issue:42915|issue:43824", + "issue:43824|issue:43975", + "issue:43653|issue:44112", + "issue:43452|issue:46512", + "issue:43950|issue:46285", + "issue:43262|issue:43450", + "issue:43334|issue:43525", + "issue:45200|issue:46899", + "issue:45200|issue:46932" + ], + "prepared_review_unit_hash": "0cb1483f57cfb619d72fc8aba8810be93426ffe973c4406d24d48402233febc4", + "component_budget": { + "node_count": 505, + "item_count": 505, + "soft_pair_count": 1854, + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 26, + "serialized_chars": 35372, + "estimated_input_tokens": 8843, + "estimated_eval_tokens": 17942 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "All soft-edge candidates look like false positives: they span unrelated tokenizer, config, import, generation, multimodal, and training bugs, not the same underlying change or defect.", + "confidence": 0.91, + "canonical_issue_reason": "issue:44521 is the most active/open report in the set with strong inbound discussion and a concrete, reproducible bug description, so it is the best single anchor if one issue must represent this mixed cluster.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:44521 has the clearest ongoing signal and the broadest direct follow-up potential among these unrelated issues, making it the most suitable representative issue.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43550", + "right": "issue:43844", + "accept": false, + "reason": "Different bugs: SDPA/torch.compile failure in Bamba-9B-v2 vs ZeRO-3 gradient instability in DeepSpeed." + }, + { + "left": "issue:29870", + "right": "issue:29942", + "accept": false, + "reason": "Bart generate() UnboundLocalError on decoder_attention_mask is unrelated to Flash Attention 2 test failures." + }, + { + "left": "issue:43450", + "right": "issue:43575", + "accept": false, + "reason": "Video processor batched-shape bug vs Qwen2-57B tensor-parallel OOM; different components and symptoms." + }, + { + "left": "issue:45520", + "right": "issue:46178", + "accept": false, + "reason": "Python 3.13 flash_attn KeyError is not the same as a BACKENDS_MAPPING tensorflow_text ValueError." + }, + { + "left": "issue:43572", + "right": "issue:44987", + "accept": false, + "reason": "Missing pad_token_idx in StableLmConfig is distinct from a model-loading failure on transformers>=5.1.0." + }, + { + "left": "issue:37428", + "right": "issue:43824", + "accept": false, + "reason": "Both are ImportErrors, but for different missing symbols/classes in different modules." + }, + { + "left": "issue:42490", + "right": "issue:43493", + "accept": false, + "reason": "Trailing-slash local-path import warning is unrelated to the SigLIP2 HF-vs-JAX discrepancy." + }, + { + "left": "issue:39692", + "right": "issue:46192", + "accept": false, + "reason": "Docs example/model-processor mismatch is not the same bug as return_tensors being ignored with text_kwargs." + }, + { + "left": "issue:44554", + "right": "issue:45356", + "accept": false, + "reason": "MPS attention correctness issue and Kimi-K2.5 tokenizer codec regression affect different code paths." + }, + { + "left": "issue:43232", + "right": "issue:44849", + "accept": false, + "reason": "Generation kwarg sync_gpus handling is unrelated to the Qwen3.5 output_hidden_states bug." + }, + { + "left": "issue:43295", + "right": "issue:43493", + "accept": false, + "reason": "processor.tokenizer regression for custom model code is unrelated to the SigLIP2 implementation discrepancy." + }, + { + "left": "issue:44521", + "right": "issue:45290", + "accept": false, + "reason": "Both involve apply_chat_template, but one is all-zero assistant masks for multimodal input and the other is a tool-call crash; different failure modes." + }, + { + "left": "issue:44466", + "right": "issue:45127", + "accept": false, + "reason": "Serialization of lm_head.weight by device is not the same as LoRA merge collapse with extended vocabulary." + }, + { + "left": "issue:45381", + "right": "issue:46209", + "accept": false, + "reason": "Qwen2.5-VL video position ids bug is unrelated to FP8 inference failure on B200." + }, + { + "left": "issue:28218", + "right": "issue:35141", + "accept": false, + "reason": "Tokenizer extra-space insertion and embedding reinitialization in post_init are separate issues." + }, + { + "left": "issue:44737", + "right": "issue:46894", + "accept": false, + "reason": "XLNet CPU placement bug is unrelated to the CohereASR training-loss double-shift problem." + }, + { + "left": "issue:43278", + "right": "issue:43856", + "accept": false, + "reason": "BF16-to-FP32 dtype drift in evaluate is not the same as MoE training memory inefficiency." + }, + { + "left": "issue:42915", + "right": "issue:43824", + "accept": false, + "reason": "Qwen3Moe FineGrainedFP8Config failure and missing Qwen2_5_VLForConditionalGeneration import are different defects." + }, + { + "left": "issue:43824", + "right": "issue:43975", + "accept": false, + "reason": "A missing exported class ImportError is unrelated to incorrect detokenization in v5." + }, + { + "left": "issue:43653", + "right": "issue:44112", + "accept": false, + "reason": "BigBirdTokenizer special-token registration bug is unrelated to a stale GraniteSpeech CI test." + }, + { + "left": "issue:43452", + "right": "issue:46512", + "accept": false, + "reason": "gguf_file loading failures and from_config(dtype=...) leaving config.dtype stale are different lifecycle problems." + }, + { + "left": "issue:43950", + "right": "issue:46285", + "accept": false, + "reason": "Non-persistent buffer corruption in from_pretrained is not the same as CLIP text-model structural inconsistency." + }, + { + "left": "issue:43262", + "right": "issue:43450", + "accept": false, + "reason": "Audio chat_template sampling-rate default bug is unrelated to batched video processor shape output." + }, + { + "left": "issue:43334", + "right": "issue:43525", + "accept": false, + "reason": "Both are pad_token_id AttributeErrors, but on different model/config classes; insufficient evidence of one root cause." + }, + { + "left": "issue:45200", + "right": "issue:46899", + "accept": false, + "reason": "Gemma4 token-type-id defaults and 4-bit quantized embedder dtype casting are separate multimodal bugs." + }, + { + "left": "issue:45200", + "right": "issue:46932", + "accept": false, + "reason": "Same family, but token-type-id defaults and quantized vision/audio embedder casting do not look like the same underlying fix." + } + ] + }, + "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": 26, + "review_unit_index": 30, + "review_unit_count": 61, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-505", + "nodes": [ + "issue:28282", + "issue:29942", + "issue:32090", + "issue:33357", + "issue:37428", + "issue:38175", + "issue:41720", + "issue:43335", + "issue:43388", + "issue:43425", + "issue:43504", + "issue:43526", + "issue:43550", + "issue:43606", + "issue:43611", + "issue:43653", + "issue:43819", + "issue:43827", + "issue:43901", + "issue:44112", + "issue:44206", + "issue:44442", + "issue:44466", + "issue:44479", + "issue:44561", + "issue:44610", + "issue:44792", + "issue:44857", + "issue:44898", + "issue:45083", + "issue:45198", + "issue:45200", + "issue:45325", + "issue:45412", + "issue:45520", + "issue:45563", + "issue:45584", + "issue:45593", + "issue:45759", + "issue:46170", + "issue:46192", + "issue:46209", + "issue:46234", + "issue:46285", + "issue:46424", + "issue:46512", + "issue:46821", + "issue:46858" + ], + "soft_pairs": [ + "issue:44442|issue:45198", + "issue:44792|issue:45563", + "issue:44898|issue:45412", + "issue:44479|issue:46192", + "issue:45200|issue:45325", + "issue:43819|issue:44112", + "issue:41720|issue:46209", + "issue:43335|issue:43550", + "issue:44561|issue:46234", + "issue:43827|issue:43901", + "issue:44610|issue:44857", + "issue:43819|issue:45083", + "issue:43388|issue:43526", + "issue:46512|issue:46821", + "issue:45759|issue:46285", + "issue:44857|issue:45593", + "issue:44206|issue:45584", + "issue:43606|issue:45083", + "issue:38175|issue:45412", + "issue:46424|issue:46858", + "issue:43335|issue:43653", + "issue:37428|issue:43425", + "issue:28282|issue:33357", + "issue:29942|issue:32090", + "issue:43550|issue:44112", + "issue:43611|issue:44466", + "issue:43335|issue:43504", + "issue:45520|issue:46170" + ], + "prepared_review_unit_hash": "8c6e550b0e7f2b07f48848c547f9ebe8bccbf49c9fa071e2b1c9b364b1c4de14", + "component_budget": { + "node_count": 505, + "item_count": 505, + "soft_pair_count": 1854, + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 28, + "serialized_chars": 36413, + "estimated_input_tokens": 9104, + "estimated_eval_tokens": 18464 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The pairwise similarities are mostly superficial; these issues describe different models, different subsystems, or different failure modes, so none of the soft edges should be collapsed.", + "confidence": 0.95, + "canonical_issue_reason": "issue:44479 is the clearest self-contained regression report in the set, with explicit affected models and a concrete user-facing failure.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:44479 is the best representative issue because it is specific, reproducible, and describes a broad regression affecting multiple Qwen VL variants.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44442", + "right": "issue:45198", + "accept": false, + "reason": "Different tokenizer families and failure modes: AutoTokenizer loading FastSpeech2ConformerTokenizer vs Wav2Vec2 save_pretrained/tokenization behavior." + }, + { + "left": "issue:44792", + "right": "issue:45563", + "accept": false, + "reason": "Both mention generation-related tests, but one is a janus image generation failure and the other is a stale warning about num_return_sequences." + }, + { + "left": "issue:44898", + "right": "issue:45412", + "accept": false, + "reason": "Unrelated bugs: Perceiver image classification resolution handling vs RT-DETR memory not being released." + }, + { + "left": "issue:44479", + "right": "issue:46192", + "accept": false, + "reason": "Both touch multimodal inputs, but one is a video-input regression in Qwen VL models while the other is text_kwargs/return_tensors precedence." + }, + { + "left": "issue:45200", + "right": "issue:45325", + "accept": false, + "reason": "Different models and code paths: Gemma 4 text-only mm_token_type_ids defaulting vs Qwen2.5-VL rope index scaling for still images." + }, + { + "left": "issue:43819", + "right": "issue:44112", + "accept": false, + "reason": "Different components entirely: DAC latent conversion mismatch vs a GraniteSpeech CI device-override test failure." + }, + { + "left": "issue:41720", + "right": "issue:46209", + "accept": false, + "reason": "Both are CUDA/runtime issues, but they affect different models, hardware, and root causes." + }, + { + "left": "issue:43335", + "right": "issue:43550", + "accept": false, + "reason": "SwitchTransformers sparse-layer config bug is unrelated to Bamba torch.compile/SDPA failure." + }, + { + "left": "issue:44561", + "right": "issue:46234", + "accept": false, + "reason": "Both involve import_utils helpers, but they are missing different symbols and stem from different compatibility breaks." + }, + { + "left": "issue:43827", + "right": "issue:43901", + "accept": false, + "reason": "Both are docs issues, but they concern different docs pages and different API behaviors." + }, + { + "left": "issue:44610", + "right": "issue:44857", + "accept": false, + "reason": "Different vision models and failures: incorrect processor image size vs mixed-precision loss crash." + }, + { + "left": "issue:43819", + "right": "issue:45083", + "accept": false, + "reason": "No shared bug beyond being in the same general audio/multimodal area; DAC latent mismatch is unrelated to qwen3_omni_moe feature-length arithmetic." + }, + { + "left": "issue:43388", + "right": "issue:43526", + "accept": false, + "reason": "Both mention labels, but one is a distributed metrics batching bug and the other is an image processor reduce_labels bug." + }, + { + "left": "issue:46512", + "right": "issue:46821", + "accept": false, + "reason": "Completely different issues: stale config.dtype after from_config vs TimesFM window_size AttributeError." + }, + { + "left": "issue:45759", + "right": "issue:46285", + "accept": false, + "reason": "Both involve model/config structure, but the concrete problems and affected architectures differ." + }, + { + "left": "issue:44857", + "right": "issue:45593", + "accept": false, + "reason": "Different detection-model training bugs: AMP/CUDA crash versus auxiliary losses being skipped when denoising is disabled." + }, + { + "left": "issue:44206", + "right": "issue:45584", + "accept": false, + "reason": "Different audio pipeline failures affecting different models and different trigger conditions." + }, + { + "left": "issue:43606", + "right": "issue:45083", + "accept": false, + "reason": "CPU offload device mismatch for bark-small is unrelated to qwen3_omni_moe feature-length helper behavior." + }, + { + "left": "issue:38175", + "right": "issue:45412", + "accept": false, + "reason": "Different symptoms and subsystems: zero probabilities in SigLIP2 vs RT-DETR memory leak." + }, + { + "left": "issue:46424", + "right": "issue:46858", + "accept": false, + "reason": "Both are cache-related, but one is about max_cache_length semantics and the other is a device-tensor return breaking static-cache slice arithmetic." + }, + { + "left": "issue:43335", + "right": "issue:43653", + "accept": false, + "reason": "Distinct bugs in different subsystems: SwitchTransformers sparse-layer construction vs BigBird tokenizer special-token registration." + }, + { + "left": "issue:37428", + "right": "issue:43425", + "accept": false, + "reason": "Missing import symbol in flash_attention_utils is not the same bug as Torch 2.10 incompatibility." + }, + { + "left": "issue:28282", + "right": "issue:33357", + "accept": false, + "reason": "One is a missing PyTorch ImportError, the other is a MacOS bus error in a specific CLIP model." + }, + { + "left": "issue:29942", + "right": "issue:32090", + "accept": false, + "reason": "Unrelated failures: Flash Attention 2 test regressions vs a Trainer broadcast TypeError." + }, + { + "left": "issue:43550", + "right": "issue:44112", + "accept": false, + "reason": "Different causes and code paths: Bamba torch.compile/SDPA failure vs GraniteSpeech stale device-override CI test." + }, + { + "left": "issue:43611", + "right": "issue:44466", + "accept": false, + "reason": "Both relate to model loading/serialization, but one is about base_model_prefix compatibility and the other about lm_head.weight serialization differing by device." + }, + { + "left": "issue:43335", + "right": "issue:43504", + "accept": false, + "reason": "Different models and bugs: sparse-layer creation in SwitchTransformers vs legacy-field loading failure in BeitForSemanticSegmentation." + }, + { + "left": "issue:45520", + "right": "issue:46170", + "accept": false, + "reason": "Separate environment-specific import problems: flash_attn key lookup on Python 3.13 vs Windows ROCm/torch.distributed dependency issues." + } + ] + }, + "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": 26, + "review_unit_index": 31, + "review_unit_count": 61, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-505", + "nodes": [ + "issue:26413", + "issue:28218", + "issue:29127", + "issue:31501", + "issue:32090", + "issue:34689", + "issue:35141", + "issue:37428", + "issue:39692", + "issue:42490", + "issue:43278", + "issue:43335", + "issue:43526", + "issue:43606", + "issue:43701", + "issue:43720", + "issue:43819", + "issue:43825", + "issue:43854", + "issue:44112", + "issue:44206", + "issue:44351", + "issue:44448", + "issue:44493", + "issue:44514", + "issue:44521", + "issue:44617", + "issue:44743", + "issue:44779", + "issue:44855", + "issue:44857", + "issue:45072", + "issue:45083", + "issue:45127", + "issue:45245", + "issue:45290", + "issue:45310", + "issue:45459", + "issue:45478", + "issue:45588", + "issue:45593", + "issue:45823", + "issue:45859", + "issue:46234", + "issue:46326", + "issue:46431", + "issue:46566", + "issue:46697" + ], + "soft_pairs": [ + "issue:44855|issue:45083", + "issue:44448|issue:44743", + "issue:37428|issue:44351", + "issue:44514|issue:44521", + "issue:43335|issue:45083", + "issue:45290|issue:46326", + "issue:43335|issue:45593", + "issue:43335|issue:44112", + "issue:29127|issue:32090", + "issue:43278|issue:43825", + "issue:45588|issue:45859", + "issue:42490|issue:43854", + "issue:34689|issue:35141", + "issue:43720|issue:44112", + "issue:43819|issue:44493", + "issue:39692|issue:43526", + "issue:46431|issue:46566", + "issue:43606|issue:44857", + "issue:45310|issue:45478", + "issue:44448|issue:44779", + "issue:28218|issue:32090", + "issue:39692|issue:44493", + "issue:44617|issue:46697", + "issue:45459|issue:46234", + "issue:26413|issue:31501", + "issue:45127|issue:45245", + "issue:44206|issue:45823", + "issue:43701|issue:45083", + "issue:44857|issue:45072" + ], + "prepared_review_unit_hash": "55ee729ed80bf1d4f4f4da54c7464eb1329f5e7d541505f1081585e731908f10", + "component_budget": { + "node_count": 505, + "item_count": 505, + "soft_pair_count": 1854, + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 29, + "serialized_chars": 36608, + "estimated_input_tokens": 9152, + "estimated_eval_tokens": 18560 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly a loose thematic grouping of unrelated bug reports. The only clear duplicate pair is the Qwen3.5 MoE `from_pretrained` regression (45310/45478); the rest are different code paths, models, or symptom classes.", + "confidence": 0.93, + "canonical_issue_reason": "Issue 45310 is the best canonical issue: it is the earlier report of the exact same Qwen3.5 MoE `from_pretrained` failure later repeated in 45478.", + "canonical_pr_reason": null, + "best_issue_reason": "45310 is the strongest representative of the cluster because it has the clearest true-duplicate match and the most stable, specific problem statement.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44855", + "right": "issue:45083", + "accept": false, + "reason": "Both are bugs, but they affect unrelated modules and failures: DeBERTa import parsing vs a qwen3_omni_moe helper behavior issue." + }, + { + "left": "issue:44448", + "right": "issue:44743", + "accept": false, + "reason": "Different models and different problems: Pegasus output drift across versions vs Qwen3.5 recurrent-state reset with cache." + }, + { + "left": "issue:37428", + "right": "issue:44351", + "accept": false, + "reason": "Both are import errors, but for different missing symbols in different subsystems; not the same underlying bug." + }, + { + "left": "issue:44514", + "right": "issue:44521", + "accept": false, + "reason": "Both involve `apply_chat_template`, but one is a batched-padding crash and the other is incorrect assistant masks for multimodal inputs." + }, + { + "left": "issue:43335", + "right": "issue:45083", + "accept": false, + "reason": "Unrelated issues: SwitchTransformers config bug vs qwen3_omni_moe feature-extraction behavior." + }, + { + "left": "issue:45290", + "right": "issue:46326", + "accept": false, + "reason": "Both touch chat templating/tokenization, but the failing paths differ: tool-call handling vs `continue_final_message` on mistral_common tokenizers." + }, + { + "left": "issue:43335", + "right": "issue:45593", + "accept": false, + "reason": "Different underlying bugs in different models: SwitchTransformers config creation vs D-FINE auxiliary-loss behavior." + }, + { + "left": "issue:43335", + "right": "issue:44112", + "accept": false, + "reason": "No shared code path; one is a model config bug and the other is a CI failure about stale device overrides." + }, + { + "left": "issue:29127", + "right": "issue:32090", + "accept": false, + "reason": "Completely different failures: LayoutLMv3 error-message clarity vs Trainer `_gpu_broadcast_one` type handling." + }, + { + "left": "issue:43278", + "right": "issue:43825", + "accept": false, + "reason": "Different symptoms and code paths: dtype drift in evaluation vs a pipeline task-support error message." + }, + { + "left": "issue:45588", + "right": "issue:45859", + "accept": false, + "reason": "Both involve offload/attention-related code, but one is `s_aux=None` handling in flash attention and the other is a CPU-offload incompatibility in rotary embeddings." + }, + { + "left": "issue:42490", + "right": "issue:43854", + "accept": false, + "reason": "A local-path import warning is unrelated to a unit-test model-loading failure for GLM-4.7-Flash." + }, + { + "left": "issue:34689", + "right": "issue:35141", + "accept": false, + "reason": "Different regressions: Llama 3.2 Vision model loading vs output embedding reinitialization after resizing token embeddings." + }, + { + "left": "issue:43720", + "right": "issue:44112", + "accept": false, + "reason": "Unrelated: packed-weight loading failure in BitNet vs a stale CI device-override test." + }, + { + "left": "issue:43819", + "right": "issue:44493", + "accept": false, + "reason": "Different bug classes: DAC latent/forward mismatch vs unexpected position-id keys in many models." + }, + { + "left": "issue:39692", + "right": "issue:43526", + "accept": false, + "reason": "Docs example errors for SigLIP2 are not the same as the BEiT Fast processor label-reduction bug." + }, + { + "left": "issue:46431", + "right": "issue:46566", + "accept": false, + "reason": "Different problems: missing chat template on a processor vs DiffusionGemma remaining on meta with `device_map=\"auto\"`." + }, + { + "left": "issue:43606", + "right": "issue:44857", + "accept": false, + "reason": "Both are runtime failures, but one is device mismatch with CPU offload and the other is an AMP/CUDA crash in a loss function." + }, + { + "left": "issue:45310", + "right": "issue:45478", + "accept": true, + "reason": "Exact same title and same Qwen3.5 MoE `from_pretrained` regression; these are clear duplicates of the same bug." + }, + { + "left": "issue:44448", + "right": "issue:44779", + "accept": false, + "reason": "Different models and failure modes: Pegasus output mismatch vs Deepseek tokenizer regression." + }, + { + "left": "issue:28218", + "right": "issue:32090", + "accept": false, + "reason": "Tokenizer spacing after added tokens is unrelated to a Trainer broadcast type error." + }, + { + "left": "issue:39692", + "right": "issue:44493", + "accept": false, + "reason": "Documentation/example issues are unrelated to unexpected position-id keys in model loading." + }, + { + "left": "issue:44617", + "right": "issue:46697", + "accept": false, + "reason": "CUDA OOM in a video model is not the same bug as over-padding images in Swin2SR." + }, + { + "left": "issue:45459", + "right": "issue:46234", + "accept": false, + "reason": "Different import-related issues: tokenizer error masking vs a missing utility symbol." + }, + { + "left": "issue:26413", + "right": "issue:31501", + "accept": false, + "reason": "Both touch training flow, but one is checkpoint resume epoch-iterator failure and the other is a training job that never starts." + }, + { + "left": "issue:45127", + "right": "issue:45245", + "accept": false, + "reason": "Unrelated: LoRA merge collapse with extended vocab vs a category-count limit error." + }, + { + "left": "issue:44206", + "right": "issue:45823", + "accept": false, + "reason": "Different model-specific failures: unsupported feature-extractor arg vs device-mapped forward mismatch." + }, + { + "left": "issue:44857", + "right": "issue:45072", + "accept": false, + "reason": "Both mention dtype issues, but they are different models and distinct failure paths." + }, + { + "left": "issue:43701", + "right": "issue:45083", + "accept": false, + "reason": "Checkpoint key mismatch is unrelated to a qwen3_omni_moe helper behavior bug." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is well grounded in the packet and stays conservative on the soft-edge pairs. The only accepted duplicate, issue:45310 / issue:45478, is clearly supported by the identical title and matching Qwen3.5 MoE from_pretrained failure. The rest of the pair judgments are appropriately cautious 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": 26, + "review_unit_index": 32, + "review_unit_count": 61, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-505", + "nodes": [ + "issue:10000", + "issue:15354", + "issue:25251", + "issue:27957", + "issue:29870", + "issue:29942", + "issue:32090", + "issue:33290", + "issue:33357", + "issue:34689", + "issue:35141", + "issue:37428", + "issue:38175", + "issue:40444", + "issue:42175", + "issue:42490", + "issue:43278", + "issue:43329", + "issue:43381", + "issue:43526", + "issue:43550", + "issue:43575", + "issue:43873", + "issue:43927", + "issue:43976", + "issue:44360", + "issue:44448", + "issue:44466", + "issue:44493", + "issue:44561", + "issue:44568", + "issue:44855", + "issue:44977", + "issue:45083", + "issue:45229", + "issue:45290", + "issue:45381", + "issue:45440", + "issue:45491", + "issue:45593", + "issue:45698", + "issue:45920", + "issue:46192", + "issue:46399", + "issue:46421", + "issue:46655", + "issue:46858", + "issue:47000" + ], + "soft_pairs": [ + "issue:32090|issue:33357", + "issue:42490|issue:43575", + "issue:45593|issue:47000", + "issue:43550|issue:45083", + "issue:27957|issue:34689", + "issue:29942|issue:35141", + "issue:38175|issue:44493", + "issue:29870|issue:32090", + "issue:43278|issue:43329", + "issue:45290|issue:45381", + "issue:45920|issue:46192", + "issue:44448|issue:44977", + "issue:10000|issue:15354", + "issue:44360|issue:44561", + "issue:44977|issue:45229", + "issue:44855|issue:46399", + "issue:29870|issue:33290", + "issue:45290|issue:46655", + "issue:43927|issue:45698", + "issue:40444|issue:43526", + "issue:43278|issue:43381", + "issue:37428|issue:42175", + "issue:32090|issue:33290", + "issue:46421|issue:46858", + "issue:43873|issue:45229", + "issue:42490|issue:43873", + "issue:25251|issue:32090", + "issue:43976|issue:45440", + "issue:44448|issue:44466", + "issue:44568|issue:45491" + ], + "prepared_review_unit_hash": "f6d2ae50e3f343379a60104f959718e73a1ea529e1002fa09715f44d75961210", + "component_budget": { + "node_count": 505, + "item_count": 505, + "soft_pair_count": 1854, + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 30, + "serialized_chars": 36980, + "estimated_input_tokens": 9245, + "estimated_eval_tokens": 18746 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This set does not form a single duplicate cluster; it\u2019s a collection of unrelated transformer bugs/regressions across different models and subsystems. None of the proposed soft edges look like the same underlying issue, so all are rejected.", + "confidence": 0.24, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": "issue:45920 is the strongest representative because it describes a broad, concrete v5 regression affecting multiple model families with clear user impact.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:32090", + "right": "issue:33357", + "accept": false, + "reason": "Different failures and stacks: Trainer GPU broadcast NoneType vs MacOS bus error in CLIP." + }, + { + "left": "issue:42490", + "right": "issue:43575", + "accept": false, + "reason": "Import warning for local paths is unrelated to Qwen2.5 tensor-parallel OOM on model load." + }, + { + "left": "issue:45593", + "right": "issue:47000", + "accept": false, + "reason": "Different detection bugs: missing auxiliary losses vs infeasible RTDetr cost matrix." + }, + { + "left": "issue:43550", + "right": "issue:45083", + "accept": false, + "reason": "torch.compile+SDPA crash in Bamba is not the same bug as qwen3_omni_moe feature-extraction helper behavior." + }, + { + "left": "issue:27957", + "right": "issue:34689", + "accept": false, + "reason": "Flash Attention 2 support for XLMRoberta is unrelated to Llama 3.2 Vision model loading breakage." + }, + { + "left": "issue:29942", + "right": "issue:35141", + "accept": false, + "reason": "Failing Flash Attention 2 tests do not match token-embedding resize/reinit behavior." + }, + { + "left": "issue:38175", + "right": "issue:44493", + "accept": false, + "reason": "Siglip2 zero-probability output issue and unexpected position-id keys are different symptoms and code paths." + }, + { + "left": "issue:29870", + "right": "issue:32090", + "accept": false, + "reason": "BART generate decoder mask assignment bug is unrelated to Trainer _gpu_broadcast_one NoneType error." + }, + { + "left": "issue:43278", + "right": "issue:43329", + "accept": false, + "reason": "Embedding dtype drift in eval is a different problem from undefined video-token helper variables." + }, + { + "left": "issue:45290", + "right": "issue:45381", + "accept": false, + "reason": "Chat-template tool-call crash and Qwen2.5-VL video position-id bug affect different functions." + }, + { + "left": "issue:45920", + "right": "issue:46192", + "accept": false, + "reason": "Both are tokenizer-related, but one is wrong token IDs and the other is return_tensors being ignored." + }, + { + "left": "issue:44448", + "right": "issue:44977", + "accept": false, + "reason": "Pegasus v4/v5 output divergence is not the same concrete bug as Qwen3.5 flash-attention generation failure." + }, + { + "left": "issue:10000", + "right": "issue:15354", + "accept": false, + "reason": "German DistilBert issue and torch.jit.script GeneratorExp export failure are unrelated." + }, + { + "left": "issue:44360", + "right": "issue:44561", + "accept": false, + "reason": "DSA indexer missing ReLU is unrelated to removal of is_torch_fx_available breaking trust_remote_code." + }, + { + "left": "issue:44977", + "right": "issue:45229", + "accept": false, + "reason": "Qwen3.5 flash-attention generation bug and Gemma4 multi-GPU OOM are different failure modes." + }, + { + "left": "issue:44855", + "right": "issue:46399", + "accept": false, + "reason": "Python 3.13 IndentationError from torch.jit.script parsing is unrelated to a missing decorator in apply_rotary_pos_emb registration." + }, + { + "left": "issue:29870", + "right": "issue:33290", + "accept": false, + "reason": "BART generation bug and DeepSpeed Adafactor OOM are different subsystems and symptoms." + }, + { + "left": "issue:45290", + "right": "issue:46655", + "accept": false, + "reason": "Assistant-tool-call chat template crash is not the same as SmolVLM video padding miscalculation." + }, + { + "left": "issue:43927", + "right": "issue:45698", + "accept": false, + "reason": "One is custom token IDs lost in config save/load; the other is the wrong custom module loaded after save_pretrained." + }, + { + "left": "issue:40444", + "right": "issue:43526", + "accept": false, + "reason": "Qwen2.5-VL iterable multi-image training failure and BeitImageProcessorFast reduce_labels bug are unrelated." + }, + { + "left": "issue:43278", + "right": "issue:43381", + "accept": false, + "reason": "Eval-time BF16->FP32 embedding dtype change is not the same bug as gradient checkpointing being disallowed in eval mode." + }, + { + "left": "issue:37428", + "right": "issue:42175", + "accept": false, + "reason": "Missing flash-support import and tensorflow not included in torch extra are different packaging/import issues." + }, + { + "left": "issue:32090", + "right": "issue:33290", + "accept": false, + "reason": "Trainer broadcast NoneType error is unrelated to OOM with Adafactor in DeepSpeed." + }, + { + "left": "issue:46421", + "right": "issue:46858", + "accept": false, + "reason": "Both involve cache/compile, but one is spurious recompilation from lazy init and the other is device-tensor seq-length arithmetic failure." + }, + { + "left": "issue:43873", + "right": "issue:45229", + "accept": false, + "reason": "Quantization offloading not working and Gemma4 multi-GPU CUDA OOM are not the same underlying bug." + }, + { + "left": "issue:42490", + "right": "issue:43873", + "accept": false, + "reason": "A trailing-slash local-model import warning is unrelated to quantization offloading behavior." + }, + { + "left": "issue:25251", + "right": "issue:32090", + "accept": false, + "reason": "Pipeline top_k nesting change has no overlap with Trainer GPU broadcast NoneType errors." + }, + { + "left": "issue:43976", + "right": "issue:45440", + "accept": false, + "reason": "Python-version compatibility issue is unrelated to DeepseekV3MoE divergence from remote implementation." + }, + { + "left": "issue:44448", + "right": "issue:44466", + "accept": false, + "reason": "Different regression types: model output mismatch vs lm_head serialization/tied-weights inconsistency." + }, + { + "left": "issue:44568", + "right": "issue:45491", + "accept": false, + "reason": "Tokenizer BOS/EOS omission and Gemma3 NaN embeddings are different code paths and failure modes." + } + ] + }, + "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": 26, + "review_unit_index": 33, + "review_unit_count": 61, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-505", + "nodes": [ + "issue:29127", + "issue:29870", + "issue:29942", + "issue:30064", + "issue:33357", + "issue:34567", + "issue:34689", + "issue:39692", + "issue:41720", + "issue:42915", + "issue:42994", + "issue:43065", + "issue:43278", + "issue:43381", + "issue:43412", + "issue:43575", + "issue:43611", + "issue:43653", + "issue:43701", + "issue:43756", + "issue:43994", + "issue:44186", + "issue:44206", + "issue:44265", + "issue:44361", + "issue:44442", + "issue:44479", + "issue:44561", + "issue:44743", + "issue:44779", + "issue:44871", + "issue:44977", + "issue:45198", + "issue:45229", + "issue:45245", + "issue:45372", + "issue:45406", + "issue:45412", + "issue:45440", + "issue:45459", + "issue:45491", + "issue:45563", + "issue:45588", + "issue:45712", + "issue:45803", + "issue:45823", + "issue:46121", + "issue:46693" + ], + "soft_pairs": [ + "issue:29127|issue:34567", + "issue:44561|issue:44779", + "issue:44871|issue:44977", + "issue:33357|issue:39692", + "issue:45588|issue:45823", + "issue:43381|issue:43701", + "issue:43278|issue:43575", + "issue:45198|issue:45803", + "issue:41720|issue:45229", + "issue:44977|issue:45563", + "issue:44186|issue:45198", + "issue:44361|issue:45198", + "issue:43611|issue:46121", + "issue:29870|issue:34689", + "issue:43653|issue:44743", + "issue:44206|issue:44479", + "issue:29942|issue:30064", + "issue:42915|issue:42994", + "issue:45372|issue:45588", + "issue:43756|issue:45440", + "issue:43653|issue:44442", + "issue:43412|issue:45588", + "issue:43065|issue:43994", + "issue:44265|issue:45198", + "issue:45491|issue:46693", + "issue:45459|issue:45712", + "issue:29127|issue:33357", + "issue:43065|issue:45412", + "issue:43575|issue:43701", + "issue:45372|issue:45406", + "issue:45245|issue:45563" + ], + "prepared_review_unit_hash": "5166ccbe164be9486c4b5c38c37dc3a34821569f103ebedcb448c9ff8ed1e7e2", + "component_budget": { + "node_count": 505, + "item_count": 505, + "soft_pair_count": 1854, + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 31, + "serialized_chars": 37086, + "estimated_input_tokens": 9272, + "estimated_eval_tokens": 18800 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "These items are a broad mix of unrelated bug reports across tokenizers, generation, export, processor loading, and model-specific regressions; none of the soft pairs look like true duplicates of the same underlying bug/change.", + "confidence": 0.92, + "canonical_issue_reason": "issue:46693 is the strongest representative issue: it is current, clearly framed as a concrete regression, and has active discussion/inbound references, making it a good cluster anchor despite the cluster being heterogeneous.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:46693 is the best overall issue to surface because it is recent, broadly impactful, and sufficiently concrete to track and triage.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:29127", + "right": "issue:34567", + "accept": false, + "reason": "Different subsystems and failure modes: LayoutLMv3 error messaging vs TrainerState token counting." + }, + { + "left": "issue:44561", + "right": "issue:44779", + "accept": false, + "reason": "One is a v5 compatibility break for trust_remote_code models; the other is a DeepSeek tokenizer correctness regression." + }, + { + "left": "issue:44871", + "right": "issue:44977", + "accept": false, + "reason": "Gemma eos_token_id mismatch and Qwen flash-attention generation bug are unrelated." + }, + { + "left": "issue:33357", + "right": "issue:39692", + "accept": false, + "reason": "MacOS bus error on CLIP is not the same as SigLIP2 documentation/example issues." + }, + { + "left": "issue:45588", + "right": "issue:45823", + "accept": false, + "reason": "Flash-attention sinkless-model crash and Gemma4 device-map mismatch are different code paths." + }, + { + "left": "issue:43381", + "right": "issue:43701", + "accept": false, + "reason": "Eval-mode gradient checkpointing limitation is unrelated to resume_from_checkpoint key mismatch." + }, + { + "left": "issue:43278", + "right": "issue:43575", + "accept": false, + "reason": "Training/eval dtype drift and tensor-parallel OOM are separate issues." + }, + { + "left": "issue:45198", + "right": "issue:45803", + "accept": false, + "reason": "Wav2Vec2 save/tokenization failure is unrelated to Fuyu's bare-except error handling bug." + }, + { + "left": "issue:41720", + "right": "issue:45229", + "accept": false, + "reason": "CUDA assert during auto device mapping differs from multi-GPU OOM on Gemma4." + }, + { + "left": "issue:44977", + "right": "issue:45563", + "accept": false, + "reason": "Generation failure with flash-attention and a stale warning in paged generate() are not the same bug." + }, + { + "left": "issue:44186", + "right": "issue:45198", + "accept": false, + "reason": "LayoutLMv2 tokenizer crashes on NER padding/truncation; Wav2Vec2 has save/tokenization failure." + }, + { + "left": "issue:44361", + "right": "issue:45198", + "accept": false, + "reason": "MLuke tokenizer AttributeError is unrelated to Wav2Vec2 save/tokenization issues." + }, + { + "left": "issue:43611", + "right": "issue:46121", + "accept": false, + "reason": "base_model_prefix loading break and rope-parameter JSON list TypeError are different changes." + }, + { + "left": "issue:29870", + "right": "issue:34689", + "accept": false, + "reason": "BART generate() UnboundLocalError and Llama 3.2 Vision loading break are unrelated." + }, + { + "left": "issue:43653", + "right": "issue:44743", + "accept": false, + "reason": "BigBird special-token registration and Qwen recurrent-state reset are different tokenizer/model bugs." + }, + { + "left": "issue:44206", + "right": "issue:44479", + "accept": false, + "reason": "Unsupported center arg in image processing is unrelated to the video-input regression across Qwen VL models." + }, + { + "left": "issue:29942", + "right": "issue:30064", + "accept": false, + "reason": "Failing Flash Attention 2 tests and void segmentation-map processing are not the same underlying bug." + }, + { + "left": "issue:42915", + "right": "issue:42994", + "accept": false, + "reason": "FineGrainedFP8Config failure and quantized model saving failure are different quantization issues." + }, + { + "left": "issue:45372", + "right": "issue:45588", + "accept": false, + "reason": "ReasoningEffort import break in Gemma4 processor loading is separate from flash_attention s_aux=None crash." + }, + { + "left": "issue:43756", + "right": "issue:45440", + "accept": false, + "reason": "Smollm3 RoPE-layer mismatch and DeepseekV3MoE divergence are model-specific but distinct." + }, + { + "left": "issue:43653", + "right": "issue:44442", + "accept": false, + "reason": "BigBird tokenizer decode bug and AutoTokenizer loading FastSpeech2ConformerTokenizer are unrelated." + }, + { + "left": "issue:43412", + "right": "issue:45588", + "accept": false, + "reason": "Executorch export/training-guard failure is unrelated to flash_attention integration crash." + }, + { + "left": "issue:43065", + "right": "issue:43994", + "accept": false, + "reason": "Sam3 dummy Conv2d cleanup and SigLIP2 nonsensical outputs are different issues." + }, + { + "left": "issue:44265", + "right": "issue:45198", + "accept": false, + "reason": "torch.export with torch_compilable_check and Wav2Vec2 save/tokenization failure are unrelated." + }, + { + "left": "issue:45491", + "right": "issue:46693", + "accept": false, + "reason": "Both mention FA2, but one is a NaN correctness bug and the other is a performance regression; not the same concrete fix." + }, + { + "left": "issue:45459", + "right": "issue:45712", + "accept": false, + "reason": "Tokenizer error masking via bare except and leftover dummy classes are unrelated repository issues." + }, + { + "left": "issue:29127", + "right": "issue:33357", + "accept": false, + "reason": "LayoutLMv3 error messaging and CLIP MacOS bus error do not share the same bug." + }, + { + "left": "issue:43065", + "right": "issue:45412", + "accept": false, + "reason": "Dummy class cleanup in SAM3 is unrelated to RT-DETR memory-release behavior." + }, + { + "left": "issue:43575", + "right": "issue:43701", + "accept": false, + "reason": "Tensor-parallel OOM on Qwen2 is unrelated to checkpoint key mismatch." + }, + { + "left": "issue:45372", + "right": "issue:45406", + "accept": false, + "reason": "Gemma4 processor import failure and serve-time missing _tokenizer attribute are related by model family but not clearly the same concrete bug." + }, + { + "left": "issue:45245", + "right": "issue:45563", + "accept": false, + "reason": "Category-count runtime error and stale num_return_sequences warning 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": 26, + "review_unit_index": 34, + "review_unit_count": 61, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-505", + "nodes": [ + "issue:27957", + "issue:29127", + "issue:33290", + "issue:34689", + "issue:35141", + "issue:42490", + "issue:42915", + "issue:43066", + "issue:43334", + "issue:43381", + "issue:43421", + "issue:43526", + "issue:43531", + "issue:43575", + "issue:43646", + "issue:43824", + "issue:43873", + "issue:43931", + "issue:43994", + "issue:44077", + "issue:44186", + "issue:44279", + "issue:44387", + "issue:44442", + "issue:44462", + "issue:44464", + "issue:44514", + "issue:44857", + "issue:45072", + "issue:45250", + "issue:45372", + "issue:45381", + "issue:45406", + "issue:45412", + "issue:45440", + "issue:45459", + "issue:45491", + "issue:45584", + "issue:45588", + "issue:45593", + "issue:45834", + "issue:46121", + "issue:46369", + "issue:46399", + "issue:46640", + "issue:46693", + "issue:46701", + "issue:46821" + ], + "soft_pairs": [ + "issue:42490|issue:43381", + "issue:29127|issue:35141", + "issue:45440|issue:45593", + "issue:43994|issue:45412", + "issue:45372|issue:45459", + "issue:45406|issue:46701", + "issue:43421|issue:45491", + "issue:45412|issue:46369", + "issue:43575|issue:44387", + "issue:43066|issue:44462", + "issue:44442|issue:45072", + "issue:43526|issue:45834", + "issue:44464|issue:45584", + "issue:43824|issue:44279", + "issue:42915|issue:43931", + "issue:43526|issue:46693", + "issue:27957|issue:33290", + "issue:27957|issue:29127", + "issue:44857|issue:45588", + "issue:43994|issue:46369", + "issue:43526|issue:44186", + "issue:33290|issue:34689", + "issue:33290|issue:35141", + "issue:43334|issue:43531", + "issue:46121|issue:46399", + "issue:45588|issue:46821", + "issue:43381|issue:43873", + "issue:45250|issue:46640", + "issue:44514|issue:45381", + "issue:43646|issue:44077" + ], + "prepared_review_unit_hash": "b9921f961547dd6e27140cb06ac2625a94f855b59f1380a917e1d257edf6331d", + "component_budget": { + "node_count": 505, + "item_count": 505, + "soft_pair_count": 1854, + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 30, + "serialized_chars": 36514, + "estimated_input_tokens": 9129, + "estimated_eval_tokens": 18514 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "These soft matches are mostly title-level similarity only; the paired issues describe different models, subsystems, or failure modes, so none should be deduplicated.", + "confidence": 0.94, + "canonical_issue_reason": "No clear canonical issue: the set is heterogeneous and the apparent similarities are mostly broad phrasing, not the same underlying bug.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:46640 is the closest umbrella for the only recurring theme (Flash Attention 2), but it is still too generic to serve as a true duplicate hub.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:42490", + "right": "issue:43381", + "accept": false, + "reason": "Local-path import warning vs gradient-checkpointing eval-mode behavior; unrelated bugs." + }, + { + "left": "issue:29127", + "right": "issue:35141", + "accept": false, + "reason": "LayoutLMv3 error-message clarity vs token-embedding resize/reinit; different code paths." + }, + { + "left": "issue:45440", + "right": "issue:45593", + "accept": false, + "reason": "DeepseekV3MoE output divergence vs D-FINE auxiliary-loss logic; unrelated model bugs." + }, + { + "left": "issue:43994", + "right": "issue:45412", + "accept": false, + "reason": "SigLIP2 bad outputs vs RT-DETR memory release; different symptoms and subsystems." + }, + { + "left": "issue:45372", + "right": "issue:45459", + "accept": false, + "reason": "Gemma processor import failure vs tokenizer-error masking when protobuf is missing; not the same bug." + }, + { + "left": "issue:45406", + "right": "issue:46701", + "accept": false, + "reason": "Gemma4Processor missing _tokenizer vs DiffusionGemma generation output/pipeline failure; different issues." + }, + { + "left": "issue:43421", + "right": "issue:45491", + "accept": false, + "reason": "Tokenizer post-processor update feature vs Gemma3 NaN embeddings with sliding-window attention; unrelated." + }, + { + "left": "issue:45412", + "right": "issue:46369", + "accept": false, + "reason": "RT-DETR deletion memory leak vs SAM3-LiteText memory overflow; both memory-related but distinct bugs." + }, + { + "left": "issue:43575", + "right": "issue:44387", + "accept": false, + "reason": "Both are OOM reports, but one is TP load OOM and the other is int4 quantization reserved-memory regression." + }, + { + "left": "issue:43066", + "right": "issue:44462", + "accept": false, + "reason": "Wrong tokenizer decoder type in v5 vs AutoTokenizer ignoring tokenizer.json; different loading failures." + }, + { + "left": "issue:44442", + "right": "issue:45072", + "accept": false, + "reason": "Tokenizer loading failure vs dtype mismatch in bfloat16 inference; unrelated." + }, + { + "left": "issue:43526", + "right": "issue:45834", + "accept": false, + "reason": "BEiT label reduction bug vs Kosmos2.5 long-OCR index error; different models and failure modes." + }, + { + "left": "issue:44464", + "right": "issue:45584", + "accept": false, + "reason": "Chunked generation inconsistency with compiled forward vs Whisper empty-transcription failure after token alignment; not the same bug." + }, + { + "left": "issue:43824", + "right": "issue:44279", + "accept": false, + "reason": "Missing Qwen2.5-VL import symbol vs vague dependency issue; too broad and not a shared concrete defect." + }, + { + "left": "issue:42915", + "right": "issue:43931", + "accept": false, + "reason": "Qwen3MoE FP8 config failure vs Qwen3-VL weight-shape mismatch; different models and root causes." + }, + { + "left": "issue:43334", + "right": "issue:43531", + "accept": false, + "reason": "Missing pad_token_id on Qwen3-VL vs sliding_window issue on Qwen3-MoE; unrelated." + }, + { + "left": "issue:46121", + "right": "issue:46399", + "accept": false, + "reason": "RoPE validation TypeError vs missing decorator for apply_rotary_pos_emb registration; unrelated code paths." + }, + { + "left": "issue:45588", + "right": "issue:46821", + "accept": false, + "reason": "flash_attention.py s_aux=None crash vs TimesFM window_size AttributeError; same broad area only, not same bug." + }, + { + "left": "issue:43381", + "right": "issue:43873", + "accept": false, + "reason": "Gradient checkpointing in eval mode vs offloading under quantization; unrelated." + }, + { + "left": "issue:45250", + "right": "issue:46640", + "accept": false, + "reason": "Both mention Flash Attention 2, but one is a generic feature/title and the other a specific issue report; not the same concrete change/bug." + }, + { + "left": "issue:44514", + "right": "issue:45381", + "accept": false, + "reason": "Batched apply_chat_template padding crash vs Qwen2.5-VL video position-id bug; different processing stages." + }, + { + "left": "issue:43646", + "right": "issue:44077", + "accept": false, + "reason": "Custom model initialization breakage vs patchtsmixer post_init policy; related only at a very high level." + }, + { + "left": "issue:27957", + "right": "issue:33290", + "accept": false, + "reason": "XLMRoberta Flash Attention 2 report vs Adafactor/deepspeed OOM; unrelated." + }, + { + "left": "issue:27957", + "right": "issue:29127", + "accept": false, + "reason": "Flash Attention 2 on XLMRoberta vs LayoutLMv3 error clarity; no underlying overlap." + }, + { + "left": "issue:44857", + "right": "issue:45588", + "accept": false, + "reason": "LwDetrImageLoss AMP crash vs flash-attention sinkless-model AttributeError; different failures." + }, + { + "left": "issue:43994", + "right": "issue:46369", + "accept": false, + "reason": "SigLIP2 nonsensical outputs vs SAM3-LiteText memory overflow; unrelated." + }, + { + "left": "issue:43526", + "right": "issue:44186", + "accept": false, + "reason": "BEiTImageProcessorFast reduce_labels bug vs LayoutLMv2Tokenizer NER/padding crash; different components." + }, + { + "left": "issue:33290", + "right": "issue:34689", + "accept": false, + "reason": "Adafactor/deepspeed OOM vs Llama 3.2 Vision model-loading breakage; unrelated." + }, + { + "left": "issue:33290", + "right": "issue:35141", + "accept": false, + "reason": "Optimizer OOM vs embedding reinitialization bug; different code paths." + }, + { + "left": "issue:43334", + "right": "issue:43531", + "accept": false, + "reason": "Qwen3-VL missing pad_token_id vs Qwen3-MoE sliding-window issue; unrelated." + }, + { + "left": "issue:46121", + "right": "issue:46399", + "accept": false, + "reason": "RoPE validation TypeError vs Qwen3.5 kernel registration decorator issue; unrelated." + }, + { + "left": "issue:45588", + "right": "issue:46821", + "accept": false, + "reason": "flash_attention.py AttributeError vs TimesFM window_size AttributeError; same generic error type only, not same bug." + }, + { + "left": "issue:43381", + "right": "issue:43873", + "accept": false, + "reason": "Eval-mode gradient checkpointing restriction vs quantization offloading behavior; unrelated." + }, + { + "left": "issue:45250", + "right": "issue:46640", + "accept": false, + "reason": "Generic Flash Attention 2 feature/title overlap is insufficient to merge as duplicates." + }, + { + "left": "issue:44514", + "right": "issue:45381", + "accept": false, + "reason": "Batched processor crash vs video vision-position-id bug; not the same underlying defect." + }, + { + "left": "issue:43646", + "right": "issue:44077", + "accept": false, + "reason": "Both mention initialization/post_init, but the reported problems are different and not mergeable into one fix." + } + ] + }, + "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": 26, + "review_unit_index": 35, + "review_unit_count": 61, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-505", + "nodes": [ + "issue:27957", + "issue:29870", + "issue:42175", + "issue:42490", + "issue:43232", + "issue:43262", + "issue:43316", + "issue:43388", + "issue:43493", + "issue:43525", + "issue:43643", + "issue:43720", + "issue:43749", + "issue:43761", + "issue:43827", + "issue:43873", + "issue:43931", + "issue:44062", + "issue:44077", + "issue:44079", + "issue:44190", + "issue:44206", + "issue:44442", + "issue:44479", + "issue:44496", + "issue:44743", + "issue:44811", + "issue:44855", + "issue:44871", + "issue:44898", + "issue:44913", + "issue:44977", + "issue:44987", + "issue:45003", + "issue:45137", + "issue:45216", + "issue:45245", + "issue:45325", + "issue:45440", + "issue:45759", + "issue:45803", + "issue:45859", + "issue:45907", + "issue:46178", + "issue:46234", + "issue:46697", + "issue:46829", + "issue:46932" + ], + "soft_pairs": [ + "issue:43761|issue:45859", + "issue:44977|issue:45245", + "issue:44743|issue:44811", + "issue:43262|issue:44206", + "issue:44442|issue:44898", + "issue:42490|issue:44743", + "issue:43720|issue:44442", + "issue:45003|issue:45440", + "issue:42175|issue:46234", + "issue:27957|issue:29870", + "issue:44077|issue:44479", + "issue:43493|issue:43827", + "issue:44062|issue:45907", + "issue:43316|issue:43525", + "issue:43873|issue:44871", + "issue:45325|issue:46697", + "issue:43643|issue:44913", + "issue:43749|issue:43931", + "issue:44855|issue:45137", + "issue:43316|issue:45759", + "issue:44987|issue:46178", + "issue:44442|issue:45803", + "issue:44190|issue:44496", + "issue:43761|issue:44079", + "issue:43931|issue:45216", + "issue:46829|issue:46932", + "issue:43232|issue:43388" + ], + "prepared_review_unit_hash": "40de04a84bfb01f5b7b27bafff921f590ed7a8272542e87444b7a2ac8543bc9d", + "component_budget": { + "node_count": 505, + "item_count": 505, + "soft_pair_count": 1854, + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 27, + "serialized_chars": 36167, + "estimated_input_tokens": 9042, + "estimated_eval_tokens": 18340 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is heterogeneous: the issues cover many unrelated models, loading paths, processors, and generation bugs. None of the soft pairs look like the same underlying bug/change, so all should be kept separate.", + "confidence": 0.97, + "canonical_issue_reason": "No single canonical issue stands out because the items span unrelated subsystems and failure modes.", + "canonical_pr_reason": null, + "best_issue_reason": "No issue here is a strong global representative; the cluster does not form one duplicate family.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43761", + "right": "issue:45859", + "accept": false, + "reason": "Both involve Qwen/rotary behavior, but one is a CLIPVision hidden-states regression and the other is CPU offload incompatibility in Qwen3_5Moe; different bugs and code paths." + }, + { + "left": "issue:44977", + "right": "issue:45245", + "accept": false, + "reason": "Qwen3.5 flash-attention generation failure is unrelated to a category-count runtime error in a separate vision/classification path." + }, + { + "left": "issue:44743", + "right": "issue:44811", + "accept": false, + "reason": "One is recurrent-state reset in modular_qwen3_5 cache handling; the other is Whisper processor batch_decode skip_special_tokens handling. Different components." + }, + { + "left": "issue:43262", + "right": "issue:44206", + "accept": false, + "reason": "Audio processor sampling-rate default bug and LasrFeatureExtractor unsupported center argument are distinct processor issues." + }, + { + "left": "issue:44442", + "right": "issue:44898", + "accept": false, + "reason": "FastSpeech2ConformerTokenizer loading failure is not the same as Perceiver interpolation/padding failure; different tokenizer vs image-model code." + }, + { + "left": "issue:42490", + "right": "issue:44743", + "accept": false, + "reason": "Local-path import warning and Qwen recurrent cache reset are unrelated." + }, + { + "left": "issue:43720", + "right": "issue:44442", + "accept": false, + "reason": "BitNet packed-weight unpacking during accelerate loading is a model-loading bug, not the tokenizer-loading failure described here." + }, + { + "left": "issue:45003", + "right": "issue:45440", + "accept": false, + "reason": "Unsafe sys.modules access in modeling_utils is a generic import issue, while DeepseekV3MoE divergence is a model-implementation mismatch." + }, + { + "left": "issue:42175", + "right": "issue:46234", + "accept": false, + "reason": "TensorFlow missing from a pip extra and a missing import_utils availability symbol are separate packaging/import problems." + }, + { + "left": "issue:27957", + "right": "issue:29870", + "accept": false, + "reason": "Flash Attention 2 support for XLMRoberta and a BART generate decoder_attention_mask bug are different models and failures." + }, + { + "left": "issue:44077", + "right": "issue:44479", + "accept": false, + "reason": "patchtsmixer post_init restriction and Qwen video-input regression are unrelated model-specific changes." + }, + { + "left": "issue:43493", + "right": "issue:43827", + "accept": false, + "reason": "SigLIP2 HF/JAX discrepancy and docs still referencing pipeline() are unrelated; one is model parity, the other documentation cleanup." + }, + { + "left": "issue:44062", + "right": "issue:45907", + "accept": false, + "reason": "AddedToken duplicate special-arg error and list_repo_files revision forwarding are unrelated API bugs." + }, + { + "left": "issue:43316", + "right": "issue:43525", + "accept": false, + "reason": "Gemma3TextConfig API discrepancy and missing pad_token_id on Llama4Config are different config-attribute problems." + }, + { + "left": "issue:43873", + "right": "issue:44871", + "accept": false, + "reason": "Quantization/offloading behavior and Gemma-3 eos_token_id inconsistency are not the same bug." + }, + { + "left": "issue:45325", + "right": "issue:46697", + "accept": false, + "reason": "Qwen2.5-VL rope indexing for still images and Swin2SR over-padding images are different image/video preprocessing bugs." + }, + { + "left": "issue:43643", + "right": "issue:44913", + "accept": false, + "reason": "trust_remote_code missing fields in AutoConfig and GPTNeoX rotary_pct not persisting on reload are distinct config-loading issues." + }, + { + "left": "issue:43749", + "right": "issue:43931", + "accept": false, + "reason": "FSDP CPU RAM efficient loading and Qwen3-VL weight-shape mismatch are separate model-loading paths." + }, + { + "left": "issue:44855", + "right": "issue:45137", + "accept": false, + "reason": "Python 3.13 TorchScript parsing failure in DebertaV2 and DeepSpeed ZeRO3 deque pop error are unrelated runtime issues." + }, + { + "left": "issue:43316", + "right": "issue:45759", + "accept": false, + "reason": "Both touch composite configs, but one is a Gemma3TextConfig API discrepancy and the other is AutoModelForCausalLM not unwrapping text_config for Qwen multimodal configs; different concrete behavior." + }, + { + "left": "issue:44987", + "right": "issue:46178", + "accept": false, + "reason": "Physical-intelligence/fast loading failure and tensorflow_text backend registration error are different backend/import issues." + }, + { + "left": "issue:44442", + "right": "issue:45803", + "accept": false, + "reason": "Tokenizer loading failure and Fuyu batch feature exception handling are not the same bug." + }, + { + "left": "issue:44190", + "right": "issue:44496", + "accept": false, + "reason": "Local dataset loading in a training script and unrecognized model/config.json loading are unrelated load-time errors." + }, + { + "left": "issue:43761", + "right": "issue:44079", + "accept": false, + "reason": "CLIPVision hidden_states regression and ModelOutput key assignment when None are different output-object bugs." + }, + { + "left": "issue:43931", + "right": "issue:45216", + "accept": false, + "reason": "Weight-shape mismatch on model load and incorrect Qwen3.5 save_pretrained output are different regressions." + }, + { + "left": "issue:46829", + "right": "issue:46932", + "accept": false, + "reason": "VideoPrism hidden_states misassignment and Gemma4 quantized embedders dtype casting are unrelated model-output vs dtype-casting issues." + }, + { + "left": "issue:43232", + "right": "issue:43388", + "accept": false, + "reason": "Generation kwargs update after sync_gpus and gather_for_metrics dropping tuple labels are distinct distributed-training 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": 26, + "review_unit_index": 36, + "review_unit_count": 61, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-505", + "nodes": [ + "issue:16998", + "issue:22355", + "issue:25251", + "issue:27957", + "issue:28282", + "issue:32090", + "issue:42617", + "issue:42907", + "issue:43295", + "issue:43316", + "issue:43404", + "issue:43441", + "issue:43452", + "issue:43550", + "issue:43575", + "issue:43606", + "issue:43646", + "issue:43701", + "issue:43716", + "issue:43742", + "issue:43761", + "issue:43825", + "issue:44077", + "issue:44162", + "issue:44188", + "issue:44242", + "issue:44464", + "issue:44496", + "issue:44534", + "issue:44610", + "issue:44792", + "issue:44871", + "issue:44913", + "issue:45584", + "issue:45632", + "issue:45698", + "issue:45803", + "issue:46050", + "issue:46170", + "issue:46234", + "issue:46291", + "issue:46365", + "issue:46369", + "issue:46421", + "issue:46424", + "issue:46506", + "issue:46821", + "issue:46894" + ], + "soft_pairs": [ + "issue:43761|issue:44188", + "issue:43701|issue:46506", + "issue:42617|issue:43575", + "issue:43575|issue:43742", + "issue:43550|issue:43701", + "issue:27957|issue:32090", + "issue:42907|issue:45698", + "issue:25251|issue:28282", + "issue:43606|issue:43701", + "issue:16998|issue:22355", + "issue:43295|issue:43825", + "issue:43295|issue:44871", + "issue:44871|issue:45584", + "issue:46234|issue:46291", + "issue:44464|issue:46421", + "issue:44077|issue:44534", + "issue:43316|issue:43716", + "issue:43316|issue:46821", + "issue:44871|issue:46050", + "issue:43646|issue:44913", + "issue:43452|issue:44496", + "issue:45803|issue:46365", + "issue:44162|issue:44792", + "issue:43404|issue:43441", + "issue:45632|issue:46424", + "issue:44242|issue:46894", + "issue:44610|issue:46369", + "issue:45803|issue:46170" + ], + "prepared_review_unit_hash": "2d33e6266108702e537718d3ce2d8a1ae0cd74894cae45eb359a4b78494afb34", + "component_budget": { + "node_count": 505, + "item_count": 505, + "soft_pair_count": 1854, + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 28, + "serialized_chars": 35855, + "estimated_input_tokens": 8964, + "estimated_eval_tokens": 18184 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is a set of mostly unrelated issues that only look superficially similar by shared terms like loading, configs, cache, generation, or performance. None of the soft pairs appear to describe the same concrete bug/change, so all soft edges should be rejected.", + "confidence": 0.96, + "canonical_issue_reason": "issue:45698 is the strongest representative issue: it has the most discussion and inbound references, and it describes a concrete, reproducible save_pretrained/from_pretrained custom-module loading failure.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45698 is also the best overall issue candidate because it is well-scoped, high-signal, and centrally about a user-facing loading regression.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43761", + "right": "issue:44188", + "accept": false, + "reason": "Both mention regressions around forward/attention behavior, but one is CLIPVisionModel hidden_states being None and the other is torch.compile branching causing diverging attention kernels; different code paths and failure modes." + }, + { + "left": "issue:43701", + "right": "issue:46506", + "accept": false, + "reason": "resume_from_checkpoint key mismatch is a checkpoint/state dict naming issue, while PPDocLayoutV3 CPU vs CUDA differences are numerical/device divergence; not the same bug." + }, + { + "left": "issue:42617", + "right": "issue:43575", + "accept": false, + "reason": "3d_parallel.py failing is a script/runtime problem, while Qwen2 tp causing OOM is a model-loading memory footprint issue; they are not the same underlying defect." + }, + { + "left": "issue:43575", + "right": "issue:43742", + "accept": false, + "reason": "Both involve loading models, but one is tensor-parallel OOM for Qwen2-57B-A14B-Instruct and the other is a key error for MobileLLM-125M; unrelated failures." + }, + { + "left": "issue:43550", + "right": "issue:43701", + "accept": false, + "reason": "torch.compile with SDPA on Bamba-9B-v2 is an execution/graphing issue, not the same as a checkpoint key mismatch on resume_from_checkpoint." + }, + { + "left": "issue:27957", + "right": "issue:32090", + "accept": false, + "reason": "XLMRoberta Flash Attention 2 support and Trainer GPU broadcast NoneType errors are different subsystems and distinct failure modes." + }, + { + "left": "issue:42907", + "right": "issue:45698", + "accept": false, + "reason": "Saving dequantized Ministral/Devstral models is a quantization-save issue, while 45698 is wrong custom module resolution after save_pretrained; different bugs." + }, + { + "left": "issue:25251", + "right": "issue:28282", + "accept": false, + "reason": "top_k changing pipeline output shape is an API/return-format issue; AutoModel missing PyTorch is an environment import error, not a duplicate." + }, + { + "left": "issue:43606", + "right": "issue:43701", + "accept": false, + "reason": "CPU offload device mismatch in bark-small is a device placement bug, not the same as checkpoint key mismatch when resuming training." + }, + { + "left": "issue:16998", + "right": "issue:22355", + "accept": false, + "reason": "model_max_length behavior for DeBERTa-V3 and missing transformers.onnx are unrelated configuration vs import/package issues." + }, + { + "left": "issue:43295", + "right": "issue:43825", + "accept": false, + "reason": "processor.tokenizer/image-to-tokenizer regression in v4.57.5 and a misleading pipeline translation error message are different regressions in different code paths." + }, + { + "left": "issue:43295", + "right": "issue:44871", + "accept": false, + "reason": "Both involve multimodal model config inconsistencies, but one is processor.tokenizer/image passing regression and the other is eos_token_id mismatch; different concrete issues." + }, + { + "left": "issue:44871", + "right": "issue:45584", + "accept": false, + "reason": "Gemma-3 eos_token_id mismatch and Whisper empty transcription after align_special_tokens are separate tokenizer/generation bugs." + }, + { + "left": "issue:46234", + "right": "issue:46291", + "accept": false, + "reason": "Missing is_falcon_mamba_ssm_available export and kernels v0.15 requiring version/revision are both import-time issues, but they involve different packages and root causes." + }, + { + "left": "issue:44464", + "right": "issue:46421", + "accept": false, + "reason": "Chunked generation inconsistency with compiled forward and cache lazy initialization causing recompilation are both compile/cache-related, but not the same behavior or fix." + }, + { + "left": "issue:44077", + "right": "issue:44534", + "accept": false, + "reason": "patchtsmixer post_init allowance and non-persistent buffers filled with junk are unrelated model-initialization/config regressions." + }, + { + "left": "issue:43316", + "right": "issue:43716", + "accept": false, + "reason": "Gemma3TextConfig API discrepancy and Mistral-3 image preprocessor dtype mismatch are different model-family issues with different symptoms." + }, + { + "left": "issue:43316", + "right": "issue:46821", + "accept": false, + "reason": "Gemma3TextConfig API mismatch and TimesFM window_size AttributeError are unrelated config/API bugs." + }, + { + "left": "issue:44871", + "right": "issue:46050", + "accept": false, + "reason": "Gemma-3 eos_token_id inconsistency and sinq quantized save/load breakage are different serialization/config problems." + }, + { + "left": "issue:43646", + "right": "issue:44913", + "accept": false, + "reason": "Transformers 5.0.0 breaking custom model initialization and GPTNeoX rotary_pct not persisting are distinct config-loading regressions." + }, + { + "left": "issue:43452", + "right": "issue:44496", + "accept": false, + "reason": "gguf_file breaking AutoTokenizer/AutoModel loading and an unrecognized model/config.json error are different model-loading failures." + }, + { + "left": "issue:45803", + "right": "issue:46365", + "accept": false, + "reason": "FuyuBatchFeature swallowing KeyboardInterrupt is a code-path exception-handling bug, while Trainer import failing until datasets is uninstalled is an environment/package conflict." + }, + { + "left": "issue:44162", + "right": "issue:44792", + "accept": false, + "reason": "ESM2 being broken and janus test_model_generate_images failure are unrelated model/test failures." + }, + { + "left": "issue:43404", + "right": "issue:43441", + "accept": false, + "reason": "Mistral3 lm_head weight tying affects image-text generation behavior; Ministral-3 FlashAttention failure is an attention-kernel compatibility issue." + }, + { + "left": "issue:45632", + "right": "issue:46424", + "accept": false, + "reason": "trust_remote_code cache path collisions and max_cache_length only honored for Quantized Cache are both cache-related, but one is a path collision bug and the other is a config-option propagation bug." + }, + { + "left": "issue:44242", + "right": "issue:46894", + "accept": false, + "reason": "Load balancing loss not added when output_router_logits=False and CohereASR training-loss double-shift are both training-loss bugs, but they affect different mechanisms and fixes." + }, + { + "left": "issue:44610", + "right": "issue:46369", + "accept": false, + "reason": "OmDet-Turbo processor producing 640px inputs vs model expecting 224px is an image-size preprocessing mismatch; sam3-litetext memory overflow is a resource usage issue." + }, + { + "left": "issue:45803", + "right": "issue:46170", + "accept": false, + "reason": "Bare except swallowing errors in FuyuBatchFeature is unrelated to Windows/ROCm import errors caused by torch.distributed dependency." + } + ] + }, + "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": 26, + "review_unit_index": 37, + "review_unit_count": 61, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-505", + "nodes": [ + "issue:35141", + "issue:41950", + "issue:42490", + "issue:43065", + "issue:43352", + "issue:43388", + "issue:43475", + "issue:43504", + "issue:43525", + "issue:43531", + "issue:43540", + "issue:43572", + "issue:43650", + "issue:43673", + "issue:43688", + "issue:44190", + "issue:44514", + "issue:44534", + "issue:44617", + "issue:44734", + "issue:44743", + "issue:44792", + "issue:44855", + "issue:44869", + "issue:44913", + "issue:44964", + "issue:44977", + "issue:45072", + "issue:45127", + "issue:45200", + "issue:45245", + "issue:45406", + "issue:45542", + "issue:45584", + "issue:45588", + "issue:45600", + "issue:45632", + "issue:45698", + "issue:46082", + "issue:46178", + "issue:46202", + "issue:46234", + "issue:46617", + "issue:46640", + "issue:46697", + "issue:46832", + "issue:46858", + "issue:46932" + ], + "soft_pairs": [ + "issue:43065|issue:44617", + "issue:43540|issue:44514", + "issue:45542|issue:46234", + "issue:43650|issue:46202", + "issue:43504|issue:45698", + "issue:43572|issue:46178", + "issue:44792|issue:44977", + "issue:44792|issue:45600", + "issue:45127|issue:45200", + "issue:43388|issue:43688", + "issue:43673|issue:44743", + "issue:43352|issue:46640", + "issue:43525|issue:43531", + "issue:44534|issue:44913", + "issue:41950|issue:43475", + "issue:44190|issue:44964", + "issue:35141|issue:42490", + "issue:44734|issue:45406", + "issue:45072|issue:46932", + "issue:45632|issue:46617", + "issue:44869|issue:46832", + "issue:44855|issue:45588", + "issue:46082|issue:46697", + "issue:45245|issue:46858", + "issue:44617|issue:45584" + ], + "prepared_review_unit_hash": "19c2297a7d59ead90c136766bc48909ea63b1f7986d9052b87d864885bfc63a1", + "component_budget": { + "node_count": 505, + "item_count": 505, + "soft_pair_count": 1854, + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 25, + "serialized_chars": 35271, + "estimated_input_tokens": 8818, + "estimated_eval_tokens": 17892 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The cluster is mostly heterogeneous, with one plausible duplicate thread around `trust_remote_code` cache-path handling. Most other pairs are only loosely related by subsystem and should stay separate.", + "confidence": 0.87, + "canonical_issue_reason": "`issue:45632` is the clearest root-cause description in the only plausible duplicate thread: cache-path resolution for `trust_remote_code` models. It best represents the shared bug surface with `issue:46617`.", + "canonical_pr_reason": null, + "best_issue_reason": "`issue:45632` is the strongest global issue anchor because it is the most specific, root-cause-oriented cache-path report and has the clearest duplicate counterpart in this set.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43065", + "right": "issue:44617", + "accept": false, + "reason": "Different bugs: a dummy conv layer in Sam3PixelDecoder vs a CUDA OOM in Sam3Video." + }, + { + "left": "issue:43540", + "right": "issue:44514", + "accept": false, + "reason": "Both involve video/multimodal flows, but one is a Qwen3OmniMoe video-input validation error and the other is batched chat-template padding behavior." + }, + { + "left": "issue:45542", + "right": "issue:46234", + "accept": false, + "reason": "Unrelated import/backends problems: TensorFlow/tensorboard backend detection vs a missing `is_falcon_mamba_ssm_available` symbol." + }, + { + "left": "issue:43650", + "right": "issue:46202", + "accept": false, + "reason": "No shared underlying bug; one is a vague data request and the other is a Gemma token roundtrip issue." + }, + { + "left": "issue:43504", + "right": "issue:45698", + "accept": false, + "reason": "Both concern model loading, but the concrete failures differ: legacy field handling vs wrong custom module selection after save/load." + }, + { + "left": "issue:43572", + "right": "issue:46178", + "accept": false, + "reason": "Different problems: a missing config field after v5 update vs an unsupported backend entry (`tensorflow_text`)." + }, + { + "left": "issue:44792", + "right": "issue:44977", + "accept": false, + "reason": "Different models and failure modes: janus image-generation test failure vs Qwen3.5 flash-attention generation bug." + }, + { + "left": "issue:44792", + "right": "issue:45600", + "accept": false, + "reason": "CI/test failure versus auto-mapping references to removed configs; not the same underlying change or bug." + }, + { + "left": "issue:45127", + "right": "issue:45200", + "accept": false, + "reason": "Both touch model setup, but one is LoRA/vocab collapse and the other is a Gemma 4 multimodal token-id defaulting issue." + }, + { + "left": "issue:43388", + "right": "issue:43688", + "accept": false, + "reason": "Different training bugs: metric gathering drops labels vs auxiliary-loss normalization." + }, + { + "left": "issue:43673", + "right": "issue:44743", + "accept": false, + "reason": "Both mention cache/state handling, but the concrete failures and code paths are different enough that they are not the same bug." + }, + { + "left": "issue:43352", + "right": "issue:46640", + "accept": false, + "reason": "Too broad to merge: one is a Nemotron-specific FA2 support error, the other is a generic flash-attention issue with no concrete shared code path." + }, + { + "left": "issue:43525", + "right": "issue:43531", + "accept": false, + "reason": "Different config/model bugs: missing `pad_token_id` in Llama4Config vs sliding-window behavior in Qwen3-MoE." + }, + { + "left": "issue:44534", + "right": "issue:44913", + "accept": false, + "reason": "Both are configuration/state related, but one is about non-persistent buffers being initialized with junk and the other about `rotary_pct` not persisting on reload." + }, + { + "left": "issue:41950", + "right": "issue:43475", + "accept": false, + "reason": "Related only by processor/vision area; one is wrong processor selection, the other is a missing output attribute in SAM 3 video." + }, + { + "left": "issue:44190", + "right": "issue:44964", + "accept": false, + "reason": "Different loading failures: a local dataset issue in a training script vs a model-load failure for Phi-4 multimodal." + }, + { + "left": "issue:35141", + "right": "issue:42490", + "accept": false, + "reason": "Unrelated: embedding reinitialization after resize vs an import warning from a trailing-slash local model path." + }, + { + "left": "issue:44734", + "right": "issue:45406", + "accept": false, + "reason": "Distinct serve/multimodal crashes: KV-cache continuation tensor indexing vs missing `_tokenizer` on Gemma4Processor." + }, + { + "left": "issue:45072", + "right": "issue:46932", + "accept": false, + "reason": "Different dtype issues in different contexts; one is a CI inference mismatch, the other is quantized Gemma4 embedder casting." + }, + { + "left": "issue:45632", + "right": "issue:46617", + "accept": true, + "reason": "Both point to the same `trust_remote_code` cache-path bug surface: a bad cache path/collision can plausibly produce the reported `FileNotFoundError`." + }, + { + "left": "issue:44869", + "right": "issue:46832", + "accept": false, + "reason": "Different model families and failures: Whisper word-timestamp decoding vs RT-DETR feature-level handling." + }, + { + "left": "issue:44855", + "right": "issue:45588", + "accept": false, + "reason": "Python 3.13 parse/import failure is unrelated to a flash-attention crash on `s_aux=None`." + }, + { + "left": "issue:46082", + "right": "issue:46697", + "accept": false, + "reason": "Unrelated model issues: LlamaConfig validation vs Swin2SR over-padding." + }, + { + "left": "issue:45245", + "right": "issue:46858", + "accept": false, + "reason": "Different low-level bugs: category-count overflow vs device-tensor slice arithmetic in static cache prefill." + }, + { + "left": "issue:44617", + "right": "issue:45584", + "accept": false, + "reason": "Different SAM3 OOM vs Whisper empty-transcription failure; only the general ML domain overlaps." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet and stays conservative overall. The only accepted soft edge, issue:45632 vs issue:46617, is plausibly the same trust_remote_code cache-path failure surface. The other rejections are cautious and supported by the issue titles, with no obvious overstatement." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 26, + "review_unit_index": 38, + "review_unit_count": 61, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-505", + "nodes": [ + "issue:25251", + "issue:27957", + "issue:29870", + "issue:35141", + "issue:39692", + "issue:41720", + "issue:43065", + "issue:43295", + "issue:43335", + "issue:43550", + "issue:43644", + "issue:43673", + "issue:43827", + "issue:43931", + "issue:43994", + "issue:44060", + "issue:44077", + "issue:44112", + "issue:44190", + "issue:44242", + "issue:44464", + "issue:44479", + "issue:44610", + "issue:44617", + "issue:44734", + "issue:44779", + "issue:44855", + "issue:44877", + "issue:44913", + "issue:44933", + "issue:45127", + "issue:45245", + "issue:45276", + "issue:45290", + "issue:45357", + "issue:45362", + "issue:45381", + "issue:45563", + "issue:45584", + "issue:45588", + "issue:45600", + "issue:46369", + "issue:46421", + "issue:46489", + "issue:46655", + "issue:46752", + "issue:46897", + "issue:47013" + ], + "soft_pairs": [ + "issue:43827|issue:43994", + "issue:29870|issue:35141", + "issue:43673|issue:46421", + "issue:43931|issue:44060", + "issue:43644|issue:44913", + "issue:44242|issue:46897", + "issue:25251|issue:27957", + "issue:45563|issue:45584", + "issue:44933|issue:45362", + "issue:43335|issue:44855", + "issue:44077|issue:44877", + "issue:45381|issue:46369", + "issue:43550|issue:44190", + "issue:41720|issue:45245", + "issue:44479|issue:44734", + "issue:43065|issue:44610", + "issue:43335|issue:44190", + "issue:45588|issue:47013", + "issue:44617|issue:46369", + "issue:39692|issue:44190", + "issue:44464|issue:45563", + "issue:45127|issue:45357", + "issue:45381|issue:46655", + "issue:43295|issue:45276", + "issue:44112|issue:45600", + "issue:44779|issue:46489", + "issue:45290|issue:46752" + ], + "prepared_review_unit_hash": "a977f79ee9dfb258c11fa6b34afd4802c1a8d9e1aeba3196b2215d1de12fc7e4", + "component_budget": { + "node_count": 505, + "item_count": 505, + "soft_pair_count": 1854, + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 27, + "serialized_chars": 35923, + "estimated_input_tokens": 8981, + "estimated_eval_tokens": 18218 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "Most pairs are thematically related but not the same underlying bug. The only strong duplicate-level match is the DeepSeek tokenizer regression pair; the rest differ in concrete failure mode, model, or code path.", + "confidence": 0.87, + "canonical_issue_reason": "issue:44779 is the broadest/root report for the DeepSeek tokenizer regression in v5, and issue:46489 reads like a narrower follow-up that explicitly calls out a gap in the prior fix.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:44779 is the best anchor because it states the regression plainly, is broader in scope, and can subsume the later DeepSeek-Coder v1 follow-up.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43827", + "right": "issue:43994", + "accept": false, + "reason": "Both mention pipelines, but one is a docs/API-removal problem and the other is a model/output correctness issue for SigLIP2; different bugs." + }, + { + "left": "issue:29870", + "right": "issue:35141", + "accept": false, + "reason": "One is a BART generate() decoder mask bug, the other is a resize_token_embeddings/post_init weight reinit bug; not the same code path." + }, + { + "left": "issue:43673", + "right": "issue:46421", + "accept": false, + "reason": "Both involve chunked prefill/cache behavior, but one is missing cache state and the other is spurious recompilation from lazy initialization; adjacent, not duplicate." + }, + { + "left": "issue:43931", + "right": "issue:44060", + "accept": false, + "reason": "Weight-shape mismatch on Qwen3-VL loading is a different failure from the Qwen3-Next tied-weights warning bug." + }, + { + "left": "issue:43644", + "right": "issue:44913", + "accept": false, + "reason": "A generic non-persistent-buffer junk-fill regression is not the same as GPTNeoXConfig's rotary_pct reload defaulting." + }, + { + "left": "issue:44242", + "right": "issue:46897", + "accept": false, + "reason": "These are both training-loss issues, but they affect different models and different loss computation paths." + }, + { + "left": "issue:25251", + "right": "issue:27957", + "accept": false, + "reason": "Top_k nested-list output from pipeline is unrelated to XLMRoberta Flash Attention 2 support." + }, + { + "left": "issue:45563", + "right": "issue:45584", + "accept": false, + "reason": "A stale warning in paged generate() is unrelated to Whisper empty-transcription failure after align_special_tokens." + }, + { + "left": "issue:44933", + "right": "issue:45362", + "accept": false, + "reason": "A nonexistent image_utils import is a different issue from Qwen3.5 chat crashes." + }, + { + "left": "issue:43335", + "right": "issue:44855", + "accept": false, + "reason": "Sparse-layer config construction and Python 3.13 IndentationError on import are unrelated bugs." + }, + { + "left": "issue:44077", + "right": "issue:44877", + "accept": false, + "reason": "Optional post_init handling in patchtsmixer is not the same as strict-config loading for granite_speech." + }, + { + "left": "issue:45381", + "right": "issue:46369", + "accept": false, + "reason": "Qwen video vision_position_ids regression and Sam3-LiteText memory overflow are different code paths." + }, + { + "left": "issue:43550", + "right": "issue:44190", + "accept": false, + "reason": "torch.compile/SDPA model failure and local dataset loading failure are unrelated." + }, + { + "left": "issue:41720", + "right": "issue:45245", + "accept": false, + "reason": "Qwen3 device-mapping cudaErrorAssert has nothing in common with the 2^24 category limit error." + }, + { + "left": "issue:44479", + "right": "issue:44734", + "accept": false, + "reason": "Video-input regression across Qwen models is different from KV-cache continuation tensor indexing in transformers serve." + }, + { + "left": "issue:43065", + "right": "issue:44610", + "accept": false, + "reason": "A dummy Conv2d in Sam3PixelDecoder is unrelated to OmDet-Turbo's processor/model image-size mismatch." + }, + { + "left": "issue:43335", + "right": "issue:44190", + "accept": false, + "reason": "Config sparsity construction and local dataset loading are unrelated failures." + }, + { + "left": "issue:45588", + "right": "issue:47013", + "accept": false, + "reason": "Flash-attention sinkless-model AttributeError and xLSTM unpacking crash are different implementation bugs." + }, + { + "left": "issue:44617", + "right": "issue:46369", + "accept": false, + "reason": "Sam3Video OOM and Sam3-LiteText memory overflow are both memory-related but not the same bug." + }, + { + "left": "issue:39692", + "right": "issue:44190", + "accept": false, + "reason": "SigLIP2 docs/model mismatch is unrelated to local dataset loading in the image-classification script." + }, + { + "left": "issue:44464", + "right": "issue:45563", + "accept": false, + "reason": "Chunked generation inconsistency under compiled forward is not the same as a stale warning on num_return_sequences." + }, + { + "left": "issue:45127", + "right": "issue:45357", + "accept": false, + "reason": "LoRA merge collapse with extended vocab and Qwen3.5 visual-encoder save_pretrained regression are different model-state bugs." + }, + { + "left": "issue:45381", + "right": "issue:46655", + "accept": false, + "reason": "Qwen video vision-position regression and SmolVLM video padding calculation are different processor issues." + }, + { + "left": "issue:43295", + "right": "issue:45276", + "accept": false, + "reason": "Processor/tokenizer API regression and gemma4 resize_token_embeddings not updating outputs are different bugs." + }, + { + "left": "issue:44112", + "right": "issue:45600", + "accept": false, + "reason": "GraniteSpeech CI stale device override and removed Sam3LiteText mappings are separate CI breakages." + }, + { + "left": "issue:44779", + "right": "issue:46489", + "accept": true, + "reason": "Both report the same DeepSeek tokenizer regression in v5; 46489 explicitly describes a remaining gap in the prior fix." + }, + { + "left": "issue:45290", + "right": "issue:46752", + "accept": false, + "reason": "Assistant tool-call empty-content chat-template crash and empty-list chat-template IndexError are related only at a high level; not the same specific failure." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet, and the soft-edge calls stay conservative. The only accepted duplicate-level pair, issue:44779 and issue:46489, is well-supported by the titles and the note that 46489 is a gap in the prior fix. No other pair is overstated as the same underlying bug." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 26, + "review_unit_index": 39, + "review_unit_count": 61, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-505", + "nodes": [ + "issue:25251", + "issue:27957", + "issue:29870", + "issue:32090", + "issue:35141", + "issue:36246", + "issue:38617", + "issue:41628", + "issue:41950", + "issue:42175", + "issue:42796", + "issue:43404", + "issue:43526", + "issue:43531", + "issue:43540", + "issue:43582", + "issue:43643", + "issue:43673", + "issue:43854", + "issue:44016", + "issue:44112", + "issue:44246", + "issue:44279", + "issue:44295", + "issue:44462", + "issue:44496", + "issue:44521", + "issue:44734", + "issue:44737", + "issue:44938", + "issue:44964", + "issue:45003", + "issue:45216", + "issue:45276", + "issue:45381", + "issue:45542", + "issue:45584", + "issue:45600", + "issue:45632", + "issue:45803", + "issue:45907", + "issue:46082", + "issue:46121", + "issue:46234", + "issue:46506", + "issue:46697", + "issue:46858", + "issue:46932" + ], + "soft_pairs": [ + "issue:43540|issue:44734", + "issue:43526|issue:46697", + "issue:45276|issue:45584", + "issue:44521|issue:45381", + "issue:44737|issue:45803", + "issue:44462|issue:45907", + "issue:25251|issue:29870", + "issue:44938|issue:45542", + "issue:43854|issue:44964", + "issue:43673|issue:46858", + "issue:44246|issue:45003", + "issue:43643|issue:44496", + "issue:45632|issue:46234", + "issue:44964|issue:45600", + "issue:42175|issue:44279", + "issue:27957|issue:35141", + "issue:44112|issue:44964", + "issue:42796|issue:44521", + "issue:36246|issue:41628", + "issue:46082|issue:46121", + "issue:32090|issue:35141", + "issue:43404|issue:45216", + "issue:36246|issue:38617", + "issue:43531|issue:45600", + "issue:44295|issue:46506", + "issue:41950|issue:44016", + "issue:43582|issue:46932" + ], + "prepared_review_unit_hash": "0dd7132dc513c54371cc3b590beecc5610fc016d68419c5b7052a1b9df7c5a95", + "component_budget": { + "node_count": 505, + "item_count": 505, + "soft_pair_count": 1854, + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 27, + "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": "No soft-edge pair looks like the same underlying bug/change. The candidates span unrelated areas (generation, model loading, tokenizers, caches, CI, pipelines, and docs), so there is no clear duplicate cluster here.", + "confidence": 0.97, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43540", + "right": "issue:44734", + "accept": false, + "reason": "Different failures: Qwen3OmniMoe video-input processing vs transformers serve KV-cache continuation indexing." + }, + { + "left": "issue:43526", + "right": "issue:46697", + "accept": false, + "reason": "Different image-processor bugs on different models: label reduction vs over-padding." + }, + { + "left": "issue:45276", + "right": "issue:45584", + "accept": false, + "reason": "Unrelated model-specific embedding resize bug vs Whisper empty-transcription generation failure." + }, + { + "left": "issue:44521", + "right": "issue:45381", + "accept": false, + "reason": "Different multimodal issues with different symptoms and code paths: assistant masks vs vision_position_ids." + }, + { + "left": "issue:44737", + "right": "issue:45803", + "accept": false, + "reason": "XLNet CPU tensor allocation bug is unrelated to Fuyu's broad exception handling." + }, + { + "left": "issue:44462", + "right": "issue:45907", + "accept": false, + "reason": "Tokenizer file selection bug is unrelated to list_repo_files revision forwarding." + }, + { + "left": "issue:25251", + "right": "issue:29870", + "accept": false, + "reason": "Pipeline output nesting bug is unrelated to Bart decoder_attention_mask initialization." + }, + { + "left": "issue:44938", + "right": "issue:45542", + "accept": false, + "reason": "Python 3.14 import failure is unrelated to missing TensorFlow backend dependency handling." + }, + { + "left": "issue:43854", + "right": "issue:44964", + "accept": false, + "reason": "Different model-loading regressions affecting different models and likely different failing code paths." + }, + { + "left": "issue:43673", + "right": "issue:46858", + "accept": false, + "reason": "Both cache-related, but one is a missing cache API in v5 chunked prefill and the other is device-tensor arithmetic in static cache." + }, + { + "left": "issue:44246", + "right": "issue:45003", + "accept": false, + "reason": "Import latency complaint is not the same bug as unsafe sys.modules access." + }, + { + "left": "issue:43643", + "right": "issue:44496", + "accept": false, + "reason": "Different config-loading problems: missing fields with trust_remote_code vs missing model_type in config.json." + }, + { + "left": "issue:45632", + "right": "issue:46234", + "accept": false, + "reason": "Cache-path collision for local models is unrelated to missing import_utils availability helper." + }, + { + "left": "issue:44964", + "right": "issue:45600", + "accept": false, + "reason": "Different failures: model loading regression vs stale auto_mappings reference breaking CI." + }, + { + "left": "issue:42175", + "right": "issue:44279", + "accept": false, + "reason": "General dependency/install mismatch is too broad and does not match the specific TensorFlow-backend packaging bug." + }, + { + "left": "issue:27957", + "right": "issue:35141", + "accept": false, + "reason": "Flash Attention support issue is unrelated to embedding reinitialization after resize_token_embeddings." + }, + { + "left": "issue:44112", + "right": "issue:44964", + "accept": false, + "reason": "CI test flake in GraniteSpeech is unrelated to Phi-4 multimodal loading." + }, + { + "left": "issue:42796", + "right": "issue:44521", + "accept": false, + "reason": "Tokenizer BPE-marker handling is unrelated to multimodal chat-template assistant masks." + }, + { + "left": "issue:36246", + "right": "issue:41628", + "accept": false, + "reason": "Specific missing class in qwen2_5_vl is not the same as the top-level AutoImageProcessor import failure." + }, + { + "left": "issue:46082", + "right": "issue:46121", + "accept": false, + "reason": "Llama config head_dim validation is unrelated to rope parameter list typing." + }, + { + "left": "issue:32090", + "right": "issue:35141", + "accept": false, + "reason": "Trainer broadcast TypeError is unrelated to output embedding reinitialization after resize." + }, + { + "left": "issue:43404", + "right": "issue:45216", + "accept": false, + "reason": "Weight-tying bug in Mistral3ForConditionalGeneration is unrelated to Qwen3.5 save_pretrained checkpoint corruption." + }, + { + "left": "issue:36246", + "right": "issue:38617", + "accept": false, + "reason": "Both are import errors, but for different symbols and different modules with no shared code path shown." + }, + { + "left": "issue:43531", + "right": "issue:45600", + "accept": false, + "reason": "Qwen3-MoE sliding_window behavior is unrelated to auto_mappings CI breakage." + }, + { + "left": "issue:44295", + "right": "issue:46506", + "accept": false, + "reason": "position_ids buffer access bug is unrelated to CPU/CUDA output divergence in PPDocLayoutV3." + }, + { + "left": "issue:41950", + "right": "issue:44016", + "accept": false, + "reason": "Video-classification pipeline bug is unrelated to a notebook syntax error." + }, + { + "left": "issue:43582", + "right": "issue:46932", + "accept": false, + "reason": "Apple Silicon warmup TypeError is unrelated to Gemma4 quantized embedder dtype casting." + } + ] + }, + "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": 26, + "review_unit_index": 40, + "review_unit_count": 61, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-505", + "nodes": [ + "issue:39692", + "issue:42907", + "issue:43295", + "issue:43299", + "issue:43381", + "issue:43441", + "issue:43479", + "issue:43526", + "issue:43650", + "issue:43673", + "issue:43824", + "issue:43901", + "issue:43976", + "issue:44016", + "issue:44062", + "issue:44075", + "issue:44220", + "issue:44279", + "issue:44327", + "issue:44336", + "issue:44368", + "issue:44496", + "issue:44521", + "issue:44683", + "issue:44779", + "issue:44792", + "issue:44933", + "issue:44987", + "issue:45042", + "issue:45083", + "issue:45216", + "issue:45335", + "issue:45375", + "issue:45405", + "issue:45446", + "issue:45584", + "issue:45588", + "issue:45593", + "issue:45759", + "issue:46082", + "issue:46123", + "issue:46369", + "issue:46399", + "issue:46424", + "issue:46506", + "issue:46515", + "issue:46693", + "issue:46752" + ], + "soft_pairs": [ + "issue:43673|issue:46424", + "issue:43295|issue:45335", + "issue:43650|issue:44220", + "issue:43976|issue:44987", + "issue:45083|issue:46506", + "issue:44016|issue:44220", + "issue:44496|issue:45042", + "issue:43526|issue:43901", + "issue:44683|issue:45446", + "issue:43901|issue:44496", + "issue:42907|issue:43441", + "issue:43381|issue:45593", + "issue:43673|issue:44336", + "issue:43479|issue:45375", + "issue:44496|issue:45759", + "issue:43976|issue:44279", + "issue:44792|issue:46424", + "issue:44016|issue:44062", + "issue:43441|issue:45588", + "issue:43824|issue:44933", + "issue:44279|issue:46515", + "issue:44521|issue:46693", + "issue:45335|issue:45584", + "issue:46123|issue:46369", + "issue:44327|issue:46752", + "issue:43299|issue:45216", + "issue:44279|issue:45405", + "issue:46082|issue:46399", + "issue:43295|issue:44779", + "issue:39692|issue:43901", + "issue:44075|issue:44368" + ], + "prepared_review_unit_hash": "d2ce2975146acd803fd652eb0cb7376f94ab5597ec6c97b1130c737c48a95a54", + "component_budget": { + "node_count": 505, + "item_count": 505, + "soft_pair_count": 1854, + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 31, + "serialized_chars": 36933, + "estimated_input_tokens": 9234, + "estimated_eval_tokens": 18724 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "Most items are unrelated single issues. The only likely duplicate pair is the FlashAttention crash on sink-less models, where the Ministral-3 report appears to be a model-specific manifestation of the same `s_aux=None` code-path failure.", + "confidence": 0.66, + "canonical_issue_reason": "issue:45588 is the most concrete and general report: it names the failing code path (`integrations/flash_attention.py`), the trigger (`s_aux=None`), and the affected class of models (sink-less).", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45588 is the best representative if these are clustered as duplicates, because it describes the underlying crash mechanism rather than a single model symptom.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43673", + "right": "issue:46424", + "accept": false, + "reason": "Both mention cache behavior, but one is a GenerationMixin cache regression and the other is a cache-length config limitation; different bugs." + }, + { + "left": "issue:43295", + "right": "issue:45335", + "accept": false, + "reason": "Multimodal processor/tokenizer regression vs T5Gemma embedding resize bug; no shared code-path or symptom." + }, + { + "left": "issue:43650", + "right": "issue:44220", + "accept": false, + "reason": "Completely unrelated titles and domains." + }, + { + "left": "issue:43976", + "right": "issue:44987", + "accept": false, + "reason": "Python support/version issue vs model loading failure; not the same defect." + }, + { + "left": "issue:45083", + "right": "issue:46506", + "accept": false, + "reason": "Different models and different failures: helper output-length logic vs CPU/CUDA divergence." + }, + { + "left": "issue:44016", + "right": "issue:44220", + "accept": false, + "reason": "Notebook syntax error and audio feature extraction bug are unrelated." + }, + { + "left": "issue:44496", + "right": "issue:45042", + "accept": false, + "reason": "Unrecognized model config error vs PIL image processor torchvision requirement; different layers." + }, + { + "left": "issue:43526", + "right": "issue:43901", + "accept": false, + "reason": "Image processor label reduction bug vs docs mismatch in a pipeline API." + }, + { + "left": "issue:44683", + "right": "issue:45446", + "accept": false, + "reason": "Both involve flex_attention, but one is a compiled-kernel failure and the other is a version-check/import bug; not clearly one bug." + }, + { + "left": "issue:43901", + "right": "issue:44496", + "accept": false, + "reason": "Docs/API wording issue vs model config loading failure; unrelated." + }, + { + "left": "issue:42907", + "right": "issue:43441", + "accept": false, + "reason": "Both mention Ministral, but one is about saving dequantized weights and the other is a FlashAttention runtime crash." + }, + { + "left": "issue:43381", + "right": "issue:45593", + "accept": false, + "reason": "Gradient checkpointing eval-mode behavior vs D-FINE auxiliary-loss behavior; no overlap." + }, + { + "left": "issue:43673", + "right": "issue:44336", + "accept": false, + "reason": "Cache-related regression and terminal ANSI formatting are unrelated." + }, + { + "left": "issue:43479", + "right": "issue:45375", + "accept": false, + "reason": "Different multimodal config bugs in different model families; similar shape but not the same underlying change." + }, + { + "left": "issue:44496", + "right": "issue:45759", + "accept": false, + "reason": "One is unrecognized-model loading, the other is `from_config` unwrapping of `text_config`; not the same bug." + }, + { + "left": "issue:43976", + "right": "issue:44279", + "accept": false, + "reason": "Python version support issue vs generic dependency issue; no common code-path." + }, + { + "left": "issue:44792", + "right": "issue:46424", + "accept": false, + "reason": "Janus test failure is unrelated to cache-length config handling." + }, + { + "left": "issue:44016", + "right": "issue:44062", + "accept": false, + "reason": "Notebook syntax error vs `AddedToken` keyword collision; unrelated." + }, + { + "left": "issue:43441", + "right": "issue:45588", + "accept": true, + "reason": "Both point to the same FlashAttention failure on sink-less models; 43441 is a model-specific symptom, while 45588 names the underlying `s_aux=None` crash path." + }, + { + "left": "issue:43824", + "right": "issue:44933", + "accept": false, + "reason": "Different import failures from different modules; not duplicates." + }, + { + "left": "issue:44279", + "right": "issue:46515", + "accept": false, + "reason": "Generic dependency issue vs torch-version incompatibility; too broad and not the same bug." + }, + { + "left": "issue:44521", + "right": "issue:46693", + "accept": false, + "reason": "Chat-template mask bug vs FA2 long-context performance regression; unrelated behavior." + }, + { + "left": "issue:45335", + "right": "issue:45584", + "accept": false, + "reason": "Embedding resize bug vs Whisper empty-transcription generation failure; different code paths." + }, + { + "left": "issue:46123", + "right": "issue:46369", + "accept": false, + "reason": "Partial-batch flag logic vs SAM memory overflow; unrelated." + }, + { + "left": "issue:44327", + "right": "issue:46752", + "accept": false, + "reason": "QA span decoding top-k bug vs empty-list chat-template crash; different APIs and failure modes." + }, + { + "left": "issue:43299", + "right": "issue:45216", + "accept": false, + "reason": "Different Qwen-related regressions affecting model loading vs checkpoint saving; not the same change." + }, + { + "left": "issue:44279", + "right": "issue:45405", + "accept": false, + "reason": "Dependency issue title is too generic and does not match the PEFT version bump problem." + }, + { + "left": "issue:46082", + "right": "issue:46399", + "accept": false, + "reason": "Different model/config issues with different triggers and subsystems." + }, + { + "left": "issue:43295", + "right": "issue:44779", + "accept": false, + "reason": "Both are v5 regressions around tokenizer/processor behavior, but they affect different models and report different symptoms." + }, + { + "left": "issue:39692", + "right": "issue:43901", + "accept": false, + "reason": "Documentation examples with model/processor mismatch vs pipeline docs wording mismatch; not the same issue." + }, + { + "left": "issue:44075", + "right": "issue:44368", + "accept": false, + "reason": "SGD optimizer args vs tie_word_embeddings warning; unrelated." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The clustering summary is conservative and stays within the packet. The only accepted edge, 43441 \u2194 45588, is phrased as a likely duplicate rather than a certainty, and the other pairs are rejected with appropriately narrow reasoning. No obvious overstatement of evidence." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 26, + "review_unit_index": 41, + "review_unit_count": 61, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-505", + "nodes": [ + "issue:33290", + "issue:41950", + "issue:42796", + "issue:42994", + "issue:43299", + "issue:43317", + "issue:43381", + "issue:43441", + "issue:43479", + "issue:43525", + "issue:43577", + "issue:43650", + "issue:43653", + "issue:43854", + "issue:43976", + "issue:44016", + "issue:44242", + "issue:44291", + "issue:44336", + "issue:44351", + "issue:44462", + "issue:44466", + "issue:44683", + "issue:44779", + "issue:45020", + "issue:45081", + "issue:45103", + "issue:45200", + "issue:45216", + "issue:45276", + "issue:45278", + "issue:45357", + "issue:45372", + "issue:45405", + "issue:45542", + "issue:45588", + "issue:45593", + "issue:45656", + "issue:45812", + "issue:45823", + "issue:45834", + "issue:45859", + "issue:45907", + "issue:46082", + "issue:46209", + "issue:46512", + "issue:46851", + "issue:47000" + ], + "soft_pairs": [ + "issue:42796|issue:43650", + "issue:43577|issue:45216", + "issue:44351|issue:45278", + "issue:43976|issue:45405", + "issue:46082|issue:46512", + "issue:41950|issue:45020", + "issue:44779|issue:45812", + "issue:43299|issue:45357", + "issue:44466|issue:45276", + "issue:43381|issue:47000", + "issue:43299|issue:43854", + "issue:43653|issue:45593", + "issue:44242|issue:45200", + "issue:43650|issue:46209", + "issue:41950|issue:45823", + "issue:43479|issue:43525", + "issue:44462|issue:45081", + "issue:44016|issue:44291", + "issue:44683|issue:45542", + "issue:45588|issue:47000", + "issue:45081|issue:45907", + "issue:43441|issue:45834", + "issue:33290|issue:45656", + "issue:42994|issue:43317", + "issue:45103|issue:45372", + "issue:44336|issue:46851", + "issue:45859|issue:46851" + ], + "prepared_review_unit_hash": "5bc45efc22a8b619ab3f86ea078cc4caae468e0808410a2316e0ba8d49527909", + "component_budget": { + "node_count": 505, + "item_count": 505, + "soft_pair_count": 1854, + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 27, + "serialized_chars": 35856, + "estimated_input_tokens": 8964, + "estimated_eval_tokens": 18184 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This set is mostly heterogeneous; the soft pairs are loose thematic similarities, not true duplicates. No PRs are present, and I would not merge any of the proposed pairs.", + "confidence": 0.13, + "canonical_issue_reason": "Issue 45278 is the broadest user-facing regression report in the set and the closest thing to an umbrella issue, but the overall cluster is too mixed to treat as one duplicate family.", + "canonical_pr_reason": null, + "best_issue_reason": "45278 is the best representative issue because it captures a wide upgrade regression with multiple import failures, making it more suitable as a cluster anchor than the narrower, model-specific reports.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:42796", + "right": "issue:43650", + "accept": false, + "reason": "Different problems: tokenizer BPE-marker handling vs a vague data request." + }, + { + "left": "issue:43577", + "right": "issue:45216", + "accept": false, + "reason": "Both involve dtype/serialization regressions, but they affect different models and different failure modes." + }, + { + "left": "issue:44351", + "right": "issue:45278", + "accept": false, + "reason": "Related to import breakage, but one is a specific missing symbol and the other is a broad upgrade regression; not clearly the same bug." + }, + { + "left": "issue:43976", + "right": "issue:45405", + "accept": false, + "reason": "Python version compatibility issue vs an unreleased dependency version bump; unrelated." + }, + { + "left": "issue:46082", + "right": "issue:46512", + "accept": false, + "reason": "Different configs and symptoms: head_dim validation vs stale config dtype." + }, + { + "left": "issue:41950", + "right": "issue:45020", + "accept": false, + "reason": "Pipeline processor lookup bug vs remote_code breakage; different code paths." + }, + { + "left": "issue:44779", + "right": "issue:45812", + "accept": false, + "reason": "Both are tokenizer regressions, but for different model families and likely different root causes." + }, + { + "left": "issue:43299", + "right": "issue:45357", + "accept": false, + "reason": "Loading failure for Qwen3VL MoE models vs save_pretrained key serialization for Qwen3.5; not the same issue." + }, + { + "left": "issue:44466", + "right": "issue:45276", + "accept": false, + "reason": "Serialization of tied weights vs resize_token_embeddings not updating Gemma4 embeddings; different behaviors." + }, + { + "left": "issue:43381", + "right": "issue:47000", + "accept": false, + "reason": "Gradient checkpointing eval-mode restriction vs RTDetrLoss infeasible cost matrix; unrelated." + }, + { + "left": "issue:43299", + "right": "issue:43854", + "accept": false, + "reason": "Different model-loading regressions with different failure points." + }, + { + "left": "issue:43653", + "right": "issue:45593", + "accept": false, + "reason": "Tokenizer special-token registration bug vs missing auxiliary losses in D-FINE." + }, + { + "left": "issue:44242", + "right": "issue:45200", + "accept": false, + "reason": "Router-load-balancing loss bug vs Gemma 4 text-only token type defaults; different subsystems." + }, + { + "left": "issue:43650", + "right": "issue:46209", + "accept": false, + "reason": "Generic data request vs FP8 inference failure; no overlap." + }, + { + "left": "issue:41950", + "right": "issue:45823", + "accept": false, + "reason": "Processor selection bug vs Gemma4 device-mapping mismatch; different failure modes." + }, + { + "left": "issue:43479", + "right": "issue:43525", + "accept": false, + "reason": "Different config initialization bugs in different model families." + }, + { + "left": "issue:44462", + "right": "issue:45081", + "accept": false, + "reason": "AutoTokenizer ignoring tokenizer.json vs Mistral regex patch crash; distinct tokenizer-loading issues." + }, + { + "left": "issue:44016", + "right": "issue:44291", + "accept": false, + "reason": "Notebook syntax error vs init_empty_weights argument mismatch; unrelated." + }, + { + "left": "issue:44683", + "right": "issue:45542", + "accept": false, + "reason": "Compiled flex_attention torch-version breakage vs TensorBoard/TF backend packaging issue." + }, + { + "left": "issue:45588", + "right": "issue:47000", + "accept": false, + "reason": "FlashAttention sink-less-model crash vs RTDetrLoss infeasible matrix; different code paths." + }, + { + "left": "issue:45081", + "right": "issue:45907", + "accept": false, + "reason": "Tokenizer regex crash vs repo file listing not forwarding revision; unrelated." + }, + { + "left": "issue:43441", + "right": "issue:45834", + "accept": false, + "reason": "FlashAttention incompatibility for Ministral-3 vs Kosmos2.5 OCR long-input index error." + }, + { + "left": "issue:33290", + "right": "issue:45656", + "accept": false, + "reason": "Both mention DeepSpeed/optimizer behavior, but one is OOM with Adafactor and the other is duplicate optimizer.step calls." + }, + { + "left": "issue:42994", + "right": "issue:43317", + "accept": false, + "reason": "Quantized model saving failure vs dequantized model loading/offload failure; opposite sides of the pipeline." + }, + { + "left": "issue:45103", + "right": "issue:45372", + "accept": false, + "reason": "Auto-docstring annotation handling bug vs Gemma 4 processor import failure; unrelated." + }, + { + "left": "issue:44336", + "right": "issue:46851", + "accept": false, + "reason": "ANSI output formatting issue vs Qwen3.5 variable-length position_ids bug." + }, + { + "left": "issue:45859", + "right": "issue:46851", + "accept": false, + "reason": "CPU-offload incompatibility in rotary embedding vs variable-length position_ids behavior; not 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": 1, + "component_count": 26, + "review_unit_index": 42, + "review_unit_count": 61, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-505", + "nodes": [ + "issue:41628", + "issue:42175", + "issue:42796", + "issue:43441", + "issue:43525", + "issue:43643", + "issue:43673", + "issue:43723", + "issue:43761", + "issue:43784", + "issue:43824", + "issue:43881", + "issue:43901", + "issue:43976", + "issue:44016", + "issue:44162", + "issue:44186", + "issue:44188", + "issue:44246", + "issue:44261", + "issue:44263", + "issue:44297", + "issue:44336", + "issue:44373", + "issue:44496", + "issue:44589", + "issue:44623", + "issue:44671", + "issue:44749", + "issue:44779", + "issue:45083", + "issue:45325", + "issue:45362", + "issue:45397", + "issue:45542", + "issue:45561", + "issue:45563", + "issue:45600", + "issue:45712", + "issue:45800", + "issue:45823", + "issue:46291", + "issue:46399", + "issue:46655", + "issue:46726", + "issue:46851", + "issue:47000" + ], + "soft_pairs": [ + "issue:44623|issue:44749", + "issue:44261|issue:44263", + "issue:45600|issue:45712", + "issue:42175|issue:43976", + "issue:44671|issue:44779", + "issue:44623|issue:45397", + "issue:43901|issue:44246", + "issue:43673|issue:43761", + "issue:43723|issue:44162", + "issue:43525|issue:43881", + "issue:44186|issue:47000", + "issue:44016|issue:45823", + "issue:43824|issue:43976", + "issue:42796|issue:43441", + "issue:45542|issue:45800", + "issue:43525|issue:43643", + "issue:44188|issue:44336", + "issue:44589|issue:45362", + "issue:45325|issue:46851", + "issue:46291|issue:46399", + "issue:44336|issue:44373", + "issue:46655|issue:46726", + "issue:44246|issue:44496", + "issue:41628|issue:43976", + "issue:43784|issue:45542", + "issue:45083|issue:46851", + "issue:44297|issue:45397", + "issue:46399|issue:46851", + "issue:45561|issue:45563" + ], + "prepared_review_unit_hash": "29edde0f7f8ca07fced4c5ed4783bc6a29866842aae6f5bea829108efcee7dea", + "component_budget": { + "node_count": 505, + "item_count": 505, + "soft_pair_count": 1854, + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 29, + "serialized_chars": 35759, + "estimated_input_tokens": 8940, + "estimated_eval_tokens": 18136 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is heterogeneous: the items span unrelated import, packaging, tokenizer, model-specific, and docs/runtime issues. I did not find any soft pair that clearly looks like the same underlying bug/change, so no soft edges are accepted.", + "confidence": 0.88, + "canonical_issue_reason": "issue:43824 is the strongest single representative: a high-engagement, user-facing import failure with a clear repro and broad impact, making it the most useful anchor in an otherwise mixed cluster.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43824 is the best issue candidate for triage because it is concrete, high-signal, and has stronger discussion/inbound references than most others here.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44623", + "right": "issue:44749", + "accept": false, + "reason": "Different bugs: missing save_pretrained files vs a tokenizer/data-filtering slowdown. No shared code-path." + }, + { + "left": "issue:44261", + "right": "issue:44263", + "accept": false, + "reason": "Unrelated model internals: RMS norm config precision issue vs torch.split return handling." + }, + { + "left": "issue:45600", + "right": "issue:45712", + "accept": false, + "reason": "Both are stale-artifact/CI-adjacent, but they target different files and failure modes (auto_mappings vs dummy_pt_objects)." + }, + { + "left": "issue:42175", + "right": "issue:43976", + "accept": false, + "reason": "Different packaging problems: missing TensorFlow in torch extra vs Python-version compatibility." + }, + { + "left": "issue:44671", + "right": "issue:44779", + "accept": false, + "reason": "Different regressions: CamemBERT MLM outputs vs Deepseek tokenizer behavior." + }, + { + "left": "issue:44623", + "right": "issue:45397", + "accept": false, + "reason": "Different model/load paths: processor save_pretrained artifact loss vs Gemma-4 ZeRO3 from_pretrained." + }, + { + "left": "issue:43901", + "right": "issue:44246", + "accept": false, + "reason": "Docs mismatch for a pipeline option vs intermittent slow import. Not the same issue." + }, + { + "left": "issue:43673", + "right": "issue:43761", + "accept": false, + "reason": "Different execution paths: generation cache during chunked prefill vs CLIPVisionModel hidden_states output." + }, + { + "left": "issue:43723", + "right": "issue:44162", + "accept": false, + "reason": "Different failures: AutoTokenizer loading regression vs ESM2 model breakage." + }, + { + "left": "issue:43525", + "right": "issue:43881", + "accept": false, + "reason": "Both are model-loading failures, but they involve different models and different missing fields/errors." + }, + { + "left": "issue:44186", + "right": "issue:47000", + "accept": false, + "reason": "Different subsystems: tokenizer batching/NER crash vs RTDetrLoss infeasible cost matrix." + }, + { + "left": "issue:44016", + "right": "issue:45823", + "accept": false, + "reason": "Notebook syntax error and Gemma4 device-mismatch are unrelated." + }, + { + "left": "issue:43824", + "right": "issue:43976", + "accept": false, + "reason": "ImportError for a specific class vs Python version support; unrelated root causes." + }, + { + "left": "issue:42796", + "right": "issue:43441", + "accept": false, + "reason": "Same model family, but tokenizer BPE handling and FlashAttention crash are distinct bugs." + }, + { + "left": "issue:45542", + "right": "issue:45800", + "accept": false, + "reason": "Different installation/runtime issues: TensorFlow backend exposure vs torch/transformers incompatibility." + }, + { + "left": "issue:43525", + "right": "issue:43643", + "accept": false, + "reason": "Both mention missing config fields, but the model/config contexts and concrete failures differ." + }, + { + "left": "issue:44188", + "right": "issue:44336", + "accept": false, + "reason": "Attention-kernel divergence on torch.compile vs ANSI code emission in loading_report are unrelated." + }, + { + "left": "issue:44589", + "right": "issue:45362", + "accept": false, + "reason": "Storage-object deserialization error vs Qwen3.5 chat loading crash; different code paths." + }, + { + "left": "issue:45325", + "right": "issue:46851", + "accept": false, + "reason": "Both involve position_ids/rope in Qwen models, but they are different models and different behaviors." + }, + { + "left": "issue:46291", + "right": "issue:46399", + "accept": false, + "reason": "Both are kernels-related, but one is an import-time dependency requirement and the other is a registration/decorator crash." + }, + { + "left": "issue:44336", + "right": "issue:44373", + "accept": false, + "reason": "Loading-report ANSI formatting vs a wrong docstring; no shared bug." + }, + { + "left": "issue:46655", + "right": "issue:46726", + "accept": false, + "reason": "Different processor bugs: video padding calculation vs missing vision-backend handling on AutoProcessor." + }, + { + "left": "issue:44246", + "right": "issue:44496", + "accept": false, + "reason": "Intermittent slow import vs unrecognized model/config missing model_type. Not the same issue." + }, + { + "left": "issue:41628", + "right": "issue:43976", + "accept": false, + "reason": "ImportError for AutoImageProcessor vs Python version compatibility; unrelated." + }, + { + "left": "issue:43784", + "right": "issue:45542", + "accept": false, + "reason": "Importing sentence-transformers breaks for a different missing symbol than the TensorFlow backend/package issue." + }, + { + "left": "issue:45083", + "right": "issue:46851", + "accept": false, + "reason": "Both touch position-related helpers, but they are different models and different failure modes." + }, + { + "left": "issue:44297", + "right": "issue:45397", + "accept": false, + "reason": "Tokenizer metadata saved incorrectly vs Gemma-4 loading under ZeRO3; unrelated." + }, + { + "left": "issue:46399", + "right": "issue:46851", + "accept": false, + "reason": "Different Qwen3.5 failures: kernels registration crash vs variable-length position_ids behavior." + }, + { + "left": "issue:45561", + "right": "issue:45563", + "accept": false, + "reason": "pytest-xdist file race vs stale warning for num_return_sequences; no shared underlying bug." + } + ] + }, + "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": 26, + "review_unit_index": 43, + "review_unit_count": 61, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-505", + "nodes": [ + "issue:36246", + "issue:41950", + "issue:42491", + "issue:42757", + "issue:42886", + "issue:43257", + "issue:43262", + "issue:43299", + "issue:43526", + "issue:43531", + "issue:43540", + "issue:43550", + "issue:43901", + "issue:44220", + "issue:44242", + "issue:44246", + "issue:44295", + "issue:44336", + "issue:44351", + "issue:44451", + "issue:44479", + "issue:44485", + "issue:44496", + "issue:44779", + "issue:44871", + "issue:45003", + "issue:45072", + "issue:45103", + "issue:45137", + "issue:45292", + "issue:45362", + "issue:45375", + "issue:45397", + "issue:45561", + "issue:45588", + "issue:45593", + "issue:45759", + "issue:46132", + "issue:46399", + "issue:46506", + "issue:46614", + "issue:46655", + "issue:46665", + "issue:46752", + "issue:46821", + "issue:46829", + "issue:46851" + ], + "soft_pairs": [ + "issue:43526|issue:45003", + "issue:45103|issue:46399", + "issue:45072|issue:46829", + "issue:43540|issue:46752", + "issue:42491|issue:43299", + "issue:43531|issue:45362", + "issue:46614|issue:46665", + "issue:44246|issue:44871", + "issue:45588|issue:46851", + "issue:44295|issue:45397", + "issue:45003|issue:46821", + "issue:44242|issue:45593", + "issue:44336|issue:44485", + "issue:42886|issue:46132", + "issue:43550|issue:46506", + "issue:44779|issue:45292", + "issue:44451|issue:46752", + "issue:36246|issue:42757", + "issue:36246|issue:44351", + "issue:43901|issue:45003", + "issue:43257|issue:45137", + "issue:41950|issue:44220", + "issue:43540|issue:46655", + "issue:44496|issue:45003", + "issue:45003|issue:45561", + "issue:43262|issue:44246", + "issue:45375|issue:45759", + "issue:44479|issue:46655" + ], + "prepared_review_unit_hash": "9bdfac12c1b123168b69fdbfd93d93d05f39d5268682d98200b6dfc18e445a89", + "component_budget": { + "node_count": 505, + "item_count": 505, + "soft_pair_count": 1854, + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 28, + "serialized_chars": 35569, + "estimated_input_tokens": 8893, + "estimated_eval_tokens": 18042 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The proposed pairs are mostly loose topical matches rather than true duplicates. I did not find any soft edge where the two issues clearly describe the same underlying bug or code path.", + "confidence": 0.91, + "canonical_issue_reason": "issue:44479 is the strongest canonical anchor in this set because it describes a concrete, version-specific regression that affects multiple related Qwen video models and is well scoped.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:44479 is the best standalone issue here: it is specific, reproducible, and broad enough to represent a meaningful regression affecting several models, unlike the other mostly narrow or unrelated reports.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43526", + "right": "issue:45003", + "accept": false, + "reason": "Different bugs: BeitImageProcessorFast label reduction vs unsafe sys.modules access in modeling_utils." + }, + { + "left": "issue:45103", + "right": "issue:46399", + "accept": false, + "reason": "Both mention internal framework behavior, but one is auto-docstring parsing with future annotations and the other is Qwen3.5 kernel registration; different code paths." + }, + { + "left": "issue:45072", + "right": "issue:46829", + "accept": false, + "reason": "Unrelated failures: dtype mismatch in model inference versus wrong hidden_states output in a video classifier." + }, + { + "left": "issue:43540", + "right": "issue:46752", + "accept": false, + "reason": "Different areas: Qwen3OmniMoe video-input processing versus empty-list handling in apply_chat_template." + }, + { + "left": "issue:42491", + "right": "issue:43299", + "accept": false, + "reason": "Both involve Qwen3/MoE loading regressions, but one is LoRA/accelerate/deepspeed compatibility and the other is direct Qwen3VL MoE loading; not the same bug." + }, + { + "left": "issue:43531", + "right": "issue:45362", + "accept": false, + "reason": "Qwen3-MoE sliding_window behavior and Qwen3.5 chat crashes are separate model/runtime issues." + }, + { + "left": "issue:46614", + "right": "issue:46665", + "accept": false, + "reason": "Stale phonemizer language state is unrelated to from_pretrained behavior inside Python classes." + }, + { + "left": "issue:44246", + "right": "issue:44871", + "accept": false, + "reason": "Different failures: missing auxiliary loss when router logits are off versus inconsistent eos_token_id config in Gemma-3." + }, + { + "left": "issue:45588", + "right": "issue:46851", + "accept": false, + "reason": "Both touch Qwen3.5, but one is flash_attention sink handling and the other is variable-length position_ids; not the same code path." + }, + { + "left": "issue:44295", + "right": "issue:45397", + "accept": false, + "reason": "A position_ids buffer access error and a gemma-4 ZeRO3 from_pretrained issue are unrelated." + }, + { + "left": "issue:45003", + "right": "issue:46821", + "accept": false, + "reason": "Different subsystems entirely: sys.modules access in modeling_utils versus TimesFM window_size handling." + }, + { + "left": "issue:44242", + "right": "issue:45593", + "accept": false, + "reason": "Both concern losses being skipped, but they are different models and different conditions; not one concrete defect." + }, + { + "left": "issue:44336", + "right": "issue:44485", + "accept": false, + "reason": "ANSI color output in loading_report is unrelated to GLM-5 RoPE implementation." + }, + { + "left": "issue:42886", + "right": "issue:46132", + "accept": false, + "reason": "Offline cache loading and AttentionInterface.register behavior changes are unrelated." + }, + { + "left": "issue:43550", + "right": "issue:46506", + "accept": false, + "reason": "Different failures: torch.compile+SDPA on Bamba versus CPU/CUDA result divergence in PPDocLayoutV3." + }, + { + "left": "issue:44779", + "right": "issue:45292", + "accept": false, + "reason": "Tokenizer output regressions and resize_token_embeddings not updating output embeddings are separate issues." + }, + { + "left": "issue:44451", + "right": "issue:46752", + "accept": false, + "reason": "ScandiBERT loading failure is unrelated to empty-list chat template handling." + }, + { + "left": "issue:36246", + "right": "issue:42757", + "accept": false, + "reason": "Both are import errors, but for different missing symbols in different packages; not the same underlying bug." + }, + { + "left": "issue:36246", + "right": "issue:44351", + "accept": false, + "reason": "Missing Qwen2_5_VLImageProcessor export and missing HybridCache export are distinct symbol/import problems." + }, + { + "left": "issue:43901", + "right": "issue:45003", + "accept": false, + "reason": "Docs mismatch for TextClassificationPipeline is unrelated to sys.modules access safety." + }, + { + "left": "issue:43257", + "right": "issue:45137", + "accept": false, + "reason": "Both involve DeepSpeed, but one is Qwen3 MoE weight conversion and the other is a ZeRO3 deque error; different failures." + }, + { + "left": "issue:41950", + "right": "issue:44220", + "accept": false, + "reason": "Video-classification processor selection and audio fbank extraction are unrelated." + }, + { + "left": "issue:43540", + "right": "issue:46655", + "accept": false, + "reason": "Both are video-input bugs, but they affect different models and different processing logic." + }, + { + "left": "issue:44496", + "right": "issue:45003", + "accept": false, + "reason": "Unrecognized-model config loading and unsafe sys.modules access are unrelated." + }, + { + "left": "issue:45003", + "right": "issue:45561", + "accept": false, + "reason": "Generic import-safety issue versus pytest-xdist file race in testing utils." + }, + { + "left": "issue:43262", + "right": "issue:44246", + "accept": false, + "reason": "Audio chat-template sampling-rate default and expert load-balancing loss omission are unrelated." + }, + { + "left": "issue:45375", + "right": "issue:45759", + "accept": false, + "reason": "Both concern Qwen multimodal configs, but one is a missing field dropped by strict config handling and the other is text_config unwrapping in from_config; different bugs." + }, + { + "left": "issue:44479", + "right": "issue:46655", + "accept": false, + "reason": "Both are video regressions, but the affected models and failure mechanisms are different; 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": 1, + "component_count": 26, + "review_unit_index": 44, + "review_unit_count": 61, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-505", + "nodes": [ + "issue:29870", + "issue:30333", + "issue:36246", + "issue:42199", + "issue:42886", + "issue:43295", + "issue:43352", + "issue:43475", + "issue:43526", + "issue:43540", + "issue:43596", + "issue:43653", + "issue:44038", + "issue:44060", + "issue:44077", + "issue:44162", + "issue:44242", + "issue:44295", + "issue:44368", + "issue:44466", + "issue:44479", + "issue:44623", + "issue:44749", + "issue:44857", + "issue:44877", + "issue:44977", + "issue:44987", + "issue:45137", + "issue:45200", + "issue:45229", + "issue:45250", + "issue:45292", + "issue:45335", + "issue:45375", + "issue:45412", + "issue:45459", + "issue:45600", + "issue:45859", + "issue:46123", + "issue:46170", + "issue:46369", + "issue:46399", + "issue:46506", + "issue:46726", + "issue:46851", + "issue:46894", + "issue:46897", + "issue:47000" + ], + "soft_pairs": [ + "issue:44162|issue:45200", + "issue:45859|issue:46399", + "issue:45292|issue:47000", + "issue:44295|issue:44749", + "issue:45137|issue:45600", + "issue:43475|issue:46369", + "issue:43295|issue:45292", + "issue:44466|issue:45335", + "issue:36246|issue:43596", + "issue:44242|issue:44857", + "issue:44877|issue:45375", + "issue:44623|issue:46726", + "issue:43526|issue:46897", + "issue:44987|issue:45459", + "issue:44038|issue:46506", + "issue:29870|issue:30333", + "issue:44479|issue:46170", + "issue:43526|issue:46894", + "issue:44060|issue:44077", + "issue:42199|issue:46851", + "issue:44368|issue:44977", + "issue:45412|issue:46123", + "issue:43653|issue:44162", + "issue:43352|issue:45250", + "issue:42886|issue:45229", + "issue:44877|issue:46170", + "issue:43475|issue:43540" + ], + "prepared_review_unit_hash": "fc7648eb25efd8c97ba9f69664eaa4fb56ddcbadcb7dfebdc00bba28861c0ca7", + "component_budget": { + "node_count": 505, + "item_count": 505, + "soft_pair_count": 1854, + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 27, + "serialized_chars": 36034, + "estimated_input_tokens": 9009, + "estimated_eval_tokens": 18274 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "All soft edges look like false positives: the items span unrelated bugs across model families, tokenizer/config loading, and runtime errors. No pair is close enough to be the same concrete underlying issue, so none should be merged.", + "confidence": 0.94, + "canonical_issue_reason": "issue:45292 is the most general, self-contained API-level bug in the set and the best representative anchor if one issue must be chosen.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45292 is the cleanest model-agnostic report and the best dedupe anchor, even though it is not actually a duplicate of the others.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44162", + "right": "issue:45200", + "accept": false, + "reason": "Different models and failures: ESM2 breakage vs Gemma4 mm_token_type_ids defaults." + }, + { + "left": "issue:45859", + "right": "issue:46399", + "accept": false, + "reason": "Both involve Qwen3.5, but one is CPU-offload compatibility and the other is a registration/decorator crash; different bugs." + }, + { + "left": "issue:45292", + "right": "issue:47000", + "accept": false, + "reason": "resize_token_embeddings vs RT-DETR cost matrix infeasible are unrelated subsystems and symptoms." + }, + { + "left": "issue:44295", + "right": "issue:44749", + "accept": false, + "reason": "Position_ids buffer access error is unrelated to the reported data-processing slowdown/regression." + }, + { + "left": "issue:45137", + "right": "issue:45600", + "accept": false, + "reason": "DeepSpeed ZeRO3 deque underflow and Sam3LiteText auto_mappings CI breakage are different code paths." + }, + { + "left": "issue:43475", + "right": "issue:46369", + "accept": false, + "reason": "Both mention SAM 3, but one is a missing attribute in video output and the other is a memory overflow in litetext." + }, + { + "left": "issue:43295", + "right": "issue:45292", + "accept": false, + "reason": "Custom processor/tokenizer regression is not the same bug as resize_token_embeddings not updating output embeddings." + }, + { + "left": "issue:44466", + "right": "issue:45335", + "accept": false, + "reason": "Serialization inconsistency with tied weights is a different issue from resize_token_embeddings not affecting decoder.embed_tokens." + }, + { + "left": "issue:36246", + "right": "issue:43596", + "accept": false, + "reason": "Qwen2_5_VL import failure and BertModel/deepspeed IndexError are unrelated." + }, + { + "left": "issue:44242", + "right": "issue:44857", + "accept": false, + "reason": "MoE load-balancing loss gating is unrelated to LwDetrImageLoss AMP/CUDA crashes." + }, + { + "left": "issue:44877", + "right": "issue:45375", + "accept": false, + "reason": "Both touch config strictness, but they involve different models and different missing-field behaviors; not the same bug." + }, + { + "left": "issue:44623", + "right": "issue:46726", + "accept": false, + "reason": "Missing files on save_pretrained is different from AutoProcessor backend-missing handling." + }, + { + "left": "issue:43526", + "right": "issue:46897", + "accept": false, + "reason": "BEiT Fast label reduction bug and Florence2 double-shift training-loss bug are unrelated." + }, + { + "left": "issue:44987", + "right": "issue:45459", + "accept": false, + "reason": "Model loading regression and protobuf-exception masking are different failure modes." + }, + { + "left": "issue:44038", + "right": "issue:46506", + "accept": false, + "reason": "Qwen3-VL-Moe regression and PPDocLayoutV3 CPU/CUDA divergence are unrelated." + }, + { + "left": "issue:29870", + "right": "issue:30333", + "accept": false, + "reason": "Bart generate() UnboundLocalError is unrelated to MLFlow job-status reporting." + }, + { + "left": "issue:44479", + "right": "issue:46170", + "accept": false, + "reason": "Video-input regression for Qwen models and Windows/ROCm torch.distributed import failure are different bugs." + }, + { + "left": "issue:43526", + "right": "issue:46894", + "accept": false, + "reason": "BEiT Fast label handling and CohereASR double-shift loss are unrelated." + }, + { + "left": "issue:44060", + "right": "issue:44077", + "accept": false, + "reason": "Tied-weights warning involving dt_bias and a patchtsmixer post_init restriction are not the same underlying issue." + }, + { + "left": "issue:42199", + "right": "issue:46851", + "accept": false, + "reason": "DETR cardinality handling and Qwen3.5 variable-length position_ids are unrelated." + }, + { + "left": "issue:44368", + "right": "issue:44977", + "accept": false, + "reason": "Tie_word_embeddings warning during LoRA fine-tuning is different from flash-attention generation failure." + }, + { + "left": "issue:45412", + "right": "issue:46123", + "accept": false, + "reason": "RT-DETR memory lifecycle bug is unrelated to MaskGenerationPipeline partial-batch handling." + }, + { + "left": "issue:43653", + "right": "issue:44162", + "accept": false, + "reason": "BigBirdTokenizer special-token registration and ESM2 breakage are unrelated." + }, + { + "left": "issue:43352", + "right": "issue:45250", + "accept": false, + "reason": "Model-specific Flash Attention 2 support gap is not the same as the generic flash-attention issue." + }, + { + "left": "issue:42886", + "right": "issue:45229", + "accept": false, + "reason": "Offline tokenizer cache loading and Gemma4 multi-GPU OOM are unrelated." + }, + { + "left": "issue:44877", + "right": "issue:46170", + "accept": false, + "reason": "Strict config loading and Windows/ROCm dependency import failure are different code paths." + }, + { + "left": "issue:43475", + "right": "issue:43540", + "accept": false, + "reason": "SAM 3 Video attribute error and Qwen3OmniMoe video-input ValueError are different 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": 26, + "review_unit_index": 45, + "review_unit_count": 61, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-505", + "nodes": [ + "issue:28282", + "issue:30990", + "issue:42994", + "issue:43065", + "issue:43408", + "issue:43441", + "issue:43475", + "issue:43482", + "issue:43643", + "issue:43646", + "issue:43701", + "issue:43723", + "issue:43906", + "issue:43931", + "issue:43994", + "issue:44162", + "issue:44188", + "issue:44297", + "issue:44361", + "issue:44368", + "issue:44483", + "issue:44623", + "issue:44749", + "issue:44871", + "issue:44995", + "issue:45103", + "issue:45200", + "issue:45375", + "issue:45397", + "issue:45478", + "issue:45563", + "issue:45588", + "issue:45656", + "issue:45685", + "issue:45987", + "issue:46082", + "issue:46123", + "issue:46132", + "issue:46170", + "issue:46202", + "issue:46421", + "issue:46431", + "issue:46710", + "issue:46726", + "issue:46752", + "issue:46821", + "issue:46832", + "issue:46858" + ], + "soft_pairs": [ + "issue:46170|issue:46832", + "issue:45397|issue:45656", + "issue:43065|issue:43408", + "issue:43723|issue:46710", + "issue:44368|issue:45563", + "issue:44995|issue:45200", + "issue:43931|issue:44368", + "issue:44188|issue:46202", + "issue:45987|issue:46082", + "issue:44483|issue:46202", + "issue:43906|issue:46202", + "issue:44995|issue:45588", + "issue:44188|issue:46421", + "issue:45685|issue:46431", + "issue:45103|issue:46858", + "issue:43065|issue:46123", + "issue:44361|issue:44749", + "issue:44297|issue:44623", + "issue:43994|issue:46726", + "issue:44749|issue:45656", + "issue:43701|issue:44749", + "issue:43646|issue:44162", + "issue:44871|issue:46132", + "issue:44623|issue:45656", + "issue:43441|issue:46752", + "issue:43475|issue:43643", + "issue:42994|issue:43482", + "issue:45375|issue:45478", + "issue:43475|issue:46821", + "issue:28282|issue:30990" + ], + "prepared_review_unit_hash": "3c84c15928ab1d5198adfe7cc031e8a357a709cd47029d44ee3c3d718c67d0b8", + "component_budget": { + "node_count": 505, + "item_count": 505, + "soft_pair_count": 1854, + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 30, + "serialized_chars": 36804, + "estimated_input_tokens": 9201, + "estimated_eval_tokens": 18658 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "These issues span many unrelated regressions across model loading, caching, tokenization, pipelines, and backend-specific failures. None of the proposed soft pairs look like the same concrete bug/change under a strict duplicate triage standard.", + "confidence": 0.97, + "canonical_issue_reason": "issue:46858 is the most concrete and actionable report here: it names the exact failing path (_prefill), the broken contract (cache.get_seq_length()), and the affected scenario (static cache + torch.compile on non-CUDA backends).", + "canonical_pr_reason": null, + "best_issue_reason": "issue:46858 has the clearest symptom-to-code-path mapping and is the strongest candidate for a canonical bug report among the issues shown.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:46170", + "right": "issue:46832", + "accept": false, + "reason": "Different failures: Windows/ROCm import/type dependency issue vs RTDetrModel forward feature-level mismatch." + }, + { + "left": "issue:45397", + "right": "issue:45656", + "accept": false, + "reason": "Different code paths: Gemma-4 Zero3 from_pretrained failure vs deepspeed causing optimizer.step() to run twice." + }, + { + "left": "issue:43065", + "right": "issue:43408", + "accept": false, + "reason": "Same model family, but different bugs: dummy Conv2d implementation vs a model-type warning/misinstantiation issue." + }, + { + "left": "issue:43723", + "right": "issue:46710", + "accept": false, + "reason": "Tokenizer loading problem vs post-upgrade accuracy regression; not the same underlying defect." + }, + { + "left": "issue:44368", + "right": "issue:45563", + "accept": false, + "reason": "Unrelated warnings in different features: tie_word_embeddings warning vs stale num_return_sequences warning in paged generate." + }, + { + "left": "issue:44995", + "right": "issue:45200", + "accept": false, + "reason": "Different models and failures: stale cache/indexer crash vs Gemma 4 multimodal token type defaults." + }, + { + "left": "issue:43931", + "right": "issue:44368", + "accept": false, + "reason": "Weight-shape mismatch for Qwen3-VL is unrelated to the tie_word_embeddings warning on Qwen3.5 fine-tuning." + }, + { + "left": "issue:44188", + "right": "issue:46202", + "accept": false, + "reason": "Different concrete problems: attention-kernel divergence under torch.compile vs Gemma pipeline roundtrip failure." + }, + { + "left": "issue:45987", + "right": "issue:46082", + "accept": false, + "reason": "Cache API return-type bug vs LlamaConfig validation of head_dim; distinct issues." + }, + { + "left": "issue:44483", + "right": "issue:46202", + "accept": false, + "reason": "API request rejection in /v1/chat/completions is unrelated to Gemma pipeline roundtrip behavior." + }, + { + "left": "issue:43906", + "right": "issue:46202", + "accept": false, + "reason": "An isolated reproduction of another issue is not the same as the Gemma roundtrip failure." + }, + { + "left": "issue:44995", + "right": "issue:45588", + "accept": false, + "reason": "Different failures: GLM MoE stale index cache on second forward vs flash_attention crashing on s_aux=None." + }, + { + "left": "issue:44188", + "right": "issue:46421", + "accept": false, + "reason": "torch.compile kernel divergence is unrelated to cache lazy initialization causing recompilation." + }, + { + "left": "issue:45685", + "right": "issue:46431", + "accept": false, + "reason": "MPS histogram dtype error and missing chat template are unrelated bugs." + }, + { + "left": "issue:45103", + "right": "issue:46858", + "accept": false, + "reason": "auto_docstring annotations crash is unrelated to StaticCache.get_seq_length returning a device tensor." + }, + { + "left": "issue:43065", + "right": "issue:46123", + "accept": false, + "reason": "Different SAM3 pixel-decoder placeholder issue vs mask generation partial-batch handling bug." + }, + { + "left": "issue:44361", + "right": "issue:44749", + "accept": false, + "reason": "Tokenizer AttributeError and a severe slowdown after upgrading are not the same defect." + }, + { + "left": "issue:44297", + "right": "issue:44623", + "accept": false, + "reason": "Both involve save_pretrained, but one is tokenizer_config class metadata and the other is processor file omission; not one concrete bug." + }, + { + "left": "issue:43994", + "right": "issue:46726", + "accept": false, + "reason": "Nonsensical model outputs vs missing vision backend handling are different problems." + }, + { + "left": "issue:44749", + "right": "issue:45656", + "accept": false, + "reason": "Performance regression after upgrade vs optimizer stepping twice under deepspeed are distinct issues." + }, + { + "left": "issue:43701", + "right": "issue:44749", + "accept": false, + "reason": "Checkpoint key mismatch is unrelated to the reported slowdown regression." + }, + { + "left": "issue:43646", + "right": "issue:44162", + "accept": false, + "reason": "Custom model initialization breakage in v5 is not the same as the ESM2 regression." + }, + { + "left": "issue:44871", + "right": "issue:46132", + "accept": false, + "reason": "Gemma eos_token_id mismatch is unrelated to AttentionInterface.register behavior changes." + }, + { + "left": "issue:44623", + "right": "issue:45656", + "accept": false, + "reason": "Processor save_pretrained file loss and duplicate optimizer steps are unrelated." + }, + { + "left": "issue:43441", + "right": "issue:46752", + "accept": false, + "reason": "FlashAttention failure for Ministral-3 is unrelated to apply_chat_template on an empty list." + }, + { + "left": "issue:43475", + "right": "issue:43643", + "accept": false, + "reason": "Missing Sam3VisionEncoderOutput field and AutoConfig trust_remote_code field loss are different code paths and symptoms." + }, + { + "left": "issue:42994", + "right": "issue:43482", + "accept": false, + "reason": "Quantized model saving failure vs Qwen2.5-GGUF loading failure are different operations and bugs." + }, + { + "left": "issue:45375", + "right": "issue:45478", + "accept": false, + "reason": "Related model family, but different failures: dropped config field under @strict vs from_pretrained error on Qwen3.5 Moe." + }, + { + "left": "issue:43475", + "right": "issue:46821", + "accept": false, + "reason": "SAM3 vision encoder attribute error is unrelated to TimesFM window_size AttributeError." + }, + { + "left": "issue:28282", + "right": "issue:30990", + "accept": false, + "reason": "Missing PyTorch ImportError and Sentence Transformers stuck loading are different failure modes." + } + ] + }, + "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": 26, + "review_unit_index": 46, + "review_unit_count": 61, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-505", + "nodes": [ + "issue:36010", + "issue:39692", + "issue:41628", + "issue:42371", + "issue:42673", + "issue:42831", + "issue:42994", + "issue:43232", + "issue:43525", + "issue:43673", + "issue:43704", + "issue:43723", + "issue:43784", + "issue:43824", + "issue:43873", + "issue:43986", + "issue:44038", + "issue:44188", + "issue:44222", + "issue:44351", + "issue:44373", + "issue:44623", + "issue:44857", + "issue:44869", + "issue:44933", + "issue:44945", + "issue:44964", + "issue:45081", + "issue:45092", + "issue:45103", + "issue:45137", + "issue:45230", + "issue:45310", + "issue:45356", + "issue:45375", + "issue:45397", + "issue:45406", + "issue:45593", + "issue:45753", + "issue:46202", + "issue:46326", + "issue:46399", + "issue:46463", + "issue:46701", + "issue:46726", + "issue:46851", + "issue:46877", + "issue:47000" + ], + "soft_pairs": [ + "issue:44964|issue:45137", + "issue:44188|issue:46851", + "issue:43232|issue:43673", + "issue:41628|issue:44933", + "issue:45310|issue:45375", + "issue:39692|issue:46726", + "issue:46877|issue:47000", + "issue:46463|issue:46877", + "issue:43986|issue:46726", + "issue:44869|issue:45356", + "issue:45230|issue:45397", + "issue:44038|issue:44623", + "issue:42994|issue:43525", + "issue:44373|issue:46851", + "issue:43723|issue:46326", + "issue:42371|issue:45092", + "issue:43873|issue:44945", + "issue:43784|issue:46399", + "issue:45406|issue:46202", + "issue:41628|issue:44351", + "issue:44945|issue:45753", + "issue:45137|issue:47000", + "issue:36010|issue:41628", + "issue:45081|issue:45103", + "issue:44857|issue:46877", + "issue:44038|issue:45310", + "issue:45593|issue:46877", + "issue:46202|issue:46701", + "issue:42994|issue:43824", + "issue:44222|issue:45310", + "issue:42673|issue:43704", + "issue:42831|issue:44222" + ], + "prepared_review_unit_hash": "932272859dcce178ef32b0a385618e859d6ef473dab1041fb54685d598d8cb7b", + "component_budget": { + "node_count": 505, + "item_count": 505, + "soft_pair_count": 1854, + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 32, + "serialized_chars": 37005, + "estimated_input_tokens": 9252, + "estimated_eval_tokens": 18760 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "Mostly unrelated issues; only the Qwen3ForCausalLM VRAM leak pair is an exact duplicate. Everything else differs by symbol, model family, or failure mode enough to keep separate.", + "confidence": 0.97, + "canonical_issue_reason": "Use issue:42673 as canonical: it is the earlier, clearer report of the same Qwen3ForCausalLM VRAM leak under multiple dataloader threads, and issue:43704 is a later duplicate of the same symptom.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:42673 is the best representative overall because it is specific, reproducible, and has the clearest duplicate relationship in the set.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44964", + "right": "issue:45137", + "accept": false, + "reason": "Different failures: model loading with latest transformers vs DeepSpeed ZeRO3 deque error." + }, + { + "left": "issue:44188", + "right": "issue:46851", + "accept": false, + "reason": "Both involve generation internals, but one is attention-kernel divergence under torch.compile and the other is variable-length position_ids behavior in Qwen3.5." + }, + { + "left": "issue:43232", + "right": "issue:43673", + "accept": false, + "reason": "Generation flow issues, but one is sync_gpus/_update_model_kwargs_for_generation and the other is chunked_prefill cache availability in v5.0.0." + }, + { + "left": "issue:41628", + "right": "issue:44933", + "accept": false, + "reason": "Both are import problems, but they concern different missing names from different modules and code paths." + }, + { + "left": "issue:45310", + "right": "issue:45375", + "accept": false, + "reason": "Same model family, but one is a from_pretrained error and the other is a config field being dropped by @strict." + }, + { + "left": "issue:39692", + "right": "issue:46726", + "accept": false, + "reason": "Doc-example/model mismatch and quantization errors are not the same as AutoProcessor's missing-vision-backend error." + }, + { + "left": "issue:46877", + "right": "issue:47000", + "accept": false, + "reason": "Unrelated issues: import-time doc-lint stdout noise vs RTDetrLoss infeasible cost matrix crash." + }, + { + "left": "issue:46463", + "right": "issue:46877", + "accept": false, + "reason": "Different subsystems entirely: T5 decoder-input validation messaging vs auto_docstring import-time diagnostics." + }, + { + "left": "issue:43986", + "right": "issue:46726", + "accept": false, + "reason": "Both mention processor/backend availability, but one is a video-model load crash without torchvision and the other is an AutoProcessor image-processor error." + }, + { + "left": "issue:44869", + "right": "issue:45356", + "accept": false, + "reason": "Both are tokenizer/decoding bugs, but they affect different models and different concrete failure modes." + }, + { + "left": "issue:45230", + "right": "issue:45397", + "accept": false, + "reason": "A generic bug report is not the same as the specific gemma-4 Zero3 from_pretrained issue." + }, + { + "left": "issue:44038", + "right": "issue:44623", + "accept": false, + "reason": "Different symptoms and likely different paths: Qwen3-VL-Moe bug vs processor.save_pretrained missing files." + }, + { + "left": "issue:42994", + "right": "issue:43525", + "accept": false, + "reason": "Quantized model saving failure is unrelated to Llama4Config missing pad_token_id." + }, + { + "left": "issue:44373", + "right": "issue:46851", + "accept": false, + "reason": "Docstring wording for position_ids is not the same as Qwen3.5 variable-length input behavior." + }, + { + "left": "issue:43723", + "right": "issue:46326", + "accept": false, + "reason": "Tokenizer loading failure in v5 is not the same as continue_final_message behavior for mistral_common tokenizers." + }, + { + "left": "issue:42371", + "right": "issue:45092", + "accept": false, + "reason": "TF32 API guidance and InternVL2 v5 meta-init incompatibility are unrelated." + }, + { + "left": "issue:43873", + "right": "issue:44945", + "accept": false, + "reason": "Quantization offloading behavior and pipeline-parallel output corruption are different bugs." + }, + { + "left": "issue:43784", + "right": "issue:46399", + "accept": false, + "reason": "Sentence-transformers import NameError and Qwen3.5 kernel registration crash are different code paths." + }, + { + "left": "issue:45406", + "right": "issue:46202", + "accept": false, + "reason": "Gemma4Processor missing _tokenizer and Gemma pipeline roundtrip on \u2581 are different issues." + }, + { + "left": "issue:41628", + "right": "issue:44351", + "accept": false, + "reason": "Both are import errors from transformers, but they are for different symbols and causes." + }, + { + "left": "issue:44945", + "right": "issue:45753", + "accept": false, + "reason": "Pipeline-parallel incorrect output and Qwen3_5 infinite loop on one image are different runtime failures." + }, + { + "left": "issue:45137", + "right": "issue:47000", + "accept": false, + "reason": "DeepSpeed ZeRO3 deque underflow is unrelated to RTDetrLoss cost-matrix infeasibility." + }, + { + "left": "issue:36010", + "right": "issue:41628", + "accept": false, + "reason": "Different missing imports: GenerationMixin from transformers.generation vs AutoImageProcessor from transformers." + }, + { + "left": "issue:45081", + "right": "issue:45103", + "accept": false, + "reason": "Tokenizer regex patch crash and auto_docstring annotation handling are unrelated." + }, + { + "left": "issue:44857", + "right": "issue:46877", + "accept": false, + "reason": "Float16 AMP CUDA crash in LwDetrImageLoss is unrelated to doc-lint stdout logging." + }, + { + "left": "issue:44038", + "right": "issue:45310", + "accept": false, + "reason": "Both mention Qwen3 MoE, but the concrete crash conditions and failure modes are different." + }, + { + "left": "issue:45593", + "right": "issue:46877", + "accept": false, + "reason": "D-FINE auxiliary-loss behavior is unrelated to auto_docstring diagnostics output." + }, + { + "left": "issue:46202", + "right": "issue:46701", + "accept": false, + "reason": "Both touch Gemma pipelines, but one is a roundtrip tokenizer edge case and the other is an image-text-to-text pipeline failure." + }, + { + "left": "issue:42994", + "right": "issue:43824", + "accept": false, + "reason": "Quantized saving failure and Qwen2_5_VL import error are different problems." + }, + { + "left": "issue:44222", + "right": "issue:45310", + "accept": false, + "reason": "FP8 save_pretrained moe and Qwen3.5 MoE from_pretrained error are different bugs." + }, + { + "left": "issue:42673", + "right": "issue:43704", + "accept": true, + "reason": "Same underlying bug: Qwen3ForCausalLM leaks VRAM when used across multiple dataloader threads; the later issue is a duplicate of the same symptom." + }, + { + "left": "issue:42831", + "right": "issue:44222", + "accept": false, + "reason": "Both relate to FP8, but one is an accuracy issue and the other is a save_pretrained bug." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "Grounded overall. The Qwen3ForCausalLM pair (issue:42673 / issue:43704) is strongly supported as a duplicate by the identical titles and earlier/later timestamps, and the rest of the soft-edge rejections stay conservative based on the packet titles alone. Minor note: avoid implying reproducibility unless it\u2019s explicit in the issue body." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 26, + "review_unit_index": 47, + "review_unit_count": 61, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-505", + "nodes": [ + "issue:27957", + "issue:30333", + "issue:30990", + "issue:33290", + "issue:34689", + "issue:35141", + "issue:39692", + "issue:42175", + "issue:42222", + "issue:42673", + "issue:42898", + "issue:43232", + "issue:43295", + "issue:43479", + "issue:43504", + "issue:43723", + "issue:43784", + "issue:43824", + "issue:43874", + "issue:43976", + "issue:44016", + "issue:44079", + "issue:44496", + "issue:44568", + "issue:44623", + "issue:44734", + "issue:44779", + "issue:44792", + "issue:44869", + "issue:44933", + "issue:45092", + "issue:45103", + "issue:45200", + "issue:45310", + "issue:45356", + "issue:45405", + "issue:45561", + "issue:45600", + "issue:45712", + "issue:45812", + "issue:46077", + "issue:46123", + "issue:46231", + "issue:46693", + "issue:46829", + "issue:46877", + "issue:46909" + ], + "soft_pairs": [ + "issue:43232|issue:44792", + "issue:42175|issue:44933", + "issue:43504|issue:45561", + "issue:45561|issue:45600", + "issue:44496|issue:45561", + "issue:45405|issue:46877", + "issue:43976|issue:44933", + "issue:30333|issue:33290", + "issue:30990|issue:35141", + "issue:42898|issue:44568", + "issue:42175|issue:45712", + "issue:43479|issue:46829", + "issue:44623|issue:45310", + "issue:46077|issue:46123", + "issue:42673|issue:46693", + "issue:42898|issue:45812", + "issue:43479|issue:44869", + "issue:43723|issue:44568", + "issue:44016|issue:46231", + "issue:39692|issue:46909", + "issue:42898|issue:44779", + "issue:43479|issue:44079", + "issue:43784|issue:43824", + "issue:27957|issue:30333", + "issue:42898|issue:43295", + "issue:30333|issue:34689", + "issue:43874|issue:45092", + "issue:44734|issue:45103", + "issue:42898|issue:45356", + "issue:42222|issue:45200" + ], + "prepared_review_unit_hash": "af7d2ca785f1e43a657ffdea1a0ef40f1589edf71445078b865cb15e8d20bd19", + "component_budget": { + "node_count": 505, + "item_count": 505, + "soft_pair_count": 1854, + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 30, + "serialized_chars": 36378, + "estimated_input_tokens": 9095, + "estimated_eval_tokens": 18446 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is heterogeneous: it mixes unrelated issue types across tokenization, model loading, multimodal configs, docs, and test infra. No soft pair looks like a true duplicate or the same fixable code-path problem, and there are no PRs here. If one issue must anchor the cluster, the open and more active #45561 is the best representative, but it is not a real duplicate of the others.", + "confidence": 0.95, + "canonical_issue_reason": "#45561 is the strongest cluster representative because it is open, has multiple inbound references, and has the most current triage signal among an otherwise mixed set of issues.", + "canonical_pr_reason": null, + "best_issue_reason": "#45561 is the most suitable single issue to surface for manual review, but only as a representative anchor; the cluster does not form a coherent duplicate set.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43232", + "right": "issue:44792", + "accept": false, + "reason": "Different generation bug vs janus image-generation test failure; no shared code path." + }, + { + "left": "issue:42175", + "right": "issue:44933", + "accept": false, + "reason": "Backend dependency packaging issue vs non-existent import from image_utils; unrelated causes." + }, + { + "left": "issue:43504", + "right": "issue:45561", + "accept": false, + "reason": "Legacy pretrained-load bug vs pytest-xdist file-write race; different subsystems." + }, + { + "left": "issue:45561", + "right": "issue:45600", + "accept": false, + "reason": "Captured-info race in tests vs stale auto_mappings config references; not the same bug." + }, + { + "left": "issue:44496", + "right": "issue:45561", + "accept": false, + "reason": "Unrecognized model/config loading error vs test utility race; distinct failure modes." + }, + { + "left": "issue:45405", + "right": "issue:46877", + "accept": false, + "reason": "Dependency version bump issue vs docstring import-time stderr/stdout behavior; unrelated." + }, + { + "left": "issue:43976", + "right": "issue:44933", + "accept": false, + "reason": "Python version compatibility vs missing import symbol; different problems." + }, + { + "left": "issue:30333", + "right": "issue:33290", + "accept": false, + "reason": "MLflow reporting on failed jobs vs DeepSpeed adafactor OOM; no shared underlying defect." + }, + { + "left": "issue:30990", + "right": "issue:35141", + "accept": false, + "reason": "Loading hang in sentence-transformers vs embedding resize/post_init reinit bug; different code paths." + }, + { + "left": "issue:42898", + "right": "issue:44568", + "accept": false, + "reason": "Both are tokenizer regressions in v5, but they affect different behaviors and fixes." + }, + { + "left": "issue:42175", + "right": "issue:45712", + "accept": false, + "reason": "Optional TensorFlow backend packaging vs leftover dummy PT objects; unrelated repository hygiene issues." + }, + { + "left": "issue:43479", + "right": "issue:46829", + "accept": false, + "reason": "Config default initialization bug vs wrong output tensor field in video classification output; different models and logic." + }, + { + "left": "issue:44623", + "right": "issue:45310", + "accept": false, + "reason": "Processor save_pretrained missing files vs Qwen3.5 MoE from_pretrained error; not the same defect." + }, + { + "left": "issue:46077", + "right": "issue:46123", + "accept": false, + "reason": "Attention-mask construction bug vs batch-finalization logic in a generation pipeline; different paths." + }, + { + "left": "issue:42673", + "right": "issue:46693", + "accept": false, + "reason": "VRAM leak from multiple dataloader threads vs long-context FA2 performance regression; distinct symptoms and fixes." + }, + { + "left": "issue:42898", + "right": "issue:45812", + "accept": false, + "reason": "Both are tokenization regressions, but one is clean-up spacing and the other is wrong Granite token IDs." + }, + { + "left": "issue:43479", + "right": "issue:44869", + "accept": false, + "reason": "Multimodal config defaults vs Whisper timestamp decoding crash; unrelated components." + }, + { + "left": "issue:43723", + "right": "issue:44568", + "accept": false, + "reason": "AutoTokenizer load issue in v5 vs add_special_tokens BOS/EOS regression; different tokenizer failures." + }, + { + "left": "issue:44016", + "right": "issue:46231", + "accept": false, + "reason": "Notebook syntax error vs documentation parameter-name mismatch; different artifact types." + }, + { + "left": "issue:39692", + "right": "issue:46909", + "accept": false, + "reason": "SigLIP2 example with model/processor mismatch and quantization failure vs broken DIA docs example; separate docs issues." + }, + { + "left": "issue:42898", + "right": "issue:44779", + "accept": false, + "reason": "General tokenization cleanup regression vs Deepseek tokenizer codec regression; different concrete bugs." + }, + { + "left": "issue:43479", + "right": "issue:44079", + "accept": false, + "reason": "Multimodal config defaulting bug vs ModelOutput key assignment bug; unrelated internals." + }, + { + "left": "issue:43784", + "right": "issue:43824", + "accept": false, + "reason": "sentence-transformers import NameError vs missing Qwen2.5-VL symbol export; different import failures." + }, + { + "left": "issue:27957", + "right": "issue:30333", + "accept": false, + "reason": "Flash Attention 2 support for XLM-Roberta vs MLflow reporting after failed training; unrelated." + }, + { + "left": "issue:42898", + "right": "issue:43295", + "accept": false, + "reason": "Both are v5 regressions, but one is token cleanup behavior and the other is custom processor/tokenizer access." + }, + { + "left": "issue:30333", + "right": "issue:34689", + "accept": false, + "reason": "Failed-job MLflow tracking vs model loading breakage for Llama 3.2 Vision; different code paths." + }, + { + "left": "issue:43874", + "right": "issue:45092", + "accept": false, + "reason": "Missing image-patch helper in GLM46V processor vs InternVL2 meta-init incompatibility; different multimodal bugs." + }, + { + "left": "issue:44734", + "right": "issue:45103", + "accept": false, + "reason": "KV-cache continuation crash in serving vs auto_docstring annotation parsing crash; unrelated." + }, + { + "left": "issue:42898", + "right": "issue:45356", + "accept": false, + "reason": "Tokenizer cleanup behavior change vs Kimi-K2.5 codec/warning regression; distinct tokenizer issues." + }, + { + "left": "issue:42222", + "right": "issue:45200", + "accept": false, + "reason": "Broken VitPose model files vs Gemma 4 token-type defaulting; separate multimodal/model-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": 26, + "review_unit_index": 48, + "review_unit_count": 61, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-505", + "nodes": [ + "issue:22355", + "issue:28282", + "issue:30333", + "issue:36331", + "issue:41628", + "issue:42757", + "issue:42796", + "issue:42831", + "issue:43116", + "issue:43122", + "issue:43232", + "issue:43479", + "issue:43644", + "issue:43723", + "issue:44060", + "issue:44075", + "issue:44190", + "issue:44222", + "issue:44263", + "issue:44295", + "issue:44297", + "issue:44360", + "issue:44373", + "issue:44448", + "issue:44484", + "issue:44485", + "issue:44623", + "issue:44933", + "issue:44977", + "issue:45071", + "issue:45092", + "issue:45125", + "issue:45245", + "issue:45310", + "issue:45405", + "issue:45656", + "issue:45712", + "issue:45834", + "issue:46082", + "issue:46132", + "issue:46209", + "issue:46231", + "issue:46424", + "issue:46431", + "issue:46614", + "issue:46710", + "issue:46762", + "issue:46932" + ], + "soft_pairs": [ + "issue:44295|issue:45656", + "issue:30333|issue:36331", + "issue:45405|issue:45712", + "issue:22355|issue:28282", + "issue:43116|issue:44190", + "issue:44222|issue:44297", + "issue:44075|issue:46231", + "issue:44060|issue:45125", + "issue:44933|issue:45712", + "issue:42796|issue:43723", + "issue:44448|issue:46132", + "issue:42831|issue:44623", + "issue:44297|issue:45656", + "issue:43479|issue:43644", + "issue:44222|issue:44623", + "issue:45245|issue:46132", + "issue:44360|issue:46762", + "issue:46132|issue:46710", + "issue:41628|issue:42757", + "issue:44297|issue:45310", + "issue:44373|issue:44485", + "issue:43723|issue:44977", + "issue:46614|issue:46932", + "issue:45071|issue:46082", + "issue:44222|issue:44484", + "issue:43122|issue:46209", + "issue:44190|issue:46082", + "issue:44263|issue:45092", + "issue:43232|issue:46424", + "issue:46209|issue:46431", + "issue:43232|issue:45834" + ], + "prepared_review_unit_hash": "ee6e7a63cf8275982f705feb69df3ad0176eb4797baa4ccda2b76b73216912a2", + "component_budget": { + "node_count": 505, + "item_count": 505, + "soft_pair_count": 1854, + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 31, + "serialized_chars": 36655, + "estimated_input_tokens": 9164, + "estimated_eval_tokens": 18584 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This is a loose similarity cluster, not a true duplicate set. The items mostly share broad themes like import errors, tokenizer/model save_pretrained behavior, or v5 regressions, but they describe different code paths and fixes. All soft edges should be rejected. No PRs are present.", + "confidence": 0.86, + "canonical_issue_reason": "issue:44297 is the most central concrete bug in the set and sits near several serialization-related similarities, but it is still too specific to serve as a real duplicate umbrella for the cluster.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:44297 is a well-scoped, actionable runtime bug with clear reproduction around tokenizer serialization, making it the best representative issue among these heterogeneous reports.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44295", + "right": "issue:45656", + "accept": false, + "reason": "Different subsystems and failures: buffer access bug vs deepspeed optimizer stepping." + }, + { + "left": "issue:30333", + "right": "issue:36331", + "accept": false, + "reason": "Unrelated training-reporting/MLflow issue versus a custom trainer signature regression." + }, + { + "left": "issue:45405", + "right": "issue:45712", + "accept": false, + "reason": "PEFT version pinning and dummy class cleanup are unrelated." + }, + { + "left": "issue:22355", + "right": "issue:28282", + "accept": false, + "reason": "Both are import/dependency errors, but they concern different missing modules and causes." + }, + { + "left": "issue:43116", + "right": "issue:44190", + "accept": false, + "reason": "Both touch example scripts, but one is multi-label classification output and the other is local dataset loading." + }, + { + "left": "issue:44222", + "right": "issue:44297", + "accept": false, + "reason": "Both involve save_pretrained, but one is FP8 model serialization and the other is tokenizer config metadata." + }, + { + "left": "issue:44075", + "right": "issue:46231", + "accept": false, + "reason": "Optimizer argument usage and a docstring parameter-name mismatch are not the same bug." + }, + { + "left": "issue:44060", + "right": "issue:45125", + "accept": false, + "reason": "Both are Qwen3_5Moe-related, but one is a tied-weights warning and the other is missing tensor-parallel plan support." + }, + { + "left": "issue:44933", + "right": "issue:45712", + "accept": false, + "reason": "Different export/import hygiene problems with different targets and failure modes." + }, + { + "left": "issue:42796", + "right": "issue:43723", + "accept": false, + "reason": "Tokenizer BPE-marker handling and tokenizer loading failures are distinct issues." + }, + { + "left": "issue:44448", + "right": "issue:46132", + "accept": false, + "reason": "Different regressions: model output drift versus attention registration behavior." + }, + { + "left": "issue:42831", + "right": "issue:44623", + "accept": false, + "reason": "FP8 accuracy and processor save_pretrained file omission are unrelated." + }, + { + "left": "issue:44297", + "right": "issue:45656", + "accept": false, + "reason": "Tokenizer serialization metadata bug versus optimizer stepping bug; no shared code path." + }, + { + "left": "issue:43479", + "right": "issue:43644", + "accept": false, + "reason": "Config default initialization and non-persistent buffer corruption are separate problems." + }, + { + "left": "issue:44222", + "right": "issue:44623", + "accept": false, + "reason": "Both mention save_pretrained, but they affect different objects and different serialization paths." + }, + { + "left": "issue:45245", + "right": "issue:46132", + "accept": false, + "reason": "Category-count validation is unrelated to attention registry behavior changes." + }, + { + "left": "issue:44360", + "right": "issue:46762", + "accept": false, + "reason": "Different model-specific numerical/architecture issues with no clear shared root cause." + }, + { + "left": "issue:46132", + "right": "issue:46710", + "accept": false, + "reason": "Both are regressions, but they affect different attention/model behavior and fixes." + }, + { + "left": "issue:41628", + "right": "issue:42757", + "accept": false, + "reason": "Both are import errors, but for different symbols from different packages." + }, + { + "left": "issue:44297", + "right": "issue:45310", + "accept": false, + "reason": "Tokenizer save_pretrained metadata mismatch is not the same as Qwen3.5 MoE from_pretrained failure." + }, + { + "left": "issue:44373", + "right": "issue:44485", + "accept": false, + "reason": "Docstring mismatch is not the same as a RoPE implementation bug." + }, + { + "left": "issue:43723", + "right": "issue:44977", + "accept": false, + "reason": "Tokenizer loading and flash-attention generation failures are unrelated." + }, + { + "left": "issue:46614", + "right": "issue:46932", + "accept": false, + "reason": "Stale phonemizer backend state is unrelated to Gemma4 quantized embedder dtype casting." + }, + { + "left": "issue:45071", + "right": "issue:46082", + "accept": false, + "reason": "General type-checking breakage and a specific head_dim validation edge case are different issues." + }, + { + "left": "issue:44222", + "right": "issue:44484", + "accept": false, + "reason": "A concrete serialization bug is not the same as a question about shard size defaults." + }, + { + "left": "issue:43122", + "right": "issue:46209", + "accept": false, + "reason": "Tokenizer output drift and FP8 inference failure are unrelated." + }, + { + "left": "issue:44263", + "right": "issue:45092", + "accept": false, + "reason": "torch.split indexing behavior and meta-initialization incompatibility are different code paths." + }, + { + "left": "issue:43232", + "right": "issue:46424", + "accept": false, + "reason": "Generation kwargs updates and cache length config handling are not the same bug." + }, + { + "left": "issue:46209", + "right": "issue:46431", + "accept": false, + "reason": "FP8 inference on B200 and missing chat template validation are unrelated." + }, + { + "left": "issue:43232", + "right": "issue:45834", + "accept": false, + "reason": "Generation sync behavior and OCR long-input indexing 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": 26, + "review_unit_index": 49, + "review_unit_count": 61, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-505", + "nodes": [ + "issue:22355", + "issue:27957", + "issue:28218", + "issue:30990", + "issue:38175", + "issue:39401", + "issue:41628", + "issue:42757", + "issue:42796", + "issue:42831", + "issue:42886", + "issue:42907", + "issue:42994", + "issue:43066", + "issue:43122", + "issue:43232", + "issue:43408", + "issue:43475", + "issue:43644", + "issue:43824", + "issue:44038", + "issue:44188", + "issue:44261", + "issue:44263", + "issue:44279", + "issue:44295", + "issue:44360", + "issue:44373", + "issue:44483", + "issue:44568", + "issue:44623", + "issue:44829", + "issue:45092", + "issue:45278", + "issue:45310", + "issue:45357", + "issue:45381", + "issue:45406", + "issue:45593", + "issue:45753", + "issue:46002", + "issue:46050", + "issue:46305", + "issue:46328", + "issue:46369", + "issue:46402", + "issue:46655", + "issue:46909" + ], + "soft_pairs": [ + "issue:30990|issue:38175", + "issue:43475|issue:46305", + "issue:46305|issue:46655", + "issue:44295|issue:44623", + "issue:46369|issue:46909", + "issue:44263|issue:44373", + "issue:42907|issue:46050", + "issue:44263|issue:44360", + "issue:45092|issue:45357", + "issue:42831|issue:45310", + "issue:44279|issue:46328", + "issue:45381|issue:46305", + "issue:44483|issue:45406", + "issue:22355|issue:27957", + "issue:44829|issue:46002", + "issue:41628|issue:43824", + "issue:42831|issue:44038", + "issue:44261|issue:45092", + "issue:43644|issue:46305", + "issue:43408|issue:46402", + "issue:22355|issue:28218", + "issue:42831|issue:45753", + "issue:39401|issue:44568", + "issue:42886|issue:43066", + "issue:42886|issue:42994", + "issue:42757|issue:45278", + "issue:27957|issue:30990", + "issue:43408|issue:46369", + "issue:44188|issue:44373", + "issue:42796|issue:43122", + "issue:43232|issue:45593", + "issue:45381|issue:46909" + ], + "prepared_review_unit_hash": "fcaeeb798fabb08fcb4a4a39b3c1a2d12aa8415ced280c8ad78f61d457120f60", + "component_budget": { + "node_count": 505, + "item_count": 505, + "soft_pair_count": 1854, + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 32, + "serialized_chars": 36802, + "estimated_input_tokens": 9201, + "estimated_eval_tokens": 18658 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The set is heterogeneous and the soft-similarity links are mostly superficial (same subsystem or related theme, but not the same concrete bug/change). I would not merge any of the candidate pairs as duplicates.", + "confidence": 0.62, + "canonical_issue_reason": "issue:46305 is the most actionable single issue in the set: it describes a precise, reproducible off-by-one bug in a specific code path and is an active/open report, making it the best canonical target among these mostly unrelated issues.", + "canonical_pr_reason": "No pull requests are present in the provided items.", + "best_issue_reason": "issue:46305 is the strongest issue candidate overall because it is concrete, narrow, and clearly states the failure mode; it is a better duplicate anchor than the broader or more ambiguous reports here.", + "best_pr_reason": "No pull requests are present in the provided items.", + "soft_edge_verdicts": [ + { + "left": "issue:30990", + "right": "issue:38175", + "accept": false, + "reason": "Both are model/runtime problems, but one is a loading hang and the other is a zero-probability accuracy issue; they are not the same bug." + }, + { + "left": "issue:43475", + "right": "issue:46305", + "accept": false, + "reason": "These affect different SAM3 vs MoonshineStreaming code paths and describe unrelated failures." + }, + { + "left": "issue:46305", + "right": "issue:46655", + "accept": false, + "reason": "Both mention video-related processing, but one is a streaming mask off-by-one and the other is a video padding calculation bug." + }, + { + "left": "issue:44295", + "right": "issue:44623", + "accept": false, + "reason": "Registering position_ids as a buffer and missing files on processor save_pretrained are distinct issues." + }, + { + "left": "issue:46369", + "right": "issue:46909", + "accept": false, + "reason": "Memory overflow in sam3-litetext and a broken DIA documentation example are unrelated." + }, + { + "left": "issue:44263", + "right": "issue:44373", + "accept": false, + "reason": "A torch.split indexing bug and a wrong docstring are not the same underlying problem." + }, + { + "left": "issue:42907", + "right": "issue:46050", + "accept": false, + "reason": "Both concern saving quantized models, but they target different quantization flows and failure modes." + }, + { + "left": "issue:44263", + "right": "issue:44360", + "accept": false, + "reason": "A return-value mismatch in DSA indexing is different from the missing ReLU discussion." + }, + { + "left": "issue:45092", + "right": "issue:45357", + "accept": false, + "reason": "Meta-initialization incompatibility for old InternVL2 checkpoints is not the same as incorrect visual encoder keys in Qwen3.5 save_pretrained." + }, + { + "left": "issue:42831", + "right": "issue:45310", + "accept": false, + "reason": "FineGrainedFP8 accuracy drift and Qwen3.5 from_pretrained errors are different bugs." + }, + { + "left": "issue:44279", + "right": "issue:46328", + "accept": false, + "reason": "A general dependency issue is unrelated to a performance inefficiency in SequenceFeatureExtractor.pad." + }, + { + "left": "issue:45381", + "right": "issue:46305", + "accept": false, + "reason": "They are both off-by-one style bugs, but in different models and different masking/positioning logic." + }, + { + "left": "issue:44483", + "right": "issue:45406", + "accept": false, + "reason": "An API request rejection and a Gemma4Processor tokenizer attribute crash are unrelated." + }, + { + "left": "issue:22355", + "right": "issue:27957", + "accept": false, + "reason": "Missing transformers.onnx and XLMRoberta Flash Attention 2 are completely different issues." + }, + { + "left": "issue:44829", + "right": "issue:46002", + "accept": false, + "reason": "Flash Attention 3 training degeneration and NaNs after checkout/init are different failure modes." + }, + { + "left": "issue:41628", + "right": "issue:43824", + "accept": false, + "reason": "These are separate missing-import cases for different symbols, not one concrete bug." + }, + { + "left": "issue:42831", + "right": "issue:44038", + "accept": false, + "reason": "A FineGrainedFP8 accuracy report and a Qwen3-VL-Moe bug report are too different to be duplicates." + }, + { + "left": "issue:44261", + "right": "issue:45092", + "accept": false, + "reason": "Missing rms_norm_eps in MLA q_a_layernorm and remote-code meta-init incompatibility are distinct root causes." + }, + { + "left": "issue:43644", + "right": "issue:46305", + "accept": false, + "reason": "Non-persistent buffer junk in Transformers 5.0.0 is unrelated to a streaming mask off-by-one bug." + }, + { + "left": "issue:43408", + "right": "issue:46402", + "accept": false, + "reason": "A model-type warning and missing text encoder weight loading are different problems." + }, + { + "left": "issue:22355", + "right": "issue:28218", + "accept": false, + "reason": "Importing transformers.onnx and tokenizer spacing after added tokens are unrelated." + }, + { + "left": "issue:42831", + "right": "issue:45753", + "accept": false, + "reason": "FP8 accuracy issues are not the same as Qwen3_5 looping on a specific image." + }, + { + "left": "issue:39401", + "right": "issue:44568", + "accept": false, + "reason": "Wrong offset_mapping and missing BOS/EOS insertion are both tokenizer bugs, but not the same concrete defect." + }, + { + "left": "issue:42886", + "right": "issue:43066", + "accept": false, + "reason": "Offline cache loading failure and wrong tokenizer decoder type are different issues." + }, + { + "left": "issue:42886", + "right": "issue:42994", + "accept": false, + "reason": "Tokenizer offline-cache loading and quantized model saving are unrelated." + }, + { + "left": "issue:42757", + "right": "issue:45278", + "accept": false, + "reason": "A missing huggingface_hub symbol and broad import errors after upgrade overlap only at a very high level, not as one concrete bug." + }, + { + "left": "issue:27957", + "right": "issue:30990", + "accept": false, + "reason": "Flash Attention 2 with XLMRoberta and a Sentence Transformers loading stall are distinct." + }, + { + "left": "issue:43408", + "right": "issue:46369", + "accept": false, + "reason": "SAM3 model-type mismatch and SAM3-litetext memory overflow are different failures." + }, + { + "left": "issue:44188", + "right": "issue:44373", + "accept": false, + "reason": "A torch.compile attention-kernel divergence bug is unrelated to a docstring issue." + }, + { + "left": "issue:42796", + "right": "issue:43122", + "accept": false, + "reason": "Ministral BPE-marker handling and generic tokenization differences across versions are related themes but not the same bug." + }, + { + "left": "issue:43232", + "right": "issue:45593", + "accept": false, + "reason": "Generation kwargs after sync_gpus and D-FINE auxiliary-loss behavior are unrelated." + }, + { + "left": "issue:45381", + "right": "issue:46909", + "accept": false, + "reason": "A Qwen2.5-VL video position_ids bug and a DIA documentation example problem are not duplicates." + } + ] + }, + "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": 26, + "review_unit_index": 50, + "review_unit_count": 61, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-505", + "nodes": [ + "issue:22355", + "issue:29159", + "issue:30990", + "issue:33357", + "issue:39401", + "issue:41628", + "issue:42371", + "issue:42907", + "issue:43122", + "issue:43408", + "issue:43452", + "issue:43475", + "issue:43540", + "issue:43576", + "issue:43646", + "issue:43756", + "issue:43784", + "issue:43824", + "issue:43937", + "issue:43976", + "issue:44038", + "issue:44060", + "issue:44117", + "issue:44222", + "issue:44263", + "issue:44295", + "issue:44297", + "issue:44351", + "issue:44360", + "issue:44485", + "issue:44488", + "issue:44521", + "issue:44625", + "issue:44704", + "issue:44829", + "issue:44945", + "issue:44964", + "issue:45216", + "issue:45278", + "issue:45290", + "issue:45478", + "issue:45600", + "issue:46002", + "issue:46172", + "issue:46326", + "issue:46614", + "issue:46762", + "issue:46909" + ], + "soft_pairs": [ + "issue:43540|issue:46909", + "issue:43646|issue:46002", + "issue:30990|issue:33357", + "issue:42907|issue:44945", + "issue:42907|issue:44964", + "issue:43976|issue:44945", + "issue:43756|issue:44060", + "issue:39401|issue:44488", + "issue:44295|issue:44297", + "issue:41628|issue:45278", + "issue:43576|issue:43784", + "issue:42371|issue:44263", + "issue:42907|issue:43937", + "issue:43784|issue:45600", + "issue:46172|issue:46762", + "issue:44117|issue:44704", + "issue:39401|issue:44521", + "issue:44625|issue:44829", + "issue:44704|issue:45290", + "issue:43122|issue:44521", + "issue:43408|issue:43475", + "issue:44360|issue:44485", + "issue:43122|issue:46326", + "issue:42907|issue:43452", + "issue:44038|issue:44297", + "issue:22355|issue:29159", + "issue:45216|issue:45478", + "issue:44038|issue:44295", + "issue:39401|issue:46614", + "issue:44038|issue:44222", + "issue:43824|issue:44351" + ], + "prepared_review_unit_hash": "38dc7670a268f95145de96930ea5ee3fc62cf79da15636095a40f36bab423d87", + "component_budget": { + "node_count": 505, + "item_count": 505, + "soft_pair_count": 1854, + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 31, + "serialized_chars": 36529, + "estimated_input_tokens": 9133, + "estimated_eval_tokens": 18522 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The cluster is mostly a collection of unrelated bug reports that only look superficially similar by model family, version, or subsystem. None of the soft pairs appear to describe the same concrete bug/change, so I would not merge any of them as duplicates. The broadest umbrella-style report is the 5.5.0 import-regression issue.", + "confidence": 0.93, + "canonical_issue_reason": "issue:45278 is the closest thing to a canonical issue because it is the broadest umbrella report in the set: a general import-regression after upgrading to 5.5.0, which is more representative than the model-specific or single-symptom tickets.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45278 is the best overall representative because it describes a wider failure mode and could subsume related import-breakage reports better than the narrower Qwen/SAM/tokenizer issues.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43540", + "right": "issue:46909", + "accept": false, + "reason": "Different bugs: Qwen3OmniMoe video-input validation vs a broken DIA documentation example." + }, + { + "left": "issue:43646", + "right": "issue:46002", + "accept": false, + "reason": "Both concern model initialization, but one is a Transformers 5.0 custom-model init breakage and the other is NaNs after checkout/init; not the same code-path bug." + }, + { + "left": "issue:30990", + "right": "issue:33357", + "accept": false, + "reason": "Unrelated symptoms: Sentence-Transformers loading hang vs a MacOS bus error with a CLIP model." + }, + { + "left": "issue:42907", + "right": "issue:44945", + "accept": false, + "reason": "Different model/pathology: dequantized save failure vs incorrect output under pipeline parallelism." + }, + { + "left": "issue:42907", + "right": "issue:44964", + "accept": false, + "reason": "One is about saving dequantized Ministrals/Devstrals, the other about loading Phi-4 multimodal; no concrete overlap." + }, + { + "left": "issue:43976", + "right": "issue:44945", + "accept": false, + "reason": "Python-version compatibility breakage is not the same bug as pipeline-parallel output corruption." + }, + { + "left": "issue:43756", + "right": "issue:44060", + "accept": false, + "reason": "Different model implementations and defects: Smollm3 RoPE-layer loss vs Qwen3-Next tied-weights warning." + }, + { + "left": "issue:39401", + "right": "issue:44488", + "accept": false, + "reason": "Qwen3 tokenizer offset_mapping bug is unrelated to inability to load cjvt/sleng-bert." + }, + { + "left": "issue:44295", + "right": "issue:44297", + "accept": false, + "reason": "One is a position_ids buffer access problem, the other is tokenizer_class metadata mismatch on save_pretrained." + }, + { + "left": "issue:41628", + "right": "issue:45278", + "accept": false, + "reason": "Both involve imports, but one is a specific missing AutoImageProcessor import and the other is a broad post-upgrade import-regression report; not the same concrete defect." + }, + { + "left": "issue:43576", + "right": "issue:43784", + "accept": false, + "reason": "Broken CLI/env command vs an import NameError in sentence-transformers; different failure modes." + }, + { + "left": "issue:42371", + "right": "issue:44263", + "accept": false, + "reason": "TF32 API guidance issue and a torch.split return-value complaint in a DSA indexer are unrelated." + }, + { + "left": "issue:42907", + "right": "issue:43937", + "accept": false, + "reason": "Dequantized save failure and invalid GenerationConfig in GLM-5 are separate bugs." + }, + { + "left": "issue:43784", + "right": "issue:45600", + "accept": false, + "reason": "Sentence-transformers import NameError vs stale auto_mappings references to removed SAM3 configs; different root causes." + }, + { + "left": "issue:46172", + "right": "issue:46762", + "accept": false, + "reason": "Both are numerical/model correctness complaints, but they target different models and code paths." + }, + { + "left": "issue:44117", + "right": "issue:44704", + "accept": false, + "reason": "TOKENIZER_MAPPING_NAMES None-handling is not the same as AutoProcessor not forwarding kwargs to cached_file." + }, + { + "left": "issue:39401", + "right": "issue:44521", + "accept": false, + "reason": "Qwen3 tokenizer offset mapping and multimodal assistant_mask generation are different tokenizer/template bugs." + }, + { + "left": "issue:44625", + "right": "issue:44829", + "accept": false, + "reason": "Qwen3.5 num_labels propagation and flash_attention_3 degenerate training are unrelated." + }, + { + "left": "issue:44704", + "right": "issue:45290", + "accept": false, + "reason": "AutoProcessor kwargs forwarding and chat-template crashes with tool-call messages are separate code paths." + }, + { + "left": "issue:43122", + "right": "issue:44521", + "accept": false, + "reason": "Version-to-version tokenization differences are not the same bug as multimodal assistant_masks being all zeros." + }, + { + "left": "issue:43408", + "right": "issue:43475", + "accept": false, + "reason": "Both are SAM 3 Video related, but one is a model-type warning and the other is a missing attribute on encoder output; different defects." + }, + { + "left": "issue:44360", + "right": "issue:44485", + "accept": false, + "reason": "DSA ReLU discussion and GLM-5 RoPE implementation are not the same issue." + }, + { + "left": "issue:43122", + "right": "issue:46326", + "accept": false, + "reason": "General tokenization changes across versions are not the same as the mistral_common continue_final_message bug." + }, + { + "left": "issue:42907", + "right": "issue:43452", + "accept": false, + "reason": "Dequantized save failures and gguf_file breaking from_pretrained are different loading/saving paths." + }, + { + "left": "issue:44038", + "right": "issue:44297", + "accept": false, + "reason": "Qwen3-VL-Moe bug and tokenizer_class metadata mismatch are unrelated." + }, + { + "left": "issue:22355", + "right": "issue:29159", + "accept": false, + "reason": "ONNX missing-module import error and slow-vs-fast tokenizer inconsistency are unrelated." + }, + { + "left": "issue:45216", + "right": "issue:45478", + "accept": false, + "reason": "Both mention Qwen3.5 and 5.4.0+, but one is a bad saved checkpoint and the other is a from_pretrained error; not enough evidence for the same concrete bug." + }, + { + "left": "issue:44038", + "right": "issue:44295", + "accept": false, + "reason": "Qwen3-VL-Moe bug and position_ids buffer access bug do not share the same failure mode." + }, + { + "left": "issue:39401", + "right": "issue:46614", + "accept": false, + "reason": "Qwen3 tokenizer offset mapping and stale phonemizer language state are unrelated tokenizer bugs." + }, + { + "left": "issue:44038", + "right": "issue:44222", + "accept": false, + "reason": "Qwen3-VL-Moe loading issue and FP8 save_pretrained moe issue are separate model/save-path problems." + }, + { + "left": "issue:43824", + "right": "issue:44351", + "accept": false, + "reason": "Both are import errors, but they involve different missing symbols and likely different export/mapping problems." + } + ] + }, + "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": 26, + "review_unit_index": 51, + "review_unit_count": 61, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-505", + "nodes": [ + "issue:22355", + "issue:29127", + "issue:36296", + "issue:38617", + "issue:42175", + "issue:42371", + "issue:42673", + "issue:42757", + "issue:42898", + "issue:43066", + "issue:43116", + "issue:43122", + "issue:43278", + "issue:43493", + "issue:43611", + "issue:43618", + "issue:43874", + "issue:43937", + "issue:43976", + "issue:44038", + "issue:44075", + "issue:44188", + "issue:44351", + "issue:44355", + "issue:44485", + "issue:44610", + "issue:44625", + "issue:44779", + "issue:44936", + "issue:44938", + "issue:44945", + "issue:45072", + "issue:45083", + "issue:45200", + "issue:45278", + "issue:45357", + "issue:45446", + "issue:45478", + "issue:45753", + "issue:46002", + "issue:46231", + "issue:46328", + "issue:46614", + "issue:46726", + "issue:46829", + "issue:46877", + "issue:46909", + "issue:46981" + ], + "soft_pairs": [ + "issue:43116|issue:45200", + "issue:43122|issue:44779", + "issue:42757|issue:44351", + "issue:43278|issue:44945", + "issue:36296|issue:43116", + "issue:22355|issue:29127", + "issue:42175|issue:44938", + "issue:43493|issue:45753", + "issue:46231|issue:46877", + "issue:44188|issue:46328", + "issue:42898|issue:46614", + "issue:38617|issue:45278", + "issue:43937|issue:44485", + "issue:42673|issue:45072", + "issue:43066|issue:44779", + "issue:43611|issue:46002", + "issue:44038|issue:45478", + "issue:43976|issue:45446", + "issue:45357|issue:45478", + "issue:43116|issue:43493", + "issue:44355|issue:46231", + "issue:44625|issue:46002", + "issue:44610|issue:46909", + "issue:45083|issue:45446", + "issue:42371|issue:43874", + "issue:46726|issue:46981", + "issue:44075|issue:44936", + "issue:43618|issue:46829" + ], + "prepared_review_unit_hash": "68acbf5715a9611cb1a95b651d59e1f313caa3b71d83cd8b1fa1acd0aab14c80", + "component_budget": { + "node_count": 505, + "item_count": 505, + "soft_pair_count": 1854, + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 28, + "serialized_chars": 35725, + "estimated_input_tokens": 8932, + "estimated_eval_tokens": 18120 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is a loose collection of mostly unrelated Transformers issues with some superficial overlap around v5 regressions, imports, tokenization, and model-specific bugs. I did not find any soft-edge pair that is confidently the same underlying defect.", + "confidence": 0.57, + "canonical_issue_reason": "issue:45278 is the broadest and most umbrella-like report in the set, covering multiple import failures after upgrading to 5.5.0.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45278 is the best representative issue for this cluster because it is general, recent, and plausibly subsumes several import-related regression reports.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43116", + "right": "issue:45200", + "accept": false, + "reason": "Different bugs: example script multi-label output handling vs Gemma 4 text-only mm_token_type_ids defaulting." + }, + { + "left": "issue:43122", + "right": "issue:44779", + "accept": false, + "reason": "Both are tokenizer regressions, but they describe different models and different failure modes; not enough evidence of one underlying bug." + }, + { + "left": "issue:42757", + "right": "issue:44351", + "accept": false, + "reason": "Both are import errors, but for different symbols and different modules; likely separate breakages." + }, + { + "left": "issue:43278", + "right": "issue:44945", + "accept": false, + "reason": "BF16/FP32 dtype drift in evaluation is unrelated to incorrect LLM output under pipeline parallelism." + }, + { + "left": "issue:36296", + "right": "issue:43116", + "accept": false, + "reason": "Tensor-parallel training bug and multi-label classification output issue are unrelated." + }, + { + "left": "issue:22355", + "right": "issue:29127", + "accept": false, + "reason": "One is a missing module import, the other is a LayoutLMv3 error message clarity issue." + }, + { + "left": "issue:42175", + "right": "issue:44938", + "accept": false, + "reason": "Environment/backend packaging issue vs Python 3.14 load failure; different problems." + }, + { + "left": "issue:43493", + "right": "issue:45753", + "accept": false, + "reason": "SigLIP2 implementation discrepancy is not the same as Qwen3_5 infinite-loop behavior on a specific image." + }, + { + "left": "issue:46231", + "right": "issue:46877", + "accept": false, + "reason": "Documentation parameter mismatch vs docstring tooling printing diagnostics at import time; related area, but not the same bug." + }, + { + "left": "issue:44188", + "right": "issue:46328", + "accept": false, + "reason": "Attention-kernel divergence under torch.compile is unrelated to SequenceFeatureExtractor.pad performance." + }, + { + "left": "issue:42898", + "right": "issue:46614", + "accept": false, + "reason": "General tokenization-space behavior change and stale phonemizer backend state are different tokenizer issues." + }, + { + "left": "issue:38617", + "right": "issue:45278", + "accept": false, + "reason": "Specific missing import and broad import-error regression may be related in theme, but not the same concrete code-path problem." + }, + { + "left": "issue:43937", + "right": "issue:44485", + "accept": false, + "reason": "GenerationConfig validation failure is distinct from GLM-5 RoPE implementation discussion." + }, + { + "left": "issue:42673", + "right": "issue:45072", + "accept": false, + "reason": "VRAM leak in dataloader-threaded Qwen3ForCausalLM is unrelated to bfloat16 dtype mismatches in CI." + }, + { + "left": "issue:43066", + "right": "issue:44779", + "accept": false, + "reason": "Wrong decoder-type regression and Deepseek tokenizer incorrect output are too indirect to treat as the same bug." + }, + { + "left": "issue:43611", + "right": "issue:46002", + "accept": false, + "reason": "Base-model-prefix loading breakage is unrelated to NaNs in classification heads after checkout/init." + }, + { + "left": "issue:44038", + "right": "issue:45478", + "accept": false, + "reason": "Both mention Qwen Moe models, but they target different model variants and different load-time failures." + }, + { + "left": "issue:43976", + "right": "issue:45446", + "accept": false, + "reason": "Python version support regression is unrelated to the PyTorch version check for AuxRequest import." + }, + { + "left": "issue:45357", + "right": "issue:45478", + "accept": false, + "reason": "Saving incorrect visual encoder keys and from_pretrained load errors are different directions of failure with no clear shared code-path." + }, + { + "left": "issue:43116", + "right": "issue:43493", + "accept": false, + "reason": "Multi-label classification example behavior and SigLIP2 implementation mismatch are unrelated." + }, + { + "left": "issue:44355", + "right": "issue:46231", + "accept": false, + "reason": "Running compiled Python files errors and a documentation parameter name mismatch are not the same issue." + }, + { + "left": "issue:44625", + "right": "issue:46002", + "accept": false, + "reason": "num_labels propagation bug in Qwen3.5 is unrelated to NaNs in classification heads." + }, + { + "left": "issue:44610", + "right": "issue:46909", + "accept": false, + "reason": "Processor input-size mismatch and a broken documentation example are different kinds of defects." + }, + { + "left": "issue:45083", + "right": "issue:45446", + "accept": false, + "reason": "Feature-extractor length helper behavior and a version-check import bug are unrelated." + }, + { + "left": "issue:42371", + "right": "issue:43874", + "accept": false, + "reason": "TF32 API guidance and missing image-patch-count method are unrelated." + }, + { + "left": "issue:46726", + "right": "issue:46981", + "accept": false, + "reason": "Missing-vision-backend error handling and repeated 429 retries under HF_ENDPOINT are different AutoProcessor behaviors." + }, + { + "left": "issue:44075", + "right": "issue:44936", + "accept": false, + "reason": "Unused SGD args and trainer.evaluate() failing after train() are not the same bug." + }, + { + "left": "issue:43618", + "right": "issue:46829", + "accept": false, + "reason": "Both are output-field assignment problems, but they affect different model outputs and different fields; too weak to merge." + } + ] + }, + "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": 26, + "review_unit_index": 52, + "review_unit_count": 61, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-505", + "nodes": [ + "issue:28348", + "issue:33453", + "issue:39401", + "issue:42222", + "issue:42673", + "issue:42796", + "issue:43064", + "issue:43066", + "issue:43116", + "issue:43441", + "issue:43450", + "issue:43673", + "issue:43824", + "issue:43825", + "issue:43827", + "issue:43937", + "issue:43986", + "issue:43992", + "issue:44117", + "issue:44263", + "issue:44291", + "issue:44380", + "issue:44410", + "issue:44466", + "issue:44704", + "issue:44779", + "issue:44977", + "issue:44995", + "issue:45071", + "issue:45092", + "issue:45310", + "issue:45405", + "issue:45446", + "issue:45563", + "issue:45588", + "issue:46002", + "issue:46077", + "issue:46328", + "issue:46512", + "issue:46614", + "issue:46640", + "issue:46655", + "issue:46697", + "issue:46710", + "issue:46752", + "issue:46762", + "issue:46981", + "issue:47000" + ], + "soft_pairs": [ + "issue:43116|issue:43825", + "issue:43992|issue:44704", + "issue:43066|issue:46710", + "issue:46614|issue:46752", + "issue:44995|issue:46697", + "issue:28348|issue:46640", + "issue:42673|issue:42796", + "issue:44263|issue:46762", + "issue:45092|issue:46762", + "issue:43827|issue:45446", + "issue:33453|issue:39401", + "issue:43064|issue:43673", + "issue:43937|issue:45563", + "issue:43824|issue:46981", + "issue:42222|issue:43441", + "issue:42673|issue:46655", + "issue:42222|issue:47000", + "issue:43673|issue:46328", + "issue:45405|issue:45446", + "issue:44291|issue:45071", + "issue:42222|issue:46752", + "issue:44117|issue:45588", + "issue:44380|issue:46077", + "issue:44466|issue:46002", + "issue:39401|issue:44779", + "issue:43937|issue:44977", + "issue:44410|issue:45310", + "issue:43450|issue:43986", + "issue:45071|issue:46512" + ], + "prepared_review_unit_hash": "b08157c483c3ca8d759d5859e63185d82a5eef3bf0b410536255191c81d2ef03", + "component_budget": { + "node_count": 505, + "item_count": 505, + "soft_pair_count": 1854, + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 29, + "serialized_chars": 36278, + "estimated_input_tokens": 9070, + "estimated_eval_tokens": 18396 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "No soft pair looks like the same underlying bug or change. Most are only loosely related by subsystem (tokenizers, generation, flash-attn, v5 regressions) but point to distinct code paths or model-specific failures, so they should stay separate issues.", + "confidence": 0.87, + "canonical_issue_reason": "issue:33453 is the broadest umbrella-style report in the set ('Regression in tokenizer loading') and is the most plausible canonical representative among the issue-only cluster.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:33453 is the best global representative because it is broad, concrete, and names the common failure class without overfitting to a single model.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43116", + "right": "issue:43825", + "accept": false, + "reason": "Both touch example/pipeline behavior, but one is a multi-label classification script bug and the other is a v5 pipeline translation-message mismatch." + }, + { + "left": "issue:43992", + "right": "issue:44704", + "accept": false, + "reason": "Both concern loading APIs, but they fail in different layers: missing model weights vs AutoProcessor forwarding kwargs to cached_file." + }, + { + "left": "issue:43066", + "right": "issue:46710", + "accept": false, + "reason": "These are unrelated: tokenizer decoder type in v5 versus a DeepSeek accuracy regression after upgrading Transformers." + }, + { + "left": "issue:46614", + "right": "issue:46752", + "accept": false, + "reason": "Tokenizer backend state leakage and apply_chat_template crashing on an empty list are different components and failure modes." + }, + { + "left": "issue:44995", + "right": "issue:46697", + "accept": false, + "reason": "A stale cache crash in GlmMoeDsa is unrelated to Swin2SR image over-padding in the processor." + }, + { + "left": "issue:28348", + "right": "issue:46640", + "accept": false, + "reason": "One is a feature request for GPT2 FlashAttention 2.0 support; the other is a separate generic Flash Attention 2 issue." + }, + { + "left": "issue:42673", + "right": "issue:42796", + "accept": false, + "reason": "VRAM leakage in Qwen3ForCausalLM and tokenizer BPE marker handling in Ministral are different bugs." + }, + { + "left": "issue:44263", + "right": "issue:46762", + "accept": false, + "reason": "GlmMoeDsaIndexer split return values and Minimax MSA attention numerical differences are not the same bug." + }, + { + "left": "issue:45092", + "right": "issue:46762", + "accept": false, + "reason": "Remote-code checkpoint/meta-init incompatibility is unrelated to attention numerical drift." + }, + { + "left": "issue:43827", + "right": "issue:45446", + "accept": false, + "reason": "Docs still referencing pipeline() and a PyTorch version check bug in flex_attention are different issues." + }, + { + "left": "issue:33453", + "right": "issue:39401", + "accept": false, + "reason": "Both are tokenizer regressions, but one is broad loading regression while the other is a Qwen3 offset_mapping bug." + }, + { + "left": "issue:43064", + "right": "issue:43673", + "accept": false, + "reason": "Optimizer-state corruption under FSDP2/PEFT and missing GenerationMixin cache during chunked_prefill are unrelated training vs generation issues." + }, + { + "left": "issue:43937", + "right": "issue:45563", + "accept": false, + "reason": "An invalid GenerationConfig error and a stale warning for num_return_sequences are different generation-path problems." + }, + { + "left": "issue:43824", + "right": "issue:46981", + "accept": false, + "reason": "An import failure for a specific model class and repeated 429 retries in AutoProcessor are not the same underlying bug." + }, + { + "left": "issue:42222", + "right": "issue:43441", + "accept": false, + "reason": "Vitpose model breakage and Ministral FlashAttention failure are separate model-specific issues." + }, + { + "left": "issue:42673", + "right": "issue:46655", + "accept": false, + "reason": "A VRAM leak in model usage threads is unrelated to a video processor padding calculation bug." + }, + { + "left": "issue:42222", + "right": "issue:47000", + "accept": false, + "reason": "Vitpose breakage and RTDetrLoss infeasible cost matrix errors do not share a concrete code path." + }, + { + "left": "issue:43673", + "right": "issue:46328", + "accept": false, + "reason": "Missing cache in generation and inefficient numpy-to-list conversion in padding are different performance/correctness issues." + }, + { + "left": "issue:45405", + "right": "issue:45446", + "accept": false, + "reason": "A dependency version bump problem and a flex_attention import-version check bug are unrelated." + }, + { + "left": "issue:44291", + "right": "issue:45071", + "accept": false, + "reason": "Unexpected _is_hf_initialized argument handling and PretrainedConfig type-checking regressions are distinct v5 compatibility issues." + }, + { + "left": "issue:42222", + "right": "issue:46752", + "accept": false, + "reason": "Vitpose model breakage and empty-list chat template crashing are unrelated." + }, + { + "left": "issue:44117", + "right": "issue:45588", + "accept": false, + "reason": "Tokenizer mapping assumptions and flash_attention integration crashing on s_aux=None are different subsystems." + }, + { + "left": "issue:44380", + "right": "issue:46077", + "accept": false, + "reason": "GPT2 attention scaling being ignored and Gemma4 building causal masks are both attention-related but not the same bug." + }, + { + "left": "issue:44466", + "right": "issue:46002", + "accept": false, + "reason": "Serialization of lm_head.weight and NaNs in classification heads are unrelated training/serialization issues." + }, + { + "left": "issue:39401", + "right": "issue:44779", + "accept": false, + "reason": "Both are tokenizer correctness regressions, but they affect different models and behaviors: offset_mapping vs incorrect tokenization results." + }, + { + "left": "issue:43937", + "right": "issue:44977", + "accept": false, + "reason": "Invalid GenerationConfig handling and Qwen3.5 flash-attention generation failure are distinct generation problems." + }, + { + "left": "issue:44410", + "right": "issue:45310", + "accept": false, + "reason": "Missing projections in qwen3next layer 0 and a Qwen3.5 MoE from_pretrained error are different model-loading/model-architecture bugs." + }, + { + "left": "issue:43450", + "right": "issue:43986", + "accept": false, + "reason": "Incorrect batched video-processor output shape and an AutoProcessor crash without torchvision are separate video-processing issues." + }, + { + "left": "issue:45071", + "right": "issue:46512", + "accept": false, + "reason": "PretrainedConfig type checking breakage and stale model.config.dtype after from_config(dtype=...) are different v5 API 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": 26, + "review_unit_index": 53, + "review_unit_count": 61, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-505", + "nodes": [ + "issue:38617", + "issue:39401", + "issue:41628", + "issue:42222", + "issue:42371", + "issue:42617", + "issue:43066", + "issue:43097", + "issue:43122", + "issue:43208", + "issue:43408", + "issue:43673", + "issue:43742", + "issue:43937", + "issue:44075", + "issue:44077", + "issue:44117", + "issue:44220", + "issue:44261", + "issue:44263", + "issue:44410", + "issue:44484", + "issue:44485", + "issue:44493", + "issue:44496", + "issue:44829", + "issue:44877", + "issue:44933", + "issue:45081", + "issue:45216", + "issue:45278", + "issue:45310", + "issue:45341", + "issue:45362", + "issue:45478", + "issue:45561", + "issue:45632", + "issue:45656", + "issue:46178", + "issue:46209", + "issue:46231", + "issue:46298", + "issue:46328", + "issue:46617", + "issue:46762", + "issue:46894", + "issue:46981" + ], + "soft_pairs": [ + "issue:45341|issue:45561", + "issue:44829|issue:45216", + "issue:45341|issue:46328", + "issue:45632|issue:46981", + "issue:44117|issue:45081", + "issue:42617|issue:45341", + "issue:43408|issue:43742", + "issue:38617|issue:41628", + "issue:43122|issue:46298", + "issue:45362|issue:46328", + "issue:39401|issue:46209", + "issue:42222|issue:46178", + "issue:44877|issue:46231", + "issue:44484|issue:45310", + "issue:42371|issue:44261", + "issue:43937|issue:44496", + "issue:44933|issue:45278", + "issue:42371|issue:46617", + "issue:44261|issue:46762", + "issue:44410|issue:45478", + "issue:44075|issue:45656", + "issue:38617|issue:44220", + "issue:43066|issue:43122", + "issue:43208|issue:46894", + "issue:43097|issue:44493", + "issue:44263|issue:44485", + "issue:44077|issue:46981", + "issue:43673|issue:45341" + ], + "prepared_review_unit_hash": "cd9f3a58f16a4d9bb9d5bf0f812fbae2f639aecbcade3b57915289cfaa431456", + "component_budget": { + "node_count": 505, + "item_count": 505, + "soft_pair_count": 1854, + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 28, + "serialized_chars": 35144, + "estimated_input_tokens": 8786, + "estimated_eval_tokens": 17828 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly heterogeneous issue reports. The soft-similarity pairs share broad subsystems (tokenizers, imports, models, optimizers, config loading) but not the same concrete bug or fixable code path, so none should be merged as duplicates.", + "confidence": 0.33, + "canonical_issue_reason": "If a single issue must represent the cluster, issue 45278 is the broadest umbrella-style regression report (many import errors after upgrading) and is the closest to a cluster anchor.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 45278 is the most general and triage-friendly report in this set: it describes a broad post-upgrade import regression rather than a narrow model-specific symptom.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45341", + "right": "issue:45561", + "accept": false, + "reason": "Same file area, but different bug class: a small testing_utils bug vs a xdist race on captured_info.txt." + }, + { + "left": "issue:44829", + "right": "issue:45216", + "accept": false, + "reason": "Different problems: flash-attention training degeneration vs incorrect checkpoint saving for Qwen3.5." + }, + { + "left": "issue:45341", + "right": "issue:46328", + "accept": false, + "reason": "Unrelated: testing_utils vs a SequenceFeatureExtractor.pad performance issue." + }, + { + "left": "issue:45632", + "right": "issue:46981", + "accept": false, + "reason": "Both involve loading/caching, but one is cache-path collision and the other is repeated HTTP 429 retry behavior." + }, + { + "left": "issue:44117", + "right": "issue:45081", + "accept": false, + "reason": "Both are tokenizer-related, but one is a mapping lookup bug and the other is a Mistral regex patch crash." + }, + { + "left": "issue:42617", + "right": "issue:45341", + "accept": false, + "reason": "Unrelated: a 3d_parallel.py runtime failure vs a testing_utils bug." + }, + { + "left": "issue:43408", + "right": "issue:43742", + "accept": false, + "reason": "Different model-loading issues affecting different model families and failure modes." + }, + { + "left": "issue:38617", + "right": "issue:41628", + "accept": false, + "reason": "Both are import errors, but for different missing symbols and likely different removal paths." + }, + { + "left": "issue:43122", + "right": "issue:46298", + "accept": false, + "reason": "Different tokenization regression vs a noisy logging message; no shared bug." + }, + { + "left": "issue:45362", + "right": "issue:46328", + "accept": false, + "reason": "Completely different areas: chat generation crash vs feature-extractor padding performance." + }, + { + "left": "issue:39401", + "right": "issue:46209", + "accept": false, + "reason": "Tokenizer offset mapping bug vs FP8 inference failure on B200; unrelated." + }, + { + "left": "issue:42222", + "right": "issue:46178", + "accept": false, + "reason": "Different model family breakage vs unsupported backend mapping validation." + }, + { + "left": "issue:44877", + "right": "issue:46231", + "accept": false, + "reason": "Config-loading strictness vs a documentation parameter-name mismatch." + }, + { + "left": "issue:44484", + "right": "issue:45310", + "accept": false, + "reason": "Save-pretrained shard-size question vs Qwen3.5 from_pretrained error; unrelated." + }, + { + "left": "issue:42371", + "right": "issue:44261", + "accept": false, + "reason": "TF32 API behavior change vs a model precision/eps bug in MLA layernorm." + }, + { + "left": "issue:43937", + "right": "issue:44496", + "accept": false, + "reason": "Both are validation failures, but one is GenerationConfig invalidity and the other is missing model_type in config." + }, + { + "left": "issue:44933", + "right": "issue:45278", + "accept": false, + "reason": "A specific missing import vs a broad import-regression report; not the same concrete bug." + }, + { + "left": "issue:42371", + "right": "issue:46617", + "accept": false, + "reason": "TF32 API control message vs trust_remote_code cache FileNotFoundError; unrelated." + }, + { + "left": "issue:44261", + "right": "issue:46762", + "accept": false, + "reason": "Both are numerical/model-behavior issues, but they affect different models and code paths." + }, + { + "left": "issue:44410", + "right": "issue:45478", + "accept": false, + "reason": "Different model architectures and failure modes; no shared concrete fix." + }, + { + "left": "issue:44075", + "right": "issue:45656", + "accept": false, + "reason": "Both mention optimizer behavior, but one is unused SGD args and the other is double optimizer stepping under deepspeed." + }, + { + "left": "issue:38617", + "right": "issue:44220", + "accept": false, + "reason": "ImportError for a missing symbol is unrelated to an audio feature extraction issue." + }, + { + "left": "issue:43066", + "right": "issue:43122", + "accept": false, + "reason": "Both concern tokenizer regressions around v5, but the reported symptoms are different enough that they do not clearly share one concrete code-path bug." + }, + { + "left": "issue:43208", + "right": "issue:46894", + "accept": false, + "reason": "Different model/training bugs, even if both mention loss/training behavior." + }, + { + "left": "issue:43097", + "right": "issue:44493", + "accept": false, + "reason": "Removed tie_embeddings API vs unexpected position-id keys; unrelated." + }, + { + "left": "issue:44263", + "right": "issue:44485", + "accept": false, + "reason": "Different GLM-5 issues: torch.split return handling vs RoPE implementation." + }, + { + "left": "issue:44077", + "right": "issue:46981", + "accept": false, + "reason": "Post-init policy issue vs HTTP retry behavior; unrelated." + }, + { + "left": "issue:43673", + "right": "issue:45341", + "accept": false, + "reason": "Generation cache regression vs a testing_utils bug; different 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": 26, + "review_unit_index": 54, + "review_unit_count": 61, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-505", + "nodes": [ + "issue:27957", + "issue:38617", + "issue:41628", + "issue:42757", + "issue:42886", + "issue:43012", + "issue:43064", + "issue:43097", + "issue:43208", + "issue:43352", + "issue:43441", + "issue:43452", + "issue:43531", + "issue:43577", + "issue:43673", + "issue:43874", + "issue:43931", + "issue:43986", + "issue:44038", + "issue:44117", + "issue:44261", + "issue:44263", + "issue:44315", + "issue:44360", + "issue:44410", + "issue:44483", + "issue:44484", + "issue:44485", + "issue:44829", + "issue:44987", + "issue:45003", + "issue:45071", + "issue:45341", + "issue:45372", + "issue:45375", + "issue:45440", + "issue:45464", + "issue:45478", + "issue:45542", + "issue:45588", + "issue:45753", + "issue:46050", + "issue:46123", + "issue:46172", + "issue:46369", + "issue:46897", + "issue:46981" + ], + "soft_pairs": [ + "issue:44483|issue:45464", + "issue:43931|issue:44410", + "issue:41628|issue:45071", + "issue:45071|issue:45542", + "issue:44484|issue:45478", + "issue:43931|issue:44038", + "issue:43012|issue:43577", + "issue:43874|issue:44261", + "issue:43441|issue:44829", + "issue:27957|issue:43352", + "issue:44263|issue:46172", + "issue:44360|issue:46172", + "issue:43097|issue:43452", + "issue:43986|issue:45341", + "issue:45003|issue:45341", + "issue:42886|issue:46981", + "issue:44261|issue:44485", + "issue:45753|issue:46369", + "issue:44117|issue:44987", + "issue:43531|issue:45341", + "issue:43097|issue:45588", + "issue:43874|issue:44263", + "issue:43673|issue:46123", + "issue:45440|issue:46172", + "issue:44410|issue:45375", + "issue:43986|issue:45372", + "issue:43097|issue:46050", + "issue:44038|issue:44410", + "issue:38617|issue:42757", + "issue:43064|issue:44315", + "issue:43986|issue:44484", + "issue:43208|issue:46897" + ], + "prepared_review_unit_hash": "169be7f8ba93925432a7b9e9755811edb2ccecd544d7cfaa00b65018a89ce3ed", + "component_budget": { + "node_count": 505, + "item_count": 505, + "soft_pair_count": 1854, + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 32, + "serialized_chars": 36567, + "estimated_input_tokens": 9142, + "estimated_eval_tokens": 18540 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster does not look like a true duplicate set; it spans many unrelated Transformers/HF bugs across imports, model loading, FlashAttention, training, tokenization, and infra behavior. I rejected all soft edges.", + "confidence": 0.93, + "canonical_issue_reason": "No clear canonical issue: the items are a heterogeneous mix of unrelated bugs, so there is no single duplicate target that the rest collapse into.", + "canonical_pr_reason": null, + "best_issue_reason": "No single issue is a strong global representative; any choice would be arbitrary because the cluster lacks one underlying problem.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44483", + "right": "issue:45464", + "accept": false, + "reason": "Both mention chat/completions failures, but one is a generic API complaint and the other is a Qwen3.5 streaming issue; not the same bug." + }, + { + "left": "issue:43931", + "right": "issue:44410", + "accept": false, + "reason": "Different model families and different failure modes: weight-shape mismatch vs missing projections." + }, + { + "left": "issue:41628", + "right": "issue:45071", + "accept": false, + "reason": "Unrelated import error vs config type-checking regression." + }, + { + "left": "issue:45071", + "right": "issue:45542", + "accept": false, + "reason": "Different subsystems: config typing vs TensorBoard/TF backend handling." + }, + { + "left": "issue:44484", + "right": "issue:45478", + "accept": false, + "reason": "A save_pretrained shard-size question is unrelated to a Qwen3.5-MoE from_pretrained failure." + }, + { + "left": "issue:43931", + "right": "issue:44038", + "accept": false, + "reason": "Both are Qwen3-VL-related, but one is a shape mismatch and the other is a broader v5/MoE load bug." + }, + { + "left": "issue:43012", + "right": "issue:43577", + "accept": false, + "reason": "Different dtype issues: compile-time precision warning vs model loading leaving dtype at float32." + }, + { + "left": "issue:43874", + "right": "issue:44261", + "accept": false, + "reason": "Different models and code paths: multimodal token counting vs MLA layernorm config precision." + }, + { + "left": "issue:43441", + "right": "issue:44829", + "accept": false, + "reason": "Both involve FlashAttention, but one is a model-support/load failure and the other is training degeneration under flash_attention_3." + }, + { + "left": "issue:27957", + "right": "issue:43352", + "accept": false, + "reason": "Unsupported FlashAttention 2 on different models is not enough to treat these as the same underlying bug." + }, + { + "left": "issue:44263", + "right": "issue:46172", + "accept": false, + "reason": "Different DeepSeek/GLM implementation bugs with no shared concrete code path." + }, + { + "left": "issue:44360", + "right": "issue:46172", + "accept": false, + "reason": "One is a missing ReLU in a DSA indexer; the other is a tensor dtype/transpose issue." + }, + { + "left": "issue:43097", + "right": "issue:43452", + "accept": false, + "reason": "Removed API semantics vs gguf_file/from_pretrained breakage are unrelated." + }, + { + "left": "issue:43986", + "right": "issue:45341", + "accept": false, + "reason": "Video AutoProcessor import/runtime crash is unrelated to a testing_utils bug." + }, + { + "left": "issue:45003", + "right": "issue:45341", + "accept": false, + "reason": "Unsafe sys.modules access is a different problem from the testing_utils issue." + }, + { + "left": "issue:42886", + "right": "issue:46981", + "accept": false, + "reason": "Both are hub/network related, but offline cache loading and HTTP 429 retry behavior are distinct bugs." + }, + { + "left": "issue:44261", + "right": "issue:44485", + "accept": false, + "reason": "RMS norm epsilon precision discussion is unrelated to GLM-5 RoPE implementation." + }, + { + "left": "issue:45753", + "right": "issue:46369", + "accept": false, + "reason": "An infinite-loop image bug and a memory overflow are different failure classes." + }, + { + "left": "issue:44117", + "right": "issue:44987", + "accept": false, + "reason": "Tokenizer mapping None handling and a specific model load failure are not clearly the same code-path bug." + }, + { + "left": "issue:43531", + "right": "issue:45341", + "accept": false, + "reason": "Qwen3-MoE sliding-window behavior is unrelated to testing_utils." + }, + { + "left": "issue:43097", + "right": "issue:45588", + "accept": false, + "reason": "A removed config/helper API and a flash_attention s_aux=None crash are different changes." + }, + { + "left": "issue:43874", + "right": "issue:44263", + "accept": false, + "reason": "Different GLM46V multimodal-token bug vs GLM MoE indexer split behavior." + }, + { + "left": "issue:43673", + "right": "issue:46123", + "accept": false, + "reason": "Generation cache handling and mask generation batching are different pipeline bugs." + }, + { + "left": "issue:45440", + "right": "issue:46172", + "accept": false, + "reason": "Both are DeepSeek-related, but they describe different divergences and different concrete code paths." + }, + { + "left": "issue:44410", + "right": "issue:45375", + "accept": false, + "reason": "Qwen3next missing projections and missing vision-config fields are distinct model-definition issues." + }, + { + "left": "issue:43986", + "right": "issue:45372", + "accept": false, + "reason": "AutoProcessor without torchvision and a mistral_common import error are unrelated." + }, + { + "left": "issue:43097", + "right": "issue:46050", + "accept": false, + "reason": "API removal and quantized-model save/load corruption are not the same underlying bug." + }, + { + "left": "issue:44038", + "right": "issue:44410", + "accept": false, + "reason": "Different Qwen variants and different missing-attribute/model-structure problems." + }, + { + "left": "issue:38617", + "right": "issue:42757", + "accept": false, + "reason": "Separate import errors from different modules/packages." + }, + { + "left": "issue:43064", + "right": "issue:44315", + "accept": false, + "reason": "Optimizer-state corruption under FSDP2/PEFT is unrelated to Liger Kernel not being applied." + }, + { + "left": "issue:43986", + "right": "issue:44484", + "accept": false, + "reason": "No shared bug: video processor crash vs a shard-size configuration question." + }, + { + "left": "issue:43208", + "right": "issue:46897", + "accept": false, + "reason": "Both are training bugs, but they affect different models with different concrete symptoms 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": 26, + "review_unit_index": 55, + "review_unit_count": 61, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-505", + "nodes": [ + "issue:38617", + "issue:42890", + "issue:42907", + "issue:43012", + "issue:43066", + "issue:43097", + "issue:43208", + "issue:43352", + "issue:43502", + "issue:43550", + "issue:43825", + "issue:43873", + "issue:43906", + "issue:43976", + "issue:43986", + "issue:43992", + "issue:44016", + "issue:44062", + "issue:44077", + "issue:44117", + "issue:44162", + "issue:44222", + "issue:44351", + "issue:44368", + "issue:44380", + "issue:44484", + "issue:44655", + "issue:44704", + "issue:44936", + "issue:44938", + "issue:45103", + "issue:45478", + "issue:45584", + "issue:45712", + "issue:45753", + "issue:46132", + "issue:46172", + "issue:46178", + "issue:46298", + "issue:46365", + "issue:46421", + "issue:46655", + "issue:46664", + "issue:46701", + "issue:46710", + "issue:46752", + "issue:46821" + ], + "soft_pairs": [ + "issue:46298|issue:46752", + "issue:46655|issue:46664", + "issue:42890|issue:46298", + "issue:44062|issue:44117", + "issue:43097|issue:43873", + "issue:38617|issue:44351", + "issue:43012|issue:46821", + "issue:46172|issue:46710", + "issue:44117|issue:46701", + "issue:43502|issue:43992", + "issue:44936|issue:46365", + "issue:44704|issue:45103", + "issue:44380|issue:46421", + "issue:44938|issue:46178", + "issue:43986|issue:44016", + "issue:42907|issue:44655", + "issue:42907|issue:45753", + "issue:43012|issue:43550", + "issue:45712|issue:46298", + "issue:43012|issue:44368", + "issue:43208|issue:43550", + "issue:43825|issue:44655", + "issue:43992|issue:44484", + "issue:43352|issue:43976", + "issue:43352|issue:44380", + "issue:43066|issue:43906", + "issue:43097|issue:46298", + "issue:45584|issue:46365", + "issue:44162|issue:46172", + "issue:44380|issue:46132", + "issue:44077|issue:46365", + "issue:44222|issue:45478" + ], + "prepared_review_unit_hash": "5e1831300dfa895fc78e474dd7b225282c60882248f9a83e8eb570e853d3fc2e", + "component_budget": { + "node_count": 505, + "item_count": 505, + "soft_pair_count": 1854, + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 32, + "serialized_chars": 36469, + "estimated_input_tokens": 9118, + "estimated_eval_tokens": 18492 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The cluster is heterogeneous: the soft-similarity pairs mostly share broad Transformers subsystem keywords, but they point to different bugs, regressions, or features. I did not find any pair that clearly describes the same underlying issue/change, so no soft edges are accepted.", + "confidence": 0.94, + "canonical_issue_reason": "No single canonical issue stands out because the items are about different code paths and failure modes rather than one underlying defect.", + "canonical_pr_reason": null, + "best_issue_reason": "No issue is a strong cluster representative; the set is too mixed across unrelated bugs, so a single best issue would be misleading.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:46298", + "right": "issue:46752", + "accept": false, + "reason": "Different problems: noisy logging vs an empty-list crash in apply_chat_template." + }, + { + "left": "issue:46655", + "right": "issue:46664", + "accept": false, + "reason": "Both are SmolVLM video-processor bugs, but one is padding math and the other is num_frames semantics." + }, + { + "left": "issue:42890", + "right": "issue:46298", + "accept": false, + "reason": "Test seed flakiness is unrelated to a logging message." + }, + { + "left": "issue:44062", + "right": "issue:44117", + "accept": false, + "reason": "One is an AddedToken keyword-argument error; the other is a tokenizer mapping None-handling bug." + }, + { + "left": "issue:43097", + "right": "issue:43873", + "accept": false, + "reason": "Token/embedding config removal is unrelated to quantization offloading behavior." + }, + { + "left": "issue:38617", + "right": "issue:44351", + "accept": false, + "reason": "Both are import errors, but for different missing symbols and likely different root causes." + }, + { + "left": "issue:43012", + "right": "issue:46821", + "accept": false, + "reason": "A PyTorch precision warning is unrelated to a TimesFM AttributeError." + }, + { + "left": "issue:46172", + "right": "issue:46710", + "accept": false, + "reason": "DeepSeek model code failure and a separate accuracy regression are not the same bug." + }, + { + "left": "issue:44117", + "right": "issue:46701", + "accept": false, + "reason": "Tokenizer mapping fallback logic is unrelated to the DiffusionGemma image-text pipeline failure." + }, + { + "left": "issue:43502", + "right": "issue:43992", + "accept": false, + "reason": "Local-files-only network leakage is unrelated to a missing embed_tokens weight on load." + }, + { + "left": "issue:44936", + "right": "issue:46365", + "accept": false, + "reason": "Trainer evaluate/train state bug is unrelated to an import-time datasets conflict." + }, + { + "left": "issue:44704", + "right": "issue:45103", + "accept": false, + "reason": "Passing kwargs to cached_file is unrelated to the auto_docstring annotation crash." + }, + { + "left": "issue:44380", + "right": "issue:46421", + "accept": false, + "reason": "GPT-2 attention scaling under SDPA/FlashAttention is unrelated to chunked-prefill recompilation." + }, + { + "left": "issue:44938", + "right": "issue:46178", + "accept": false, + "reason": "Python 3.14 import failure and a BACKENDS_MAPPING validation error are different compatibility issues." + }, + { + "left": "issue:43986", + "right": "issue:44016", + "accept": false, + "reason": "AutoProcessor without torchvision and a notebook syntax error are unrelated." + }, + { + "left": "issue:42907", + "right": "issue:44655", + "accept": false, + "reason": "Saving dequantized models and saving Pipeline objects are different save_pretrained problems." + }, + { + "left": "issue:42907", + "right": "issue:45753", + "accept": false, + "reason": "A save failure for dequantized models is unrelated to an image-specific infinite loop." + }, + { + "left": "issue:43012", + "right": "issue:43550", + "accept": false, + "reason": "A bfloat16 precision warning is unrelated to a torch.compile SDPA failure." + }, + { + "left": "issue:45712", + "right": "issue:46298", + "accept": false, + "reason": "Dummy object leakage in repo checks is unrelated to a logging message." + }, + { + "left": "issue:43012", + "right": "issue:44368", + "accept": false, + "reason": "These are distinct warning cases with different triggers and model/config paths." + }, + { + "left": "issue:43208", + "right": "issue:43550", + "accept": false, + "reason": "xLSTM training blockers are unrelated to the Bamba SDPA compile issue." + }, + { + "left": "issue:43825", + "right": "issue:44655", + "accept": false, + "reason": "A pipeline task-support error message is unrelated to Pipeline save_pretrained failures." + }, + { + "left": "issue:43992", + "right": "issue:44484", + "accept": false, + "reason": "Missing weight loading and the 50GB shard-size question are unrelated save_pretrained topics." + }, + { + "left": "issue:43352", + "right": "issue:43976", + "accept": false, + "reason": "A model-specific Flash Attention limitation and a Python version compatibility issue are different problems." + }, + { + "left": "issue:43352", + "right": "issue:44380", + "accept": false, + "reason": "Nemotron Flash Attention support and GPT-2 attention scaling are unrelated model bugs." + }, + { + "left": "issue:43066", + "right": "issue:43906", + "accept": false, + "reason": "Tokenizer decoder type mismatch and an isolated reproduction of another issue are not the same bug." + }, + { + "left": "issue:43097", + "right": "issue:46298", + "accept": false, + "reason": "Removed tie_embeddings behavior and a noisy log message are unrelated." + }, + { + "left": "issue:45584", + "right": "issue:46365", + "accept": false, + "reason": "Whisper empty-transcription generation failure is unrelated to Trainer import conflicts." + }, + { + "left": "issue:44162", + "right": "issue:46172", + "accept": false, + "reason": "ESM2 breakage and a DeepSeek tensor operation bug are unrelated." + }, + { + "left": "issue:44380", + "right": "issue:46132", + "accept": false, + "reason": "GPT-2 backend scaling bug is unrelated to AttentionInterface.register changing function behavior." + }, + { + "left": "issue:44222", + "right": "issue:45478", + "accept": false, + "reason": "FP8 save_pretrained behavior and a MoE from_pretrained regression are related only by model family, not by the same underlying bug." + } + ] + }, + "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": 26, + "review_unit_index": 56, + "review_unit_count": 61, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-505", + "nodes": [ + "issue:29870", + "issue:36010", + "issue:36296", + "issue:36683", + "issue:38617", + "issue:42175", + "issue:42490", + "issue:42757", + "issue:42994", + "issue:43012", + "issue:43064", + "issue:43208", + "issue:43299", + "issue:43352", + "issue:43475", + "issue:43502", + "issue:43525", + "issue:43526", + "issue:43582", + "issue:43611", + "issue:43618", + "issue:43756", + "issue:43761", + "issue:43824", + "issue:43867", + "issue:43937", + "issue:43976", + "issue:44038", + "issue:44077", + "issue:44230", + "issue:44355", + "issue:44380", + "issue:44483", + "issue:44485", + "issue:44492", + "issue:44514", + "issue:44625", + "issue:44655", + "issue:44938", + "issue:45216", + "issue:45276", + "issue:45290", + "issue:45479", + "issue:46050", + "issue:46077", + "issue:46082", + "issue:46365", + "issue:46877" + ], + "soft_pairs": [ + "issue:36683|issue:43582", + "issue:44380|issue:45479", + "issue:42175|issue:46365", + "issue:42757|issue:43824", + "issue:43064|issue:44380", + "issue:42490|issue:43012", + "issue:44483|issue:44514", + "issue:43618|issue:44380", + "issue:38617|issue:43502", + "issue:43208|issue:43526", + "issue:44038|issue:44230", + "issue:43618|issue:43761", + "issue:43756|issue:44485", + "issue:42994|issue:44655", + "issue:36010|issue:42757", + "issue:45276|issue:46077", + "issue:43299|issue:43867", + "issue:44625|issue:46365", + "issue:44483|issue:45290", + "issue:43611|issue:44655", + "issue:36683|issue:43475", + "issue:43352|issue:43937", + "issue:44355|issue:46877", + "issue:44655|issue:46050", + "issue:44625|issue:46077", + "issue:36683|issue:44077", + "issue:43976|issue:44938", + "issue:43867|issue:45216", + "issue:46077|issue:46082", + "issue:43618|issue:45216", + "issue:36683|issue:43525", + "issue:36683|issue:44625", + "issue:29870|issue:36296", + "issue:43064|issue:44492" + ], + "prepared_review_unit_hash": "b86b86580e6d3af60f1608d285a9dc43869536bcfa3128aaa44778d030061e72", + "component_budget": { + "node_count": 505, + "item_count": 505, + "soft_pair_count": 1854, + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 34, + "serialized_chars": 37684, + "estimated_input_tokens": 9421, + "estimated_eval_tokens": 19098 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The cluster is highly heterogeneous: most pairs are only superficially related by broad themes like import errors, model loading/saving, or model-family regressions, but they do not describe the same underlying bug or change. No soft edge looks strong enough to merge.", + "confidence": 0.96, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:36683", + "right": "issue:43582", + "accept": false, + "reason": "Both are runtime/configuration failures, but one is Gemma3 missing vocab_size and the other is an Apple Silicon TypeError in allocator warmup; unrelated bugs." + }, + { + "left": "issue:44380", + "right": "issue:45479", + "accept": false, + "reason": "Both involve training/inference behavior, but one is GPT2 attention scaling being ignored in SDPA/FlashAttention and the other is zero loss for single-label classification; different code paths." + }, + { + "left": "issue:42175", + "right": "issue:46365", + "accept": false, + "reason": "Both mention import trouble, but one is a packaging/backend dependency issue and the other is Trainer import conflict with datasets; not the same problem." + }, + { + "left": "issue:42757", + "right": "issue:43824", + "accept": false, + "reason": "Both are ImportErrors, but for different symbols and different causes: huggingface_hub is_offline_mode vs transformers Qwen2_5_VLForConditionalGeneration." + }, + { + "left": "issue:43064", + "right": "issue:44380", + "accept": false, + "reason": "One is an FSDP2/PEFT optimizer-state mismatch, the other is attention-scaling config ignored in backend kernels; unrelated regressions." + }, + { + "left": "issue:42490", + "right": "issue:43012", + "accept": false, + "reason": "A local path trailing-slash import warning is not the same as a bfloat16 compilation precision warning." + }, + { + "left": "issue:44483", + "right": "issue:44514", + "accept": false, + "reason": "Both touch chat/template flows, but one is an API /v1/chat/completions acceptance issue and the other is a batched apply_chat_template crash with padding=False." + }, + { + "left": "issue:43618", + "right": "issue:44380", + "accept": false, + "reason": "CLIPOutput attentions missing is a model output regression; GPT2 attention scaling ignored is a backend-specific config bug." + }, + { + "left": "issue:38617", + "right": "issue:43502", + "accept": false, + "reason": "Importing layer_type_validation and local_files_only still making API requests are different subsystems and failure modes." + }, + { + "left": "issue:43208", + "right": "issue:43526", + "accept": false, + "reason": "xLSTM training bugs and BeitImageProcessorFast reduce_labels returning one label are unrelated model/processor issues." + }, + { + "left": "issue:44038", + "right": "issue:44230", + "accept": false, + "reason": "Both are Qwen3-VL/Qwen3.5-related, but one is a loading regression in transformers 5.0 and the other is FP8 inference support; not the same fix." + }, + { + "left": "issue:43618", + "right": "issue:43761", + "accept": false, + "reason": "Both are CLIP vision regressions, but one loses attentions and the other loses hidden_states under output_hidden_states=True; distinct outputs and code paths." + }, + { + "left": "issue:43756", + "right": "issue:44485", + "accept": false, + "reason": "Smollm3 RoPE-layer loss and GLM-5 RoPE implementation discussion are different models and different issues." + }, + { + "left": "issue:42994", + "right": "issue:44655", + "accept": false, + "reason": "Quantized model saving failure and Pipeline save_pretrained failure are both serialization-related, but one targets quantized models and the other Pipeline objects." + }, + { + "left": "issue:36010", + "right": "issue:42757", + "accept": false, + "reason": "GenerationMixin import failure and is_offline_mode import failure are unrelated ImportErrors." + }, + { + "left": "issue:45276", + "right": "issue:46077", + "accept": false, + "reason": "Both are Gemma4-related, but one is resize_token_embeddings propagation and the other is bidirectional-attention mask construction; different bugs." + }, + { + "left": "issue:43299", + "right": "issue:43867", + "accept": false, + "reason": "One is Qwen3VL MoE loading breakage and the other is a state_dict sorting load error; not the same underlying loading bug." + }, + { + "left": "issue:44625", + "right": "issue:46365", + "accept": false, + "reason": "Qwen3.5 num_labels propagation and Trainer import conflict with datasets are unrelated." + }, + { + "left": "issue:44483", + "right": "issue:45290", + "accept": false, + "reason": "Both involve chat/template flows, but one is endpoint request validation and the other is tool-call assistant message handling in tokenize=True." + }, + { + "left": "issue:43611", + "right": "issue:44655", + "accept": false, + "reason": "base_model_prefix loading regression and Pipeline save_pretrained failure are different serialization/load paths." + }, + { + "left": "issue:36683", + "right": "issue:43475", + "accept": false, + "reason": "Gemma3 missing vocab_size and SAM 3 Video missing fpn_position_embeddings are both attribute errors, but on different models and fields." + }, + { + "left": "issue:43352", + "right": "issue:43937", + "accept": false, + "reason": "Nemotron FlashAttention2 support and GLM-5 GenerationConfig validity are unrelated model-specific issues." + }, + { + "left": "issue:44355", + "right": "issue:46877", + "accept": false, + "reason": "Compiled Python file errors and auto_docstring printing diagnostics at import time are separate import/runtime concerns." + }, + { + "left": "issue:44655", + "right": "issue:46050", + "accept": false, + "reason": "Pipeline save_pretrained and sinq quantized model round-trip save/load are both serialization issues, but not the same object or code path." + }, + { + "left": "issue:44625", + "right": "issue:46077", + "accept": false, + "reason": "Qwen3.5 num_labels propagation and Gemma4 bidirectional attention masks are unrelated." + }, + { + "left": "issue:36683", + "right": "issue:44077", + "accept": false, + "reason": "Gemma3 missing vocab_size and patchtsmixer post_init allowance are unrelated config/model-definition issues." + }, + { + "left": "issue:43976", + "right": "issue:44938", + "accept": false, + "reason": "Both are Python-compatibility reports, but one is about 5.1.0 on Python 3.9+ and the other is loading failure on Python 3.14; different compatibility problems." + }, + { + "left": "issue:43867", + "right": "issue:45216", + "accept": false, + "reason": "A sorted state_dict load error and a broken Qwen3.5 save_pretrained checkpoint are both save/load regressions, but they are distinct failures." + }, + { + "left": "issue:46077", + "right": "issue:46082", + "accept": false, + "reason": "Gemma4 causal-attention mask behavior and LlamaConfig head_dim validation are unrelated model/config bugs." + }, + { + "left": "issue:43618", + "right": "issue:45216", + "accept": false, + "reason": "CLIPOutput attentions missing and Qwen3.5 checkpoint corruption after save_pretrained are unrelated regressions." + }, + { + "left": "issue:36683", + "right": "issue:43525", + "accept": false, + "reason": "Gemma3 missing vocab_size and Llama4 missing pad_token_id are similar in shape but refer to different model configs and separate fixes." + }, + { + "left": "issue:36683", + "right": "issue:44625", + "accept": false, + "reason": "Gemma3 vocab_size missing and Qwen3.5 num_labels propagation are unrelated attribute/config issues." + }, + { + "left": "issue:29870", + "right": "issue:36296", + "accept": false, + "reason": "BART generate decoder_attention_mask UnboundLocalError and tensor-parallel training bug are different execution paths and failures." + }, + { + "left": "issue:43064", + "right": "issue:44492", + "accept": false, + "reason": "FSDP2 optimizer-state corruption and a cache-strategies typo have nothing in common beyond both being in Transformers." + } + ] + }, + "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": 26, + "review_unit_index": 57, + "review_unit_count": 61, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-505", + "nodes": [ + "issue:28282", + "issue:29870", + "issue:33453", + "issue:36296", + "issue:36683", + "issue:41669", + "issue:42890", + "issue:42898", + "issue:43010", + "issue:43064", + "issue:43502", + "issue:43575", + "issue:43618", + "issue:43723", + "issue:43867", + "issue:43976", + "issue:43992", + "issue:44062", + "issue:44075", + "issue:44077", + "issue:44112", + "issue:44246", + "issue:44355", + "issue:44360", + "issue:44393", + "issue:44556", + "issue:44734", + "issue:44855", + "issue:44908", + "issue:44938", + "issue:44964", + "issue:45071", + "issue:45278", + "issue:45310", + "issue:45357", + "issue:45375", + "issue:45440", + "issue:45478", + "issue:45561", + "issue:45600", + "issue:45850", + "issue:46202", + "issue:46285", + "issue:46365", + "issue:46617", + "issue:46664", + "issue:46851", + "issue:47000" + ], + "soft_pairs": [ + "issue:43010|issue:45278", + "issue:43992|issue:45310", + "issue:44075|issue:46365", + "issue:44908|issue:46851", + "issue:44393|issue:47000", + "issue:29870|issue:33453", + "issue:44938|issue:45071", + "issue:44355|issue:44855", + "issue:43867|issue:46202", + "issue:36683|issue:43064", + "issue:43867|issue:45375", + "issue:42890|issue:45561", + "issue:41669|issue:44246", + "issue:43010|issue:45850", + "issue:36296|issue:43575", + "issue:43618|issue:44077", + "issue:43010|issue:43976", + "issue:43867|issue:45357", + "issue:43992|issue:45478", + "issue:28282|issue:33453", + "issue:43618|issue:46285", + "issue:44360|issue:45440", + "issue:42890|issue:44964", + "issue:42890|issue:44112", + "issue:42890|issue:45600", + "issue:43502|issue:44062", + "issue:42898|issue:46664", + "issue:44734|issue:46617", + "issue:43723|issue:44556" + ], + "prepared_review_unit_hash": "55710c92ba50fa9b0aa88057d6b85f56a7e9dd73bdf95825bf5eb6c7047a5aa4", + "component_budget": { + "node_count": 505, + "item_count": 505, + "soft_pair_count": 1854, + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 29, + "serialized_chars": 36053, + "estimated_input_tokens": 9014, + "estimated_eval_tokens": 18284 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly a grab bag of unrelated Transformers regressions: model-loading/import failures, version-compatibility breakages, training bugs, tokenizer/processor issues, and a few CI/test flakiness reports. None of the soft-edge pairs look like the same concrete bug or change under strict duplicate triage.", + "confidence": 0.9, + "canonical_issue_reason": "issue:45278 is the broadest and most representative report of the cluster\u2019s dominant theme: post-upgrade import/load regressions in v5.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45278 is the best cluster anchor because it captures the shared high-level regression class without being overly model-specific.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43010", + "right": "issue:45278", + "accept": false, + "reason": "Different bug class: cache/layer update semantics vs broad import regressions after upgrade." + }, + { + "left": "issue:43992", + "right": "issue:45310", + "accept": false, + "reason": "Different models and different from_pretrained failures; no shared concrete code path." + }, + { + "left": "issue:44075", + "right": "issue:46365", + "accept": false, + "reason": "Optimizer argument handling is unrelated to a Trainer import/dependency issue." + }, + { + "left": "issue:44908", + "right": "issue:46851", + "accept": false, + "reason": "Scheduler kwargs bug vs Qwen3.5 variable-length position_ids issue; unrelated." + }, + { + "left": "issue:44393", + "right": "issue:47000", + "accept": false, + "reason": "Different tasks and failures: VLM bbox output vs RTDetr loss infeasibility." + }, + { + "left": "issue:29870", + "right": "issue:33453", + "accept": false, + "reason": "BART generate decoder mask bug vs tokenizer-loading regression; distinct code paths." + }, + { + "left": "issue:44938", + "right": "issue:45071", + "accept": false, + "reason": "Python-version load failure vs PretrainedConfig type-checking break; not the same defect." + }, + { + "left": "issue:44355", + "right": "issue:44855", + "accept": false, + "reason": "Both are import-related, but one is compiled-file behavior and the other is a torch.jit parsing failure on 3.13." + }, + { + "left": "issue:43867", + "right": "issue:46202", + "accept": false, + "reason": "State_dict sorting load error vs Gemma pipeline roundtrip issue; unrelated." + }, + { + "left": "issue:36683", + "right": "issue:43064", + "accept": false, + "reason": "Gemma3 config missing vocab_size is unrelated to FSDP2/PEFT optimizer-state corruption." + }, + { + "left": "issue:43867", + "right": "issue:45375", + "accept": false, + "reason": "A load error and a config-field serialization issue are different problems." + }, + { + "left": "issue:42890", + "right": "issue:45561", + "accept": false, + "reason": "Seed-related flaky tests vs xdist file race; different test failures." + }, + { + "left": "issue:41669", + "right": "issue:44246", + "accept": false, + "reason": "Import-star performance regression vs intermittent slow import; too broad to be the same bug." + }, + { + "left": "issue:43010", + "right": "issue:45850", + "accept": false, + "reason": "Specific no_grad/update bug vs vague \"many models break\" report; not the same issue." + }, + { + "left": "issue:36296", + "right": "issue:43575", + "accept": false, + "reason": "Tensor-parallel training bug vs TP OOM on one model; different concrete failures." + }, + { + "left": "issue:43618", + "right": "issue:44077", + "accept": false, + "reason": "CLIP attentions assignment bug vs patchtsmixer post_init policy change; unrelated." + }, + { + "left": "issue:43010", + "right": "issue:43976", + "accept": false, + "reason": "Cache update bug vs Python compatibility/load failure; unrelated." + }, + { + "left": "issue:43867", + "right": "issue:45357", + "accept": false, + "reason": "Model load error vs save_pretrained key mismatch; opposite directions and different causes." + }, + { + "left": "issue:43992", + "right": "issue:45478", + "accept": false, + "reason": "Both are from_pretrained complaints, but they involve different model families and different failures." + }, + { + "left": "issue:28282", + "right": "issue:33453", + "accept": false, + "reason": "Missing PyTorch installation error vs tokenizer-loading regression; unrelated." + }, + { + "left": "issue:43618", + "right": "issue:46285", + "accept": false, + "reason": "Both CLIP-related, but one is missing attentions assignment and the other is class-structure inconsistency." + }, + { + "left": "issue:44360", + "right": "issue:45440", + "accept": false, + "reason": "DSA indexer/ReLU discussion is unrelated to DeepseekV3MoE implementation divergence." + }, + { + "left": "issue:42890", + "right": "issue:44964", + "accept": false, + "reason": "SAM-HQ test flakiness vs Phi-4 multimodal load failure; unrelated." + }, + { + "left": "issue:42890", + "right": "issue:44112", + "accept": false, + "reason": "Seed-related test instability vs GraniteSpeech device-override CI issue; different bugs." + }, + { + "left": "issue:42890", + "right": "issue:45600", + "accept": false, + "reason": "Seed/test flakiness vs auto_mappings referencing removed configs; unrelated." + }, + { + "left": "issue:43502", + "right": "issue:44062", + "accept": false, + "reason": "local_files_only network leakage vs AddedToken keyword collision; different code paths." + }, + { + "left": "issue:42898", + "right": "issue:46664", + "accept": false, + "reason": "Tokenization cleanup behavior in v5 vs video processor num_frames semantics; unrelated." + }, + { + "left": "issue:44734", + "right": "issue:46617", + "accept": false, + "reason": "KV-cache continuation crash vs cache loading FileNotFoundError; different runtime paths." + }, + { + "left": "issue:43723", + "right": "issue:44556", + "accept": false, + "reason": "Tokenizer loading regression vs checkpoint reload compatibility issue; not the same underlying bug." + } + ] + }, + "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": 26, + "review_unit_index": 58, + "review_unit_count": 61, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-505", + "nodes": [ + "issue:29870", + "issue:30333", + "issue:30990", + "issue:31795", + "issue:33453", + "issue:34689", + "issue:36296", + "issue:38617", + "issue:41628", + "issue:42548", + "issue:42757", + "issue:43010", + "issue:43065", + "issue:43232", + "issue:43299", + "issue:43352", + "issue:43475", + "issue:43502", + "issue:43576", + "issue:43824", + "issue:43827", + "issue:43874", + "issue:43901", + "issue:44222", + "issue:44230", + "issue:44315", + "issue:44351", + "issue:44355", + "issue:44360", + "issue:44387", + "issue:44485", + "issue:44556", + "issue:44749", + "issue:44938", + "issue:44961", + "issue:45071", + "issue:45125", + "issue:45375", + "issue:45468", + "issue:45478", + "issue:46132", + "issue:46143", + "issue:46144", + "issue:46209", + "issue:46617", + "issue:46664", + "issue:46726" + ], + "soft_pairs": [ + "issue:44222|issue:45125", + "issue:43824|issue:46617", + "issue:43502|issue:44485", + "issue:43874|issue:46617", + "issue:43010|issue:43065", + "issue:44387|issue:44556", + "issue:43576|issue:46132", + "issue:45125|issue:45478", + "issue:30333|issue:33453", + "issue:44355|issue:44749", + "issue:42548|issue:44351", + "issue:45125|issue:45375", + "issue:30990|issue:31795", + "issue:43502|issue:44222", + "issue:43502|issue:44961", + "issue:43010|issue:43232", + "issue:46664|issue:46726", + "issue:43827|issue:44961", + "issue:30333|issue:36296", + "issue:29870|issue:31795", + "issue:43299|issue:44556", + "issue:33453|issue:34689", + "issue:44355|issue:45071", + "issue:43576|issue:45071", + "issue:42757|issue:43576", + "issue:44315|issue:44360", + "issue:43576|issue:43824", + "issue:31795|issue:38617", + "issue:44961|issue:45468", + "issue:44485|issue:44961", + "issue:44351|issue:46617", + "issue:36296|issue:43901", + "issue:43576|issue:44938", + "issue:43352|issue:44961", + "issue:43475|issue:46617", + "issue:44222|issue:44230", + "issue:46143|issue:46144", + "issue:41628|issue:42548", + "issue:44315|issue:46209" + ], + "prepared_review_unit_hash": "b495925f766b53c42935fe42d6be0e0875eda9dd3b3724fe396a4ba236666774", + "component_budget": { + "node_count": 505, + "item_count": 505, + "soft_pair_count": 1854, + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 39, + "serialized_chars": 38291, + "estimated_input_tokens": 9573, + "estimated_eval_tokens": 19402 + }, + "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 issues, with a weaker central theme around v5 loading/import/regression breakages. None of the soft pairs look like true duplicates or the same concrete fix path, so all should stay separate.", + "confidence": 0.84, + "canonical_issue_reason": "issue:43576 is the most central and broadly representative item: a v5 regression affecting the environment/CLI surface, which aligns with several other load/import/version-breakage reports in the cluster.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43576 is the best single representative issue because it captures the broadest recurring theme in the cluster (v5 regression / breakage) without being overly narrow to one model or one symbol import.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44222", + "right": "issue:45125", + "accept": false, + "reason": "Both involve Qwen/MoE/FP8-related problems, but one is about save_pretrained behavior and the other about tensor-parallel planning/loading; not the same bug." + }, + { + "left": "issue:43824", + "right": "issue:46617", + "accept": false, + "reason": "One is an import error for a specific model class, the other is a FileNotFoundError loading a trust_remote_code model from cache; different failure modes and code paths." + }, + { + "left": "issue:43502", + "right": "issue:44485", + "accept": false, + "reason": "Local-files-only network leakage and GLM-5 RoPE implementation are unrelated issues." + }, + { + "left": "issue:43874", + "right": "issue:46617", + "accept": false, + "reason": "Missing image-patch helper in a processor and cache FileNotFoundError are separate problems." + }, + { + "left": "issue:43010", + "right": "issue:43065", + "accept": false, + "reason": "Cache update no_grad decoration and a dummy Conv2d in Sam3PixelDecoder are unrelated implementation bugs." + }, + { + "left": "issue:44387", + "right": "issue:44556", + "accept": false, + "reason": "Int4 quantization memory growth and reloading checkpoints after upgrade are different regression types." + }, + { + "left": "issue:43576", + "right": "issue:46132", + "accept": false, + "reason": "A broken env command in v5 and changed behavior of AttentionInterface.register are not the same concrete bug." + }, + { + "left": "issue:45125", + "right": "issue:45478", + "accept": false, + "reason": "Both mention Qwen3.5 MoE, but one is missing _tp_plan for tensor parallelism while the other is a generic from_pretrained error; insufficient evidence of one root cause." + }, + { + "left": "issue:30333", + "right": "issue:33453", + "accept": false, + "reason": "MLFlow reporting on failed jobs and tokenizer loading regression are unrelated." + }, + { + "left": "issue:44355", + "right": "issue:44749", + "accept": false, + "reason": "Compiled Python file errors and a tokenizer performance regression do not describe the same bug." + }, + { + "left": "issue:42548", + "right": "issue:44351", + "accept": false, + "reason": "Both are import errors, but for different symbols and likely different missing exports." + }, + { + "left": "issue:45125", + "right": "issue:45375", + "accept": false, + "reason": "Missing tensor-parallel plan and missing config field silently dropped by strict validation are different issues." + }, + { + "left": "issue:30990", + "right": "issue:31795", + "accept": false, + "reason": "Loading gets stuck and confusing forward-docs wording are not the same underlying problem." + }, + { + "left": "issue:43502", + "right": "issue:44222", + "accept": false, + "reason": "Network access despite local_files_only and FP8 moe save_pretrained behavior are unrelated." + }, + { + "left": "issue:43502", + "right": "issue:44961", + "accept": false, + "reason": "API requests with local_files_only and the placeholder title 'racoon' have no relation." + }, + { + "left": "issue:43010", + "right": "issue:43232", + "accept": false, + "reason": "Both touch generation internals, but one is about no_grad on cache updates and the other about sync_gpus/model kwargs ordering; not the same concrete fix." + }, + { + "left": "issue:46664", + "right": "issue:46726", + "accept": false, + "reason": "A video num_frames semantic mismatch and an AutoProcessor vision-backend import error are different bugs." + }, + { + "left": "issue:43827", + "right": "issue:44961", + "accept": false, + "reason": "Docs still referencing pipeline after removals is unrelated to the placeholder issue." + }, + { + "left": "issue:30333", + "right": "issue:36296", + "accept": false, + "reason": "MLFlow failure reporting and tensor parallel training bug are unrelated." + }, + { + "left": "issue:29870", + "right": "issue:31795", + "accept": false, + "reason": "BART generate() UnboundLocalError and docs confusion around input_ids/past_key_values do not share the same code-path bug." + }, + { + "left": "issue:43299", + "right": "issue:44556", + "accept": false, + "reason": "Qwen3VL MoE loading regression and general checkpoint reload failure after upgrade are related only at a high level, not the same underlying defect." + }, + { + "left": "issue:33453", + "right": "issue:34689", + "accept": false, + "reason": "Tokenizer loading regression and Llama 3.2 Vision model loading breakage are separate model-loading issues." + }, + { + "left": "issue:44355", + "right": "issue:45071", + "accept": false, + "reason": "Compiled Python-file failures and PretrainedConfig type-checking breakage are unrelated." + }, + { + "left": "issue:43576", + "right": "issue:45071", + "accept": false, + "reason": "A broken env command and PretrainedConfig type-checking issues are different regression surfaces." + }, + { + "left": "issue:42757", + "right": "issue:43576", + "accept": false, + "reason": "Missing huggingface_hub import and a broken transformers env command are not the same bug." + }, + { + "left": "issue:44315", + "right": "issue:44360", + "accept": false, + "reason": "Liger kernel not applied with model_init and a missing ReLU in the DSA indexer are unrelated." + }, + { + "left": "issue:43576", + "right": "issue:43824", + "accept": false, + "reason": "The env command regression and a specific model-class import error are different failures." + }, + { + "left": "issue:31795", + "right": "issue:38617", + "accept": false, + "reason": "Docs confusion in model.forward and missing layer_type_validation import are unrelated." + }, + { + "left": "issue:44961", + "right": "issue:45468", + "accept": false, + "reason": "Placeholder issue 'racoon' is unrelated to the Gemma-4 positional encoding bug." + }, + { + "left": "issue:44485", + "right": "issue:44961", + "accept": false, + "reason": "GLM-5 RoPE discussion and the placeholder issue are unrelated." + }, + { + "left": "issue:44351", + "right": "issue:46617", + "accept": false, + "reason": "Importing HybridCache and FileNotFoundError when loading from cache are different root causes." + }, + { + "left": "issue:36296", + "right": "issue:43901", + "accept": false, + "reason": "Tensor parallel training bug and TextClassificationPipeline docs mismatch are unrelated." + }, + { + "left": "issue:43576", + "right": "issue:44938", + "accept": false, + "reason": "Broken env command in v5 and Python 3.14 load failure are different version-related regressions, but not the same concrete issue." + }, + { + "left": "issue:43352", + "right": "issue:44961", + "accept": false, + "reason": "Flash Attention 2.0 support for Nemotron and the placeholder issue are unrelated." + }, + { + "left": "issue:43475", + "right": "issue:46617", + "accept": false, + "reason": "SAM 3 Video attribute error and trust_remote_code cache FileNotFoundError are different code paths." + }, + { + "left": "issue:44222", + "right": "issue:44230", + "accept": false, + "reason": "Both mention FP8/MoE, but one is about save_pretrained and the other about inference support; not mergeable into one fix." + }, + { + "left": "issue:46143", + "right": "issue:46144", + "accept": false, + "reason": "Missing kwargs propagation and attention interface usage are closely related conceptually, but the titles describe different implementation gaps and not one concrete duplicate bug." + }, + { + "left": "issue:41628", + "right": "issue:42548", + "accept": false, + "reason": "Both are import errors, but for different top-level symbols and likely separate export regressions." + }, + { + "left": "issue:44315", + "right": "issue:46209", + "accept": false, + "reason": "Liger kernel application and FP8 inference on B200 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": 26, + "review_unit_index": 59, + "review_unit_count": 61, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-505", + "nodes": [ + "issue:31795", + "issue:33453", + "issue:41669", + "issue:42371", + "issue:42890", + "issue:43352", + "issue:43408", + "issue:43519", + "issue:43525", + "issue:43572", + "issue:43576", + "issue:43673", + "issue:43976", + "issue:44038", + "issue:44077", + "issue:44230", + "issue:44246", + "issue:44273", + "issue:44295", + "issue:44297", + "issue:44315", + "issue:44351", + "issue:44360", + "issue:44393", + "issue:44485", + "issue:44492", + "issue:44556", + "issue:44623", + "issue:44704", + "issue:44829", + "issue:44861", + "issue:44908", + "issue:45003", + "issue:45071", + "issue:45092", + "issue:45125", + "issue:45230", + "issue:45278", + "issue:45341", + "issue:45362", + "issue:45542", + "issue:45753", + "issue:45850", + "issue:46095", + "issue:46218", + "issue:46328", + "issue:46655", + "issue:46664" + ], + "soft_pairs": [ + "issue:45362|issue:45850", + "issue:43673|issue:44492", + "issue:44485|issue:46664", + "issue:44273|issue:45542", + "issue:44623|issue:45230", + "issue:44038|issue:45125", + "issue:44273|issue:45003", + "issue:44360|issue:46664", + "issue:44393|issue:45071", + "issue:41669|issue:43673", + "issue:44393|issue:45753", + "issue:43576|issue:44556", + "issue:43525|issue:45850", + "issue:43572|issue:45850", + "issue:44556|issue:45092", + "issue:44492|issue:45850", + "issue:45278|issue:45850", + "issue:42371|issue:44492", + "issue:41669|issue:46218", + "issue:44230|issue:46095", + "issue:43408|issue:44315", + "issue:44246|issue:44273", + "issue:42890|issue:44908", + "issue:43576|issue:44908", + "issue:43352|issue:44393", + "issue:43976|issue:44393", + "issue:44908|issue:45071", + "issue:44295|issue:45230", + "issue:44297|issue:45230", + "issue:41669|issue:44351", + "issue:44492|issue:46328", + "issue:44704|issue:44908", + "issue:41669|issue:45341", + "issue:43976|issue:45850", + "issue:43976|issue:46218", + "issue:44077|issue:44861", + "issue:31795|issue:33453", + "issue:43576|issue:44273", + "issue:44315|issue:44829", + "issue:43519|issue:46655" + ], + "prepared_review_unit_hash": "6ec0fa1ff31965d04d3dafdc5db88468eb4ca475c3676af91b2c4a58bcc906d2", + "component_budget": { + "node_count": 505, + "item_count": 505, + "soft_pair_count": 1854, + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 40, + "serialized_chars": 39091, + "estimated_input_tokens": 9773, + "estimated_eval_tokens": 19802 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is highly heterogeneous: it mixes documentation issues, model-specific bugs, loading/regression reports, and version-upgrade complaints. The proposed soft pairs are mostly false positives rather than true duplicates of the same underlying bug or change.", + "confidence": 0.89, + "canonical_issue_reason": "Issue 45850 is the closest thing to a cluster anchor because it is the broadest regression complaint ('Since 5.0 version it breaks too many models'). Even so, the set is too diverse for a strong true-canonical duplicate.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 45850 is the best single representative of the cluster because it captures the widest v5 regression theme and has the most umbrella-like framing, but it should not be treated as a real duplicate target for the others.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45362", + "right": "issue:45850", + "accept": false, + "reason": "Different concrete problems: Qwen3.5 chat crash vs a broad v5 breakage complaint." + }, + { + "left": "issue:43673", + "right": "issue:44492", + "accept": false, + "reason": "Chunked-prefill cache regression and a cache-strategy typo are unrelated." + }, + { + "left": "issue:44485", + "right": "issue:46664", + "accept": false, + "reason": "GLM-5 RoPE implementation and SmolVLM video-frame padding are unrelated." + }, + { + "left": "issue:44273", + "right": "issue:45542", + "accept": false, + "reason": "Lazy loading bug vs TensorFlow/tensorboard backend error; different code paths." + }, + { + "left": "issue:44623", + "right": "issue:45230", + "accept": false, + "reason": "Specific processor.save_pretrained file bug vs an unspecified generic bug report." + }, + { + "left": "issue:44038", + "right": "issue:45125", + "accept": false, + "reason": "Qwen3-VL-Moe fp8 issue vs missing tensor-parallel plan; different functionality." + }, + { + "left": "issue:44273", + "right": "issue:45003", + "accept": false, + "reason": "Both involve loading/imports, but one is lazy loading and the other is unsafe sys.modules access." + }, + { + "left": "issue:44360", + "right": "issue:46664", + "accept": false, + "reason": "DSA indexer/ReLU discussion is unrelated to SmolVLM video processing." + }, + { + "left": "issue:44393", + "right": "issue:45071", + "accept": false, + "reason": "Qwen3-VL bbox output bug and PretrainedConfig type checking are unrelated." + }, + { + "left": "issue:41669", + "right": "issue:43673", + "accept": false, + "reason": "Import-* performance issue is not the same as the chunked-prefill cache regression." + }, + { + "left": "issue:44393", + "right": "issue:45753", + "accept": false, + "reason": "Both are Qwen3-VL-family issues, but one is bbox output and the other is an infinite loop on a specific image." + }, + { + "left": "issue:43576", + "right": "issue:44556", + "accept": false, + "reason": "Broken env command and checkpoint reload incompatibility are different v5 regressions." + }, + { + "left": "issue:43525", + "right": "issue:45850", + "accept": false, + "reason": "Missing pad_token_id in Llama4Config is a model-specific config bug, not the broad v5 regression." + }, + { + "left": "issue:43572", + "right": "issue:45850", + "accept": false, + "reason": "StableLm pad_token_idx regression is specific; it is not the same issue as the broad breakage report." + }, + { + "left": "issue:44556", + "right": "issue:45092", + "accept": false, + "reason": "Both are upgrade/load compatibility complaints, but they affect different checkpoint-loading paths and failure modes." + }, + { + "left": "issue:44492", + "right": "issue:45850", + "accept": false, + "reason": "Cache-strategy typo is unrelated to the general v5 breakage complaint." + }, + { + "left": "issue:45278", + "right": "issue:45850", + "accept": false, + "reason": "Broad import errors after upgrade is related in theme, but not the same concrete underlying bug." + }, + { + "left": "issue:42371", + "right": "issue:44492", + "accept": false, + "reason": "TF32 API/settings guidance and cache-strategy typo are unrelated." + }, + { + "left": "issue:41669", + "right": "issue:46218", + "accept": false, + "reason": "Import-* slowdown and missing Qwen3_5 module import are different import problems." + }, + { + "left": "issue:44230", + "right": "issue:46095", + "accept": false, + "reason": "FP8 inference support request and weight-conversion question are unrelated." + }, + { + "left": "issue:43408", + "right": "issue:44315", + "accept": false, + "reason": "Model-type mismatch warning and Liger kernel application are different code paths." + }, + { + "left": "issue:44246", + "right": "issue:44273", + "accept": false, + "reason": "Import slowdowns and lazy-loading malfunction overlap loosely, but not the same underlying bug." + }, + { + "left": "issue:42890", + "right": "issue:44908", + "accept": false, + "reason": "Test flakiness from missing seeds is unrelated to inverse_sqrt scheduler kwargs handling." + }, + { + "left": "issue:43576", + "right": "issue:44908", + "accept": false, + "reason": "Env-command breakage and scheduler kwargs regression are unrelated." + }, + { + "left": "issue:43352", + "right": "issue:44393", + "accept": false, + "reason": "Flash Attention support limitation and Qwen3-VL bbox hallucination are different model issues." + }, + { + "left": "issue:43976", + "right": "issue:44393", + "accept": false, + "reason": "Python version compatibility and Qwen3-VL output errors are unrelated." + }, + { + "left": "issue:44908", + "right": "issue:45071", + "accept": false, + "reason": "Scheduler kwargs bug and PretrainedConfig type-checking breakage are unrelated." + }, + { + "left": "issue:44295", + "right": "issue:45230", + "accept": false, + "reason": "position_ids buffer access bug is unrelated to a generic bug report." + }, + { + "left": "issue:44297", + "right": "issue:45230", + "accept": false, + "reason": "Tokenizer save_pretrained metadata mismatch is unrelated to a generic bug report." + }, + { + "left": "issue:41669", + "right": "issue:44351", + "accept": false, + "reason": "Import-* slowdown and HybridCache import failure are different issues." + }, + { + "left": "issue:44492", + "right": "issue:46328", + "accept": false, + "reason": "Cache-strategy typo and SequenceFeatureExtractor.pad conversion inefficiency are unrelated." + }, + { + "left": "issue:44704", + "right": "issue:44908", + "accept": false, + "reason": "Both mention kwargs, but they affect different APIs and different behaviors." + }, + { + "left": "issue:41669", + "right": "issue:45341", + "accept": false, + "reason": "Import-* performance issue and testing_utils bug are unrelated." + }, + { + "left": "issue:43976", + "right": "issue:45850", + "accept": false, + "reason": "Python 3.9/3.10 compatibility issue is not the same as broad v5 model breakage." + }, + { + "left": "issue:43976", + "right": "issue:46218", + "accept": false, + "reason": "Python-version compatibility and missing Qwen3_5 import are unrelated." + }, + { + "left": "issue:44077", + "right": "issue:44861", + "accept": false, + "reason": "Optional post_init policy change and tied-weights AttributeError are unrelated." + }, + { + "left": "issue:31795", + "right": "issue:33453", + "accept": false, + "reason": "Documentation confusion in model.forward and tokenizer loading regression are unrelated." + }, + { + "left": "issue:43576", + "right": "issue:44273", + "accept": false, + "reason": "Broken env command and lazy loading malfunction are both infrastructure-ish but not the same bug." + }, + { + "left": "issue:44315", + "right": "issue:44829", + "accept": false, + "reason": "Liger kernel application and flash_attention_3 training degeneration are unrelated." + }, + { + "left": "issue:43519", + "right": "issue:46655", + "accept": false, + "reason": "Qwen3VL timestamp calculation and SmolVLM video padding are separate model-specific 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": 26, + "review_unit_index": 60, + "review_unit_count": 61, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-505", + "nodes": [ + "issue:28348", + "issue:29870", + "issue:30333", + "issue:31515", + "issue:31795", + "issue:34689", + "issue:36296", + "issue:38617", + "issue:41669", + "issue:42371", + "issue:42913", + "issue:43066", + "issue:43298", + "issue:43352", + "issue:43502", + "issue:43519", + "issue:43525", + "issue:43572", + "issue:43576", + "issue:43606", + "issue:43618", + "issue:43704", + "issue:43717", + "issue:44230", + "issue:44263", + "issue:44273", + "issue:44351", + "issue:44360", + "issue:44485", + "issue:44492", + "issue:44623", + "issue:44655", + "issue:44861", + "issue:44908", + "issue:44938", + "issue:44998", + "issue:45230", + "issue:45250", + "issue:45290", + "issue:45310", + "issue:45468", + "issue:45850", + "issue:46018", + "issue:46095", + "issue:46144", + "issue:46172", + "issue:46218", + "issue:46894" + ], + "soft_pairs": [ + "issue:44360|issue:45468", + "issue:44273|issue:44908", + "issue:44655|issue:44861", + "issue:45230|issue:45310", + "issue:45468|issue:46894", + "issue:43576|issue:46218", + "issue:43519|issue:44485", + "issue:43519|issue:44908", + "issue:41669|issue:43704", + "issue:31795|issue:34689", + "issue:43704|issue:44485", + "issue:44485|issue:45468", + "issue:42371|issue:43704", + "issue:46018|issue:46172", + "issue:43502|issue:43519", + "issue:45850|issue:46218", + "issue:43519|issue:44623", + "issue:44263|issue:45468", + "issue:43606|issue:45468", + "issue:43298|issue:46095", + "issue:43519|issue:43704", + "issue:44351|issue:46218", + "issue:43352|issue:46218", + "issue:42913|issue:43066", + "issue:28348|issue:45250", + "issue:44908|issue:44998", + "issue:44492|issue:44998", + "issue:43618|issue:44998", + "issue:44998|issue:45850", + "issue:44938|issue:44998", + "issue:44861|issue:45290", + "issue:31515|issue:38617", + "issue:43572|issue:44861", + "issue:43525|issue:44861", + "issue:44230|issue:44623", + "issue:28348|issue:43352", + "issue:29870|issue:31515", + "issue:30333|issue:31515", + "issue:31515|issue:36296", + "issue:31515|issue:31795", + "issue:44230|issue:45310", + "issue:43298|issue:44230", + "issue:43717|issue:46144" + ], + "prepared_review_unit_hash": "4d429559e184a1b64544459b7f57992fdbd5d6ff8d7cd18d8ed6cdcea2f6aaa1", + "component_budget": { + "node_count": 505, + "item_count": 505, + "soft_pair_count": 1854, + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 43, + "serialized_chars": 39626, + "estimated_input_tokens": 9907, + "estimated_eval_tokens": 20070 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly heterogeneous: the soft pairs share broad vocabulary like model bugs, loading, or tokenizer changes, but they do not look like the same underlying issue. I would not merge any of the proposed soft pairs.", + "confidence": 0.92, + "canonical_issue_reason": "No true duplicate hub is visible here; the items span unrelated bugs and feature requests. If a representative anchor is required, issue 45468 is the most concrete model-specific bug report among the set, but it should not be treated as a duplicate parent.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 45468 is the best single representative only because it is a specific, technically described bug with multiple nearby-but-not-duplicate neighbors; it is not a good canonical duplicate target for the whole cluster.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44360", + "right": "issue:45468", + "accept": false, + "reason": "Different models and different failure modes; only superficial NLP overlap around indexing/encoding." + }, + { + "left": "issue:44273", + "right": "issue:44908", + "accept": false, + "reason": "Lazy loading regression vs scheduler argument handling; unrelated code paths." + }, + { + "left": "issue:44655", + "right": "issue:44861", + "accept": false, + "reason": "Saving Pipeline objects is unrelated to tied-weight key handling." + }, + { + "left": "issue:45230", + "right": "issue:45310", + "accept": false, + "reason": "A generic bug report vs a specific Qwen3.5 MoE from_pretrained error; not the same issue." + }, + { + "left": "issue:45468", + "right": "issue:46894", + "accept": false, + "reason": "Different models and different bugs; one is a positional encoding issue, the other is a training-loss shift problem." + }, + { + "left": "issue:43576", + "right": "issue:46218", + "accept": false, + "reason": "Broken env command vs import failure for a specific model module; unrelated." + }, + { + "left": "issue:43519", + "right": "issue:44485", + "accept": false, + "reason": "Timestamp calculation in a processor vs RoPE implementation in GLM-5; different components." + }, + { + "left": "issue:43519", + "right": "issue:44908", + "accept": false, + "reason": "Processor timestamp math has nothing to do with scheduler kwargs propagation." + }, + { + "left": "issue:41669", + "right": "issue:43704", + "accept": false, + "reason": "Import star cleanup/performance issue is unrelated to a VRAM leak in Qwen3ForCausalLM." + }, + { + "left": "issue:31795", + "right": "issue:34689", + "accept": false, + "reason": "Documentation confusion vs a model loading breakage in Llama 3.2 vision; not the same bug." + }, + { + "left": "issue:43704", + "right": "issue:44485", + "accept": false, + "reason": "VRAM leak in one model and RoPE implementation in another are distinct." + }, + { + "left": "issue:44485", + "right": "issue:45468", + "accept": false, + "reason": "Both are model internals, but they concern different architectures and different fixes." + }, + { + "left": "issue:42371", + "right": "issue:43704", + "accept": false, + "reason": "TF32 configuration guidance is unrelated to a model VRAM leak." + }, + { + "left": "issue:46018", + "right": "issue:46172", + "accept": false, + "reason": "Shared-expert gating bug vs a tensor transpose/casting issue in DeepSeek-V4; not the same concrete defect." + }, + { + "left": "issue:43502", + "right": "issue:43519", + "accept": false, + "reason": "Local-files-only network requests and timestamp computation are unrelated." + }, + { + "left": "issue:45850", + "right": "issue:46218", + "accept": false, + "reason": "Broad complaint about 5.0 breakages vs a specific import error for one model." + }, + { + "left": "issue:43519", + "right": "issue:44623", + "accept": false, + "reason": "Processor timestamp logic and missing save_pretrained files are different problems." + }, + { + "left": "issue:44263", + "right": "issue:45468", + "accept": false, + "reason": "A torch.split/indexing issue is not the same as Gemma-4 audio positional encoding." + }, + { + "left": "issue:43606", + "right": "issue:45468", + "accept": false, + "reason": "CPU offload device mismatch vs positional encoding bug; unrelated." + }, + { + "left": "issue:43298", + "right": "issue:46095", + "accept": false, + "reason": "Continuous batching audio support and DeepSeek weight conversion are unrelated feature requests." + }, + { + "left": "issue:43519", + "right": "issue:43704", + "accept": false, + "reason": "Processor timestamp bug and VRAM leak are different code paths." + }, + { + "left": "issue:44351", + "right": "issue:46218", + "accept": false, + "reason": "Different import errors for different symbols/models." + }, + { + "left": "issue:43352", + "right": "issue:46218", + "accept": false, + "reason": "Flash Attention 2.0 support request for NemotronH vs Qwen module import failure; not the same issue." + }, + { + "left": "issue:42913", + "right": "issue:43066", + "accept": false, + "reason": "Both involve tokenizer/v5 changes, but one is a general behavior regression and the other is a specific decoder-type bug." + }, + { + "left": "issue:28348", + "right": "issue:45250", + "accept": false, + "reason": "GPT2-specific FA2 support request is not clearly the same as a broader Flash Attention 2.0 topic." + }, + { + "left": "issue:44908", + "right": "issue:44998", + "accept": false, + "reason": "Scheduler kwargs bug vs unrelated placeholder issue." + }, + { + "left": "issue:44492", + "right": "issue:44998", + "accept": false, + "reason": "Cache strategy typo is unrelated to the placeholder issue." + }, + { + "left": "issue:43618", + "right": "issue:44998", + "accept": false, + "reason": "CLIPOutput attentions regression is unrelated to the placeholder issue." + }, + { + "left": "issue:44998", + "right": "issue:45850", + "accept": false, + "reason": "Placeholder issue has no technical overlap with broad 5.0 breakage complaints." + }, + { + "left": "issue:44938", + "right": "issue:44998", + "accept": false, + "reason": "Python 3.14 load failure is unrelated to the placeholder issue." + }, + { + "left": "issue:44861", + "right": "issue:45290", + "accept": false, + "reason": "Tied-weight key handling and chat template tool-call crash are different bugs." + }, + { + "left": "issue:31515", + "right": "issue:38617", + "accept": false, + "reason": "Slow from_pretrained loading vs an ImportError; different failure modes." + }, + { + "left": "issue:43572", + "right": "issue:44861", + "accept": false, + "reason": "Missing pad_token_idx after v5 update is unrelated to tied-weight key crashes." + }, + { + "left": "issue:43525", + "right": "issue:44861", + "accept": false, + "reason": "Llama4Config pad_token_id attribute error is unrelated to tied-weight key handling." + }, + { + "left": "issue:44230", + "right": "issue:44623", + "accept": false, + "reason": "FP8 support request and processor.save_pretrained missing files are unrelated." + }, + { + "left": "issue:28348", + "right": "issue:43352", + "accept": false, + "reason": "Both mention Flash Attention 2.0, but they are different model-specific support requests and not the same fix." + }, + { + "left": "issue:29870", + "right": "issue:31515", + "accept": false, + "reason": "BART generate() decoder mask bug is unrelated to slow checkpoint loading." + }, + { + "left": "issue:30333", + "right": "issue:31515", + "accept": false, + "reason": "MLflow finalization bug and slow from_pretrained loading are unrelated." + }, + { + "left": "issue:31515", + "right": "issue:36296", + "accept": false, + "reason": "Slow loading vs tensor-parallel training bug; no shared code path." + }, + { + "left": "issue:31515", + "right": "issue:31795", + "accept": false, + "reason": "Performance/loading issue vs documentation confusion." + }, + { + "left": "issue:44230", + "right": "issue:45310", + "accept": false, + "reason": "FP8 inference support request is not the same as a Qwen3.5 MoE from_pretrained error." + }, + { + "left": "issue:43298", + "right": "issue:44230", + "accept": false, + "reason": "Audio continuous batching support and FP8 support are different feature requests." + }, + { + "left": "issue:43717", + "right": "issue:46144", + "accept": false, + "reason": "Mamba init_weights behavior and RoFormer attention interface usage are unrelated implementation issues." + } + ] + }, + "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": 26, + "review_unit_index": 61, + "review_unit_count": 61, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-505", + "nodes": [ + "issue:37934", + "issue:42936", + "issue:43240", + "issue:43668", + "issue:43813", + "issue:44031", + "issue:46018", + "issue:46144" + ], + "soft_pairs": [ + "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:43668|issue:43813", + "issue:43240|issue:43668", + "issue:37934|issue:43668", + "issue:42936|issue:43668" + ], + "prepared_review_unit_hash": "b9df463b070c48627378376d8cd559d177435c8eb9d04716716eb2abf127f25d", + "component_budget": { + "node_count": 505, + "item_count": 505, + "soft_pair_count": 1854, + "serialized_chars": 962358, + "estimated_input_tokens": 240590, + "estimated_eval_tokens": 481436 + }, + "budget": { + "node_count": 8, + "item_count": 8, + "soft_pair_count": 11, + "serialized_chars": 7775, + "estimated_input_tokens": 1944, + "estimated_eval_tokens": 4144 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The items are not one duplicate set; they span unrelated bugs across different models and components. The soft pairings only show weak textual similarity and do not describe the same concrete defect or change.", + "confidence": 0.95, + "canonical_issue_reason": "No single canonical issue: the cluster contains unrelated reports (norm config, tokenizer warnings, loss kwargs, typo, gating, attention interface, etc.), so there is no shared underlying bug to canonicalize.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:46144 is the most concrete, code-path-oriented report, but it is only the strongest standalone issue here, not a representative duplicate target for the whole cluster.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:42936", + "right": "issue:46018", + "accept": false, + "reason": "Different models and failures: Mask2Former ignore_value usage vs DeepSeek-V4 shared expert gating." + }, + { + "left": "issue:43668", + "right": "issue:44031", + "accept": false, + "reason": "One is a type-hint issue in ModernBERTConfig; the other is a tokenizer regex warning regression." + }, + { + "left": "issue:42936", + "right": "issue:43240", + "accept": false, + "reason": "Unrelated bug classes: missing use of ignore_value vs kwargs not forwarded to a loss function." + }, + { + "left": "issue:43240", + "right": "issue:46144", + "accept": false, + "reason": "Different code paths and symptoms: loss-kwargs forwarding vs RoFormer attention interface usage." + }, + { + "left": "issue:42936", + "right": "issue:44031", + "accept": false, + "reason": "Mask2Former loss parameter handling is unrelated to tokenizer regex warnings." + }, + { + "left": "issue:42936", + "right": "issue:46144", + "accept": false, + "reason": "Different subsystems and fixes: Mask2Former loss logic vs RoFormer attention implementation." + }, + { + "left": "issue:42936", + "right": "issue:43813", + "accept": false, + "reason": "A model-loss behavior bug is not the same as a typo in a quantization attribute name." + }, + { + "left": "issue:43668", + "right": "issue:43813", + "accept": false, + "reason": "Both are small correctness issues, but in different configs/code paths with no shared underlying defect." + }, + { + "left": "issue:43240", + "right": "issue:43668", + "accept": false, + "reason": "Loss-function kwargs forwarding is unrelated to an incorrect type hint in a config class." + }, + { + "left": "issue:37934", + "right": "issue:43668", + "accept": false, + "reason": "Different component and defect type: Llama4 norm behavior vs ModernBERT config typing." + }, + { + "left": "issue:42936", + "right": "issue:43668", + "accept": false, + "reason": "Mask2Former ignore_value handling and ModernBERT norm_eps typing 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": 2, + "component_count": 26, + "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": "de2bf2c8d2aa545ab07c34409fc85918ec612da013e8cead28b60088e3941ad7", + "component_budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 3, + "serialized_chars": 2877, + "estimated_input_tokens": 720, + "estimated_eval_tokens": 1696 + }, + "budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 3, + "serialized_chars": 2877, + "estimated_input_tokens": 720, + "estimated_eval_tokens": 1696 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These three issues are not duplicates of one another; they cover unrelated Transformer truncation, learning-rate scheduling, and Wav2Vec2 word offsets.", + "confidence": 0.97, + "canonical_issue_reason": "No single canonical issue: the items discuss unrelated feature requests with no shared bug/change path.", + "canonical_pr_reason": null, + "best_issue_reason": "None of the issues is a good cluster representative because the set is heterogeneous rather than duplicate-focused.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:10519", + "right": "issue:11307", + "accept": false, + "reason": "Different features: tokenizer truncation behavior vs Wav2Vec2 timestamp/offset extraction." + }, + { + "left": "issue:10519", + "right": "issue:11005", + "accept": false, + "reason": "Unrelated requests: truncation direction option vs ReduceLROnPlateau-like scheduler behavior." + }, + { + "left": "issue:11005", + "right": "issue:11307", + "accept": false, + "reason": "No shared underlying bug or code path; scheduler functionality and Wav2Vec2 word offsets are distinct." + } + ] + }, + "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": 26, + "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": "27141eaf5ea5079e1610f4190ed7c645bcf48eb2388521a69f10f668d6cd4836", + "component_budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 2, + "serialized_chars": 2478, + "estimated_input_tokens": 620, + "estimated_eval_tokens": 1496 + }, + "budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 2, + "serialized_chars": 2478, + "estimated_input_tokens": 620, + "estimated_eval_tokens": 1496 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These three closed issues do not appear to describe the same underlying bug: one is about Tapas tokenization mismatch, one about an auth/download error message for private models, and one about failing to load bert-base-uncased. They are at most loosely related to model loading, so I would not cluster them as duplicates.", + "confidence": 0.94, + "canonical_issue_reason": "Issue #16618 is the broadest and most central user-facing failure report in this set, with the highest discussion activity; it is the best single representative if one is needed for the cluster, though it is not a duplicate of the others.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue #16618 is the most generally framed and the most active of the three, so it is the best issue to surface as the representative of this loose cluster.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:13244", + "right": "issue:16618", + "accept": false, + "reason": "Different bug classes: tokenization mismatch versus model loading failure. They do not look like the same underlying issue." + }, + { + "left": "issue:15980", + "right": "issue:16618", + "accept": false, + "reason": "One is an authentication/error-message problem when downloading a private model, the other is a generic inability to load a public model. Not 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": 4, + "component_count": 26, + "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": "3640334fad19d25ec040c1fa22cda792d1d49cac3e477b57b7a2c38091873776", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1590, + "estimated_input_tokens": 398, + "estimated_eval_tokens": 1052 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1590, + "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 different feature-tracking issues in the same general area, not duplicates: one requests Flash Attention 2 support for more architectures, the other requests GGUF support for more architectures.", + "confidence": 0.96, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": "issue:26350 is the stronger representative issue because it is the older, more active tracker for a concrete support expansion request, but it is not a duplicate of issue:33260.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:26350", + "right": "issue:33260", + "accept": false, + "reason": "Different underlying work: Flash Attention 2 support vs GGUF support. They share a broad theme of architecture support, but not the same 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": 5, + "component_count": 26, + "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-27088-9", + "nodes": [ + "issue:27088", + "issue:28936", + "issue:38435", + "issue:43834", + "issue:44365", + "issue:44370", + "issue:46222", + "issue:46226", + "issue:46931" + ], + "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:38435|issue:46222", + "issue:38435|issue:46226", + "issue:43834|issue:44370", + "issue:43834|issue:44365", + "issue:43834|issue:46222", + "issue:43834|issue:46226", + "issue:43834|issue:46931", + "issue:44365|issue:44370", + "issue:44365|issue:46222", + "issue:44365|issue:46226", + "issue:44365|issue:46931", + "issue:44370|issue:46222", + "issue:44370|issue:46226", + "issue:44370|issue:46931", + "issue:46222|issue:46226", + "issue:46222|issue:46931", + "issue:46226|issue:46931" + ], + "prepared_review_unit_hash": "5226c857bf02b4751de7e0a075a0f6c7dfbe76a3d868ff7073b61e298167dd10", + "component_budget": { + "node_count": 9, + "item_count": 9, + "soft_pair_count": 29, + "serialized_chars": 15982, + "estimated_input_tokens": 3996, + "estimated_eval_tokens": 8248 + }, + "budget": { + "node_count": 9, + "item_count": 9, + "soft_pair_count": 29, + "serialized_chars": 15982, + "estimated_input_tokens": 3996, + "estimated_eval_tokens": 8248 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "This cluster is mostly heterogeneous translation-request issues, but it does contain two distinct duplicate-like subgroups: the placeholder-title trio (43834/44365/44370) and the exact Thai-title pair (46222/46226). Those are the only clear duplicate sets; the rest are just broad i18n similarity across different languages/topics.", + "confidence": 0.82, + "canonical_issue_reason": "issue:43834 is the earliest and best anchor for the placeholder-title duplicate subgroup, but the cluster also contains a separate exact-duplicate pair at issue:46222/issue:46226, so no single issue fully dominates the whole cluster.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43834 is the strongest single representative because it sits in the more connected placeholder-title subgroup, while still acknowledging that issue:46222/issue:46226 is a separate exact-duplicate subgroup rather than part of the same one.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:27088", + "right": "issue:44370", + "accept": false, + "reason": "Different language request versus a generic placeholder-titled issue; same broad theme, not the same underlying report." + }, + { + "left": "issue:27088", + "right": "issue:44365", + "accept": false, + "reason": "Different language request versus a generic placeholder-titled issue; same broad theme, not the same underlying report." + }, + { + "left": "issue:27088", + "right": "issue:43834", + "accept": false, + "reason": "Turkish translation request versus a generic placeholder template; only broad i18n similarity." + }, + { + "left": "issue:27088", + "right": "issue:38435", + "accept": false, + "reason": "Turkish and Romanian translation requests are different concrete tasks, despite shared category." + }, + { + "left": "issue:27088", + "right": "issue:28936", + "accept": false, + "reason": "Turkish and Spanish translation requests are different concrete tasks, despite shared category." + }, + { + "left": "issue:28936", + "right": "issue:43834", + "accept": false, + "reason": "Spanish request versus a generic placeholder-titled issue; not enough evidence they are the same issue." + }, + { + "left": "issue:28936", + "right": "issue:44365", + "accept": false, + "reason": "Spanish request versus a generic placeholder-titled issue; not enough evidence they are the same issue." + }, + { + "left": "issue:28936", + "right": "issue:38435", + "accept": false, + "reason": "Spanish and Romanian translation requests are different concrete tasks, even if related by topic." + }, + { + "left": "issue:28936", + "right": "issue:44370", + "accept": false, + "reason": "Spanish request versus a generic placeholder-titled issue; not enough evidence they are the same issue." + }, + { + "left": "issue:38435", + "right": "issue:44370", + "accept": false, + "reason": "Romanian request versus a generic placeholder-titled issue; similar theme but not the same report." + }, + { + "left": "issue:38435", + "right": "issue:44365", + "accept": false, + "reason": "Romanian request versus a generic placeholder-titled issue; similar theme but not the same report." + }, + { + "left": "issue:38435", + "right": "issue:43834", + "accept": false, + "reason": "Romanian request versus a generic placeholder template; only broad i18n similarity." + }, + { + "left": "issue:38435", + "right": "issue:46222", + "accept": false, + "reason": "Different Thai-titled issue and no concrete evidence it is the same underlying translation task." + }, + { + "left": "issue:38435", + "right": "issue:46226", + "accept": false, + "reason": "Different Thai-titled issue and no concrete evidence it is the same underlying translation task." + }, + { + "left": "issue:43834", + "right": "issue:44370", + "accept": true, + "reason": "Identical placeholder title and near-identical body size; this looks like a repost of the same issue." + }, + { + "left": "issue:43834", + "right": "issue:44365", + "accept": true, + "reason": "Identical placeholder title and near-identical body size; this looks like a repost of the same issue." + }, + { + "left": "issue:43834", + "right": "issue:46222", + "accept": false, + "reason": "Placeholder-template issue versus a Thai-titled issue; no evidence they describe the same report." + }, + { + "left": "issue:43834", + "right": "issue:46226", + "accept": false, + "reason": "Placeholder-template issue versus a Thai-titled issue; no evidence they describe the same report." + }, + { + "left": "issue:43834", + "right": "issue:46931", + "accept": false, + "reason": "Different title and only broad i18n similarity; not the same concrete issue." + }, + { + "left": "issue:44365", + "right": "issue:44370", + "accept": true, + "reason": "Identical placeholder title and near-identical body size; this looks like a repost of the same issue." + }, + { + "left": "issue:44365", + "right": "issue:46222", + "accept": false, + "reason": "Generic placeholder title versus a Thai-titled issue; insufficient evidence of duplication." + }, + { + "left": "issue:44365", + "right": "issue:46226", + "accept": false, + "reason": "Generic placeholder title versus a Thai-titled issue; insufficient evidence of duplication." + }, + { + "left": "issue:44365", + "right": "issue:46931", + "accept": false, + "reason": "Different title and only broad i18n similarity; not the same concrete issue." + }, + { + "left": "issue:44370", + "right": "issue:46222", + "accept": false, + "reason": "Generic placeholder title versus a Thai-titled issue; insufficient evidence of duplication." + }, + { + "left": "issue:44370", + "right": "issue:46226", + "accept": false, + "reason": "Generic placeholder title versus a Thai-titled issue; insufficient evidence of duplication." + }, + { + "left": "issue:44370", + "right": "issue:46931", + "accept": false, + "reason": "Different title and only broad i18n similarity; not the same concrete issue." + }, + { + "left": "issue:46222", + "right": "issue:46226", + "accept": true, + "reason": "Exact same Thai title and identical body size; this is a clear duplicate issue report." + }, + { + "left": "issue:46222", + "right": "issue:46931", + "accept": false, + "reason": "Both are Thai-titled, but the titles differ and there is no concrete evidence they are the same issue." + }, + { + "left": "issue:46226", + "right": "issue:46931", + "accept": false, + "reason": "Both are Thai-titled, but the titles differ and there is no concrete evidence they are the same issue." + } + ] + }, + "evaluator_result": { + "accept": false, + "feedback": "The cluster summary is mostly grounded, but the canonical/best-representative reasoning overstates the evidence. The packet shows at least two duplicate-like subgroups: the placeholder-title trio (43834/44365/44370) and the exact Thai-title pair (46222/46226). So 43834 is not the only clearly repeated subgroup, and the rationale should not frame it that way. The soft-edge verdicts are generally conservative, but the representative selection needs to acknowledge the second duplicate-like subgroup and avoid implying a uniquely dominant one." + }, + "evaluator_used": true, + "retried": true, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 6, + "component_count": 26, + "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-36718-21", + "nodes": [ + "issue:36718", + "issue:36817", + "issue:36879", + "issue:37015", + "issue:41115", + "issue:41211", + "issue:42503", + "issue:42738", + "issue:42954", + "issue:43671", + "issue:43837", + "issue:43909", + "issue:44121", + "issue:44205", + "issue:45027", + "issue:45177", + "issue:45322", + "issue:46133", + "issue:46645", + "issue:46653", + "issue:47004" + ], + "soft_pairs": [ + "issue:36718|issue:47004", + "issue:36718|issue:41115", + "issue:36817|issue:41211", + "issue:36817|issue:43671", + "issue:36817|issue:36879", + "issue:36817|issue:42954", + "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:42954", + "issue:42738|issue:45027", + "issue:42503|issue:42738", + "issue:42738|issue:42954", + "issue:42738|issue:46653", + "issue:42954|issue:44205", + "issue:42954|issue:43671", + "issue:42954|issue:45177", + "issue:42954|issue:46653", + "issue:43671|issue:43837", + "issue:43671|issue:45027", + "issue:43671|issue:45177", + "issue:43671|issue:44121", + "issue:43837|issue:43909", + "issue:43837|issue:45322", + "issue:43837|issue:44121", + "issue:43909|issue:44205", + "issue:43909|issue:46645", + "issue:44121|issue:45322", + "issue:43837|issue:45027", + "issue:37015|issue:45027", + "issue:43837|issue:45177", + "issue:45322|issue:46133", + "issue:42503|issue:46653", + "issue:45027|issue:46653" + ], + "prepared_review_unit_hash": "55bc652708a47394caa8c9888092d92df6274d7c56e446cc23b9cafc1710b164", + "component_budget": { + "node_count": 21, + "item_count": 21, + "soft_pair_count": 48, + "serialized_chars": 28618, + "estimated_input_tokens": 7155, + "estimated_eval_tokens": 14566 + }, + "budget": { + "node_count": 21, + "item_count": 21, + "soft_pair_count": 48, + "serialized_chars": 28618, + "estimated_input_tokens": 7155, + "estimated_eval_tokens": 14566 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are mostly unrelated model-add/support requests that only look similar at the title level. I found no pair that clearly describes the same underlying bug or change, so all soft edges are rejected.", + "confidence": 0.96, + "canonical_issue_reason": "issue:46133 is the richest self-contained model-add request here and the best anchor for this mixed cluster.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:46133 has the clearest scope and the strongest discussion/context, making it the best representative issue in this set.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:36718", + "right": "issue:47004", + "accept": false, + "reason": "Related model names, but RoMa matcher and LoMa detector+matcher are different requests and likely different code paths." + }, + { + "left": "issue:36718", + "right": "issue:41115", + "accept": false, + "reason": "One is a keypoint matcher request, the other is a MiniCPM3 architecture request; not the same change." + }, + { + "left": "issue:36817", + "right": "issue:41211", + "accept": false, + "reason": "EuroBert config support and DEIMv2 model support are separate model additions." + }, + { + "left": "issue:36817", + "right": "issue:43671", + "accept": false, + "reason": "BERT config work and Qwen3-TTS audio support are unrelated underlying changes." + }, + { + "left": "issue:36817", + "right": "issue:36879", + "accept": false, + "reason": "EuroBert config and RF-DETR model support are different model families and implementation paths." + }, + { + "left": "issue:36817", + "right": "issue:42954", + "accept": false, + "reason": "EuroBert config and XiaoMi MiMo are distinct model requests with no shared code-path bug." + }, + { + "left": "issue:36879", + "right": "issue:41211", + "accept": false, + "reason": "RF-DETR and DEIMv2 are different detection models; same subsystem, different concrete change." + }, + { + "left": "issue:36879", + "right": "issue:43909", + "accept": false, + "reason": "RF-DETR object detection and LFM2.5 Audio are unrelated model integrations." + }, + { + "left": "issue:36879", + "right": "issue:44121", + "accept": false, + "reason": "RF-DETR model support and weight-sparse transformer support are different architectural changes." + }, + { + "left": "issue:36879", + "right": "issue:43837", + "accept": false, + "reason": "RF-DETR vision model support and Qwen3-ASR audio support are unrelated requests." + }, + { + "left": "issue:36879", + "right": "issue:43671", + "accept": false, + "reason": "RF-DETR and Qwen3-TTS target different model families and tasks." + }, + { + "left": "issue:37015", + "right": "issue:43837", + "accept": false, + "reason": "NeoBERT and Qwen3-ASR are different model additions, not one shared bug/change." + }, + { + "left": "issue:37015", + "right": "issue:44121", + "accept": false, + "reason": "NeoBERT support and circuit-sparsity transformer support do not describe the same code path." + }, + { + "left": "issue:37015", + "right": "issue:43909", + "accept": false, + "reason": "NeoBERT and LFM2.5 Audio are unrelated model requests." + }, + { + "left": "issue:36879", + "right": "issue:37015", + "accept": false, + "reason": "RF-DETR and NeoBERT are separate model integrations with different implementation surfaces." + }, + { + "left": "issue:37015", + "right": "issue:42503", + "accept": false, + "reason": "NeoBERT and ModernVBERT are distinct model additions, even if both are BERT-like." + }, + { + "left": "issue:41115", + "right": "issue:43837", + "accept": false, + "reason": "MiniCPM3 architecture work and Qwen3-ASR support are different model families/tasks." + }, + { + "left": "issue:41115", + "right": "issue:44121", + "accept": false, + "reason": "MiniCPM3 architecture support and weight-sparse transformer support are not the same underlying change." + }, + { + "left": "issue:41115", + "right": "issue:42503", + "accept": false, + "reason": "MiniCPM3 and ModernVBERT are separate architecture/model requests." + }, + { + "left": "issue:41115", + "right": "issue:45027", + "accept": false, + "reason": "MiniCPM3 architecture support and Voxtral TTS support are unrelated." + }, + { + "left": "issue:41211", + "right": "issue:43671", + "accept": false, + "reason": "DEIMv2 and Qwen3-TTS are different model support requests." + }, + { + "left": "issue:42503", + "right": "issue:43837", + "accept": false, + "reason": "ModernVBERT and Qwen3-ASR are different model families and tasks." + }, + { + "left": "issue:42503", + "right": "issue:43671", + "accept": false, + "reason": "ModernVBERT models and Qwen3-TTS support are unrelated additions." + }, + { + "left": "issue:42503", + "right": "issue:42954", + "accept": false, + "reason": "ModernVBERT and XiaoMi MiMo are separate model requests." + }, + { + "left": "issue:42738", + "right": "issue:45027", + "accept": false, + "reason": "BERT-like models with RoPE and Voxtral TTS are not the same concrete change." + }, + { + "left": "issue:42503", + "right": "issue:42738", + "accept": false, + "reason": "ModernVBERT models and generic BERT-like RoPE support are different requests." + }, + { + "left": "issue:42738", + "right": "issue:42954", + "accept": false, + "reason": "Generic BERT-like RoPE work and XiaoMi MiMo support are unrelated." + }, + { + "left": "issue:42738", + "right": "issue:46653", + "accept": false, + "reason": "BERT-like RoPE support and an unrelated person-named issue are clearly not the same change." + }, + { + "left": "issue:42954", + "right": "issue:44205", + "accept": false, + "reason": "XiaoMi MiMo and SAM3-LiteText are different model additions." + }, + { + "left": "issue:42954", + "right": "issue:43671", + "accept": false, + "reason": "XiaoMi MiMo and Qwen3-TTS support are unrelated model requests." + }, + { + "left": "issue:42954", + "right": "issue:45177", + "accept": false, + "reason": "XiaoMi MiMo and DeepSeek-OCR2 are separate model additions." + }, + { + "left": "issue:42954", + "right": "issue:46653", + "accept": false, + "reason": "XiaoMi MiMo and this unrelated issue are not the same underlying change." + }, + { + "left": "issue:43671", + "right": "issue:43837", + "accept": false, + "reason": "Qwen3-TTS and Qwen3-ASR share a vendor and modality, but they are different tasks and should not be merged as duplicates." + }, + { + "left": "issue:43671", + "right": "issue:45027", + "accept": false, + "reason": "Qwen3-TTS support and Voxtral-4B-TTS support are different model integrations." + }, + { + "left": "issue:43671", + "right": "issue:45177", + "accept": false, + "reason": "Qwen3-TTS and DeepSeek-OCR2 are unrelated model requests." + }, + { + "left": "issue:43671", + "right": "issue:44121", + "accept": false, + "reason": "Qwen3-TTS support and weight-sparse transformer support are different code paths." + }, + { + "left": "issue:43837", + "right": "issue:43909", + "accept": false, + "reason": "Qwen3-ASR and LFM2.5 Audio are both audio-related, but they are distinct models and tasks." + }, + { + "left": "issue:43837", + "right": "issue:45322", + "accept": false, + "reason": "Qwen3-ASR and EUPE are different model families and problem types." + }, + { + "left": "issue:43837", + "right": "issue:44121", + "accept": false, + "reason": "ASR support and weight-sparse transformer support are unrelated." + }, + { + "left": "issue:43909", + "right": "issue:44205", + "accept": false, + "reason": "LFM2.5 Audio and SAM3-LiteText are different model additions with no shared implementation." + }, + { + "left": "issue:43909", + "right": "issue:46645", + "accept": false, + "reason": "LFM2.5 Audio and Granite SWA+sink are unrelated changes." + }, + { + "left": "issue:44121", + "right": "issue:45322", + "accept": false, + "reason": "Weight-sparse transformer support and EUPE are distinct model requests." + }, + { + "left": "issue:43837", + "right": "issue:45027", + "accept": false, + "reason": "Qwen3-ASR and Voxtral-4B-TTS are different audio model supports." + }, + { + "left": "issue:37015", + "right": "issue:45027", + "accept": false, + "reason": "NeoBERT and Voxtral-4B-TTS are unrelated model additions." + }, + { + "left": "issue:43837", + "right": "issue:45177", + "accept": false, + "reason": "Qwen3-ASR and DeepSeek-OCR2 are not the same underlying change." + }, + { + "left": "issue:45322", + "right": "issue:46133", + "accept": false, + "reason": "EUPE and TIPSv2 are both model requests, but they are different architectures from different projects." + }, + { + "left": "issue:42503", + "right": "issue:46653", + "accept": false, + "reason": "ModernVBERT models and this unrelated closed issue are not duplicates." + }, + { + "left": "issue:45027", + "right": "issue:46653", + "accept": false, + "reason": "Voxtral-4B-TTS support and this unrelated closed issue are clearly different." + } + ] + }, + "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": 26, + "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-38740-2", + "nodes": [ + "issue:38740", + "issue:46443" + ], + "soft_pairs": [ + "issue:38740|issue:46443" + ], + "prepared_review_unit_hash": "c7cf5521976580ca35c8687c1e41aaaa65477eeb461fda0505f6d79a5e47df3b", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1505, + "estimated_input_tokens": 377, + "estimated_eval_tokens": 1010 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1505, + "estimated_input_tokens": 377, + "estimated_eval_tokens": 1010 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "The two issues are not duplicates and describe unrelated work: one is a closed docs request to add Pruna as an optimization framework, while the other is an open Vision RoPE refactor. There is no shared code path or concrete change overlap.", + "confidence": 0.97, + "canonical_issue_reason": "Issue 46443 is the better anchor only because it is the active technical change request; issue 38740 is a closed documentation-related request and unrelated to the refactor.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 46443 is the most suitable representative issue in this cluster because it concerns a concrete code refactor, whereas issue 38740 is a separate docs enhancement.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:38740", + "right": "issue:46443", + "accept": false, + "reason": "Different subjects entirely: a docs add request for Pruna versus a Vision RoPE refactor. No evidence they track 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": 8, + "component_count": 26, + "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": "7f8f49be9869834a032b79e4f26da0bd7ba06a0fb3fd7dd8136a95034d88883c", + "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 not duplicates: one is a generic decoder feature request, the other is a specific model support request. There is no clear shared underlying bug or change scope.", + "confidence": 0.95, + "canonical_issue_reason": "No canonical issue: the titles and implied scopes are unrelated, so neither issue can serve as a duplicate target for the other.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue #43037 is the better representative only in the sense that it is more specific and actionable, but it is not a duplicate of #41084.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:41084", + "right": "issue:43037", + "accept": false, + "reason": "Low textual overlap and different problem domains: block decoding behavior vs. DeepSeek v3.2 model support. They do not appear to describe 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": 9, + "component_count": 26, + "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-41306-2", + "nodes": [ + "issue:41306", + "issue:41826" + ], + "soft_pairs": [ + "issue:41306|issue:41826" + ], + "prepared_review_unit_hash": "99e6a57116b83c3b8694d40c44334bddb9cc805f0030b0955761afae0580fbc5", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1584, + "estimated_input_tokens": 396, + "estimated_eval_tokens": 1048 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1584, + "estimated_input_tokens": 396, + "estimated_eval_tokens": 1048 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These two issues are related only by vague theme (tiled / memory-oriented model changes) but describe different requested capabilities: one is about exposing a TiledFusedLogitsLoss/FLCE flag in `from_pretrained()`, the other about integrating TiledMLP for lower memory use.", + "confidence": 0.93, + "canonical_issue_reason": "Issue 41306 is the more concrete, narrowly scoped request and is easier to use as a representative node for the cluster.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 41306 is the better representative because it states a specific API/flag problem, whereas 41826 is a broader architectural feature request.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:41306", + "right": "issue:41826", + "accept": false, + "reason": "Different underlying changes: one concerns enabling a FLCE/TiledFusedLogitsLoss option in `from_pretrained()`, the other concerns adding TiledMLP for memory savings. Similar subsystem, but not the same 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": 26, + "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": "3ca343b2de5a32fb021cfb05036b85c303cc714d56a51b611ebdb1ceccb1be91", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1584, + "estimated_input_tokens": 396, + "estimated_eval_tokens": 1048 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1584, + "estimated_input_tokens": 396, + "estimated_eval_tokens": 1048 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These two issues both mention custom trainer behavior, but they describe different underlying problems: one is about `remove_unused_column` customizability in `Trainer`, the other about FLOPS estimation for custom trainers. They are related at the subsystem level, but not duplicates.", + "confidence": 0.89, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": "Issue 43598 is the broader and more architectural Trainer customization request, so it is the better representative issue of the two, though it is not a duplicate of 43600.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43598", + "right": "issue:43600", + "accept": false, + "reason": "Both touch custom trainer behavior, but one concerns `remove_unused_column` customization while the other concerns FLOPS estimation logic; they are distinct bugs/feature requests and should not be merged." + } + ] + }, + "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": 26, + "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": "42f306e0c815d50517ef1bfe7d3389e2fee8e6b3b135e98c022f5ed8803dd467", + "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 related only at the high level of \"add model support,\" but they target different Qwen variants and different AutoModel heads, so they are not duplicates.", + "confidence": 0.96, + "canonical_issue_reason": "Issue 43684 is the broader and more central request: adding Qwen3-Omni registration for AutoModel and AutoModelForConditionalGeneration.", + "canonical_pr_reason": null, + "best_issue_reason": "43684 is the best representative issue for the cluster because it is the more general model-registration request and closest to the shared theme.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43684", + "right": "issue:44405", + "accept": false, + "reason": "Different underlying work: Qwen3-Omni AutoModel/AutoModelForConditionalGeneration registration vs Qwen3.5 AutoModelForSequenceClassification support. Different model family and different task head, so they should not be merged as one bug/change." + } + ] + }, + "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": 26, + "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": "afbd5ab0b46be2fe8f7eb47922bed36d63a8ef73170941c5d771a1f514b4b460", + "component_budget": { + "node_count": 4, + "item_count": 4, + "soft_pair_count": 6, + "serialized_chars": 4262, + "estimated_input_tokens": 1066, + "estimated_eval_tokens": 2388 + }, + "budget": { + "node_count": 4, + "item_count": 4, + "soft_pair_count": 6, + "serialized_chars": 4262, + "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 issues titled \"huggingface\" look like duplicates of the same report; issue #43835 (\"Fett\") does not match that cluster. I\u2019d pick #43845 as the canonical issue because it is the earliest of the matching set.", + "confidence": 0.79, + "canonical_issue_reason": "#43845 is the earliest issue with the repeated \"huggingface\" title and matches the other two same-title reports closely.", + "canonical_pr_reason": null, + "best_issue_reason": "#43845 is the best representative of the duplicate cluster because it appears earliest and is textually aligned with #43846 and #43859.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43835", + "right": "issue:43845", + "accept": false, + "reason": "Different titles; \"Fett\" does not look like the same underlying issue as \"huggingface\"." + }, + { + "left": "issue:43835", + "right": "issue:43846", + "accept": false, + "reason": "Different titles; no evidence these report the same bug." + }, + { + "left": "issue:43835", + "right": "issue:43859", + "accept": false, + "reason": "Different titles; insufficient evidence of a shared underlying issue." + }, + { + "left": "issue:43845", + "right": "issue:43846", + "accept": true, + "reason": "Same title, same topic, and created seconds apart; these look like duplicate reports." + }, + { + "left": "issue:43845", + "right": "issue:43859", + "accept": true, + "reason": "Same title and clearly overlapping report topic; likely a duplicate of the same issue." + }, + { + "left": "issue:43846", + "right": "issue:43859", + "accept": true, + "reason": "Same title and same apparent report topic; likely duplicates of one issue." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "Grounded in the packet: #43845 is the earliest of the repeated \"huggingface\" issues, #43835 has a different title, and the soft-similarity evidence supports grouping the three \"huggingface\" reports together. The verdicts stay conservative enough based on the available metadata." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 13, + "component_count": 26, + "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": "b1a02bd293f341689386bc585664ef2b25e4463c7b570020867009cba924bcdf", + "component_budget": { + "node_count": 5, + "item_count": 5, + "soft_pair_count": 9, + "serialized_chars": 5782, + "estimated_input_tokens": 1446, + "estimated_eval_tokens": 3148 + }, + "budget": { + "node_count": 5, + "item_count": 5, + "soft_pair_count": 9, + "serialized_chars": 5782, + "estimated_input_tokens": 1446, + "estimated_eval_tokens": 3148 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "All items are low-information spam/promotional issues, not a real technical duplicate set. The three \"\" issues are the closest match, but the remaining two are different kinds of promotional/support-style noise, so no soft-edge pair is safe to merge.", + "confidence": 0.92, + "canonical_issue_reason": "Issue 44371 is the best representative of the repeated generic spam placeholder issues: it is the earliest of the identical \"\" entries and matches the strongest duplicate-like pattern in the set.", + "canonical_pr_reason": null, + "best_issue_reason": "44371 is the strongest issue canonical candidate because it is one of the exact \"\" duplicates and is earliest among them; the others are either the same placeholder or clearly different spam-like content.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44371", + "right": "issue:44572", + "accept": false, + "reason": "Same generic spam title, but no concrete bug/change is described, so there is no identifiable underlying issue to safely deduplicate." + }, + { + "left": "issue:44371", + "right": "issue:44573", + "accept": false, + "reason": "These are both placeholder spam reports, but the record does not show a specific shared bug or change; too little substance to merge." + }, + { + "left": "issue:44371", + "right": "issue:44418", + "accept": false, + "reason": "Both are noisy/low-signal, but one is a spam placeholder and the other is a documentation-suggestion title; not the same underlying issue." + }, + { + "left": "issue:44371", + "right": "issue:44937", + "accept": false, + "reason": "A generic spam placeholder and a promotional-looking title are not evidence of the same concrete problem." + }, + { + "left": "issue:44418", + "right": "issue:44572", + "accept": false, + "reason": "Documentation-suggestion wording does not match the spam placeholder issue, and there is no shared technical bug description." + }, + { + "left": "issue:44418", + "right": "issue:44573", + "accept": false, + "reason": "Different low-information topics: doc enhancement suggestion versus spam placeholder; no same underlying issue is visible." + }, + { + "left": "issue:44572", + "right": "issue:44573", + "accept": false, + "reason": "Both are identical spam placeholders, but identical spam text alone is not enough to prove the same underlying bug/change." + }, + { + "left": "issue:44572", + "right": "issue:44937", + "accept": false, + "reason": "Two different spam/promotional titles with no concrete technical problem to unify." + }, + { + "left": "issue:44573", + "right": "issue:44937", + "accept": false, + "reason": "The placeholder spam issue and the promotional title do not indicate the same issue or change request." + } + ] + }, + "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": 26, + "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": "51e97648d845cc5f4f8fa4438ba73ea194c35aad1717551596fbe9a9aa246707", + "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 issues are highly similar and likely related, but the evidence is only soft similarity; there is not enough to treat them as definite duplicates. Issue #45145 is the better canonical representative because it was created first and already has an inbound reference.", + "confidence": 0.86, + "canonical_issue_reason": "Issue #45145 is the better canonical choice because it was created first and already has an inbound reference, making it the more established representative of this similar issue pair.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue #45145 is the strongest issue candidate overall due to earlier creation and existing inbound linkage.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45145", + "right": "issue:45151", + "accept": false, + "reason": "The titles match and the bodies appear similar, but the packet only provides soft similarity with no concrete duplicate evidence, so this is not safe to accept as a definitive duplicate." + } + ] + }, + "evaluator_result": { + "accept": false, + "feedback": "The packet supports that the two issues are very similar and that issue 45145 is the more established canonical choice (earlier creation, one inbound reference). However, the soft-edge verdict is too strong: the packet only provides soft similarity and the deterministic accept is false, so labeling them as definite duplicates is not conservative enough. The summary also overstates the evidence by adding \"spam issue,\" which is not grounded in the packet." + }, + "evaluator_used": true, + "retried": true, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 15, + "component_count": 26, + "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": 16, + "component_count": 26, + "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": "21e2144ac963148ca74453492f013b33a858db6c330e5d3a0d3048092b13d955", + "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": "Two closed issues contain only different personal-name/date titles and no body or linkage. There is no evidence they describe the same bug or change, so they should not be clustered as duplicates.", + "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": "Only superficial surname overlap; the titles refer to different people and dates, with no shared context, body, or links indicating the same underlying issue." + } + ] + }, + "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": 26, + "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": "5ebe17517ac735f7d673dc8f9d172084def9c92196a45e32aa3c201a9c4aa506", + "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": "These two issues do not describe the same underlying bug or change; they are separate empty-name/date entries with only superficial title similarity.", + "confidence": 0.98, + "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 just different personal-name/date strings with no shared problem statement, code path, or explicit linkage; this looks like unrelated content rather than duplicates." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 18, + "component_count": 26, + "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-45810-2", + "nodes": [ + "issue:45810", + "issue:46295" + ], + "soft_pairs": [ + "issue:45810|issue:46295" + ], + "prepared_review_unit_hash": "d223086690dc75f3c6119e8e5c8fbfd0584952a77e1c6cc4b051aa10adb271bc", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1541, + "estimated_input_tokens": 386, + "estimated_eval_tokens": 1028 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1541, + "estimated_input_tokens": 386, + "estimated_eval_tokens": 1028 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "The two issues are not duplicates: one requests a new `Qwen3_5ForTokenClassification` model class for value-model use, while the other asks for FP8 training support. They concern different features and code paths.", + "confidence": 0.95, + "canonical_issue_reason": "Issue #45810 is the more specific and concrete implementation request; #46295 is broader and less tied to a single model/API addition.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue #45810 is the best representative of the cluster only because it is the most concrete feature request, but the two issues are not actually the same underlying problem.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45810", + "right": "issue:46295", + "accept": false, + "reason": "Different underlying requests: adding a token-classification model class versus enabling FP8 training support. They do not describe the same bug or change path." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 19, + "component_count": 26, + "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": "3e57b6dad31555b89633a64e43b9ef6b6c87560ad5c30f0edf70895acd013ece", + "component_budget": { + "node_count": 5, + "item_count": 5, + "soft_pair_count": 8, + "serialized_chars": 5721, + "estimated_input_tokens": 1431, + "estimated_eval_tokens": 3118 + }, + "budget": { + "node_count": 5, + "item_count": 5, + "soft_pair_count": 8, + "serialized_chars": 5721, + "estimated_input_tokens": 1431, + "estimated_eval_tokens": 3118 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "This cluster splits into two near-identical duplicate issue pairs (45870/45871 and 45872/45873) plus a separate, thematically related financial dataset promo issue (45869).", + "confidence": 0.96, + "canonical_issue_reason": "No single canonical issue covers the full cluster cleanly, because it contains two separate near-identical duplicate pairs and a separate financial dataset promo issue. If forced to choose one anchor, 45870 is the best representative of the prediction-example pair.", + "canonical_pr_reason": null, + "best_issue_reason": "45870 is the best single issue representative only in a narrow sense: it anchors one near-identical duplicate pair and has the most visible activity among the example issues, but it does not represent the scraping-service pair or the dataset promo issue.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45869", + "right": "issue:45870", + "accept": false, + "reason": "Related by financial/DataForge theme, but the dataset bundle promo and the prediction example are different concrete issues." + }, + { + "left": "issue:45869", + "right": "issue:45871", + "accept": false, + "reason": "The dataset bundle promo and the prediction example are thematically related, but not the same underlying issue." + }, + { + "left": "issue:45870", + "right": "issue:45871", + "accept": true, + "reason": "Same title and highly similar body length; this looks like a near-identical duplicate pair for the same prediction-example issue." + }, + { + "left": "issue:45870", + "right": "issue:45872", + "accept": false, + "reason": "Both are about financial/DataForge topics, but one is a prediction example and the other is a scraping-service request; different concrete issues." + }, + { + "left": "issue:45870", + "right": "issue:45873", + "accept": false, + "reason": "Broad theme overlaps, but the prediction-example issue and the scraping-service issue are different concrete requests." + }, + { + "left": "issue:45871", + "right": "issue:45872", + "accept": false, + "reason": "Different reported topics: prediction example versus scraping service. Not the same underlying issue." + }, + { + "left": "issue:45871", + "right": "issue:45873", + "accept": false, + "reason": "Only broad thematic overlap; the issues describe different concrete requests." + }, + { + "left": "issue:45872", + "right": "issue:45873", + "accept": true, + "reason": "Same title and highly similar body length; this looks like a near-identical duplicate pair for the scraping-service issue." + } + ] + }, + "evaluator_result": { + "accept": false, + "feedback": "The clustering read is mostly grounded, but a couple of phrasings are a bit too strong for the packet. In particular, 45869 is better described as a separate, thematically related issue rather than \"unrelated,\" and the two matching pairs are supported as very strong duplicates, but \"exact duplicates\" is slightly stronger than the visible evidence unless the underlying body text is known. Please keep the summary more conservative: two near-identical duplicate pairs plus a separate financial dataset promo issue." + }, + "evaluator_used": true, + "retried": true, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 20, + "component_count": 26, + "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-46439-2", + "nodes": [ + "issue:46439", + "issue:46441" + ], + "soft_pairs": [ + "issue:46439|issue:46441" + ], + "prepared_review_unit_hash": "400a7d13ed5b4c7b91f1c73d9b69b35d6ead1bef662a6f3ed2e1a01c27104f9a", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1636, + "estimated_input_tokens": 409, + "estimated_eval_tokens": 1074 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1636, + "estimated_input_tokens": 409, + "estimated_eval_tokens": 1074 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "The two issues are related around static cache and linear-attention crashes, but the available evidence is too thin to treat them as the same underlying bug. #46439 is broader; #46441 reads like a specific Qwen3.5 KeyError symptom that is not proven to share the same root cause.", + "confidence": 0.87, + "canonical_issue_reason": "No canonical duplicate should be selected from the available evidence: #46439 is the broader report, but the titles alone do not establish that #46441 is the same bug rather than a related static-cache failure mode.", + "canonical_pr_reason": null, + "best_issue_reason": "#46439 is the best representative issue because it describes the more general mechanism and failure class, whereas #46441 is narrower and product-specific.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:46439", + "right": "issue:46441", + "accept": false, + "reason": "Reject: the titles suggest a soft relationship, but they do not demonstrate the same concrete bug. #46439 mentions early initialization corrupting linear-attention shapes; #46441 reports a Qwen3.5 static-cache KeyError. Related subsystem, but not enough to merge as one issue." + } + ] + }, + "evaluator_result": { + "accept": false, + "feedback": "Reject: the packet only supports a soft-similarity relationship between the two issues, and the analyst overstates the evidence by asserting a shared root cause (\u201csame static-cache/linear-attention initialization bug\u201d) that is not demonstrated here. Based on the titles alone, 46439 and 46441 are related, but the specific mechanism and equivalence are not sufficiently grounded to accept a same-bug merge verdict." + }, + "evaluator_used": true, + "retried": true, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 21, + "component_count": 26, + "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-46551-2", + "nodes": [ + "issue:46551", + "issue:46552" + ], + "soft_pairs": [ + "issue:46551|issue:46552" + ], + "prepared_review_unit_hash": "96ca8d09c18e12d6f371ae516eeed79225ed573f1a5f1072ba83da63f9c00ea0", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1707, + "estimated_input_tokens": 427, + "estimated_eval_tokens": 1110 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1707, + "estimated_input_tokens": 427, + "estimated_eval_tokens": 1110 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These two issues are duplicates of the same Wav2Vec2CTCTokenizer v5 special-token leak bug. The open issue is the better tracker.", + "confidence": 0.99, + "canonical_issue_reason": "Issue #46552 is open and already has inbound references, so it is the better canonical tracker for the duplicate report.", + "canonical_pr_reason": null, + "best_issue_reason": "#46552 is the best issue because it is the active open report for the exact same bug and is more likely to collect follow-up discussion and fixes.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:46551", + "right": "issue:46552", + "accept": true, + "reason": "Same title, same described behavior, and no distinguishing details; these are the same underlying bug report." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "Grounded by the packet: the two issues have identical titles, identical body length, and a soft_similarity score of 1.0, while #46552 is open and has inbound references. The canonical-tracker choice is conservative and well supported." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 22, + "component_count": 26, + "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-46657-2", + "nodes": [ + "issue:46657", + "issue:46692" + ], + "soft_pairs": [ + "issue:46657|issue:46692" + ], + "prepared_review_unit_hash": "c7d353fd89f60220e1203e8b8367a34ac164f4cfef4a6d6bd6e0a3ffa6989d8b", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1570, + "estimated_input_tokens": 393, + "estimated_eval_tokens": 1042 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1570, + "estimated_input_tokens": 393, + "estimated_eval_tokens": 1042 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These issues describe different failure modes in training object detection models: one is a tensor size mismatch error, the other is an out-of-memory condition after several epochs on RT-DETR v2. They are related to training, but not the same underlying bug.", + "confidence": 0.96, + "canonical_issue_reason": "Issue #46657 is the better canonical issue only in the sense that it is a concrete, clearly defined bug report with a specific runtime error; however, it does not cover the OOM-at-epoch problem in #46692.", + "canonical_pr_reason": null, + "best_issue_reason": "#46657 is the stronger representative for this cluster because it is more specific and actionable, with a distinct error signature. #46692 appears to be a separate training stability/memory issue rather than a duplicate.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:46657", + "right": "issue:46692", + "accept": false, + "reason": "The titles point to different concrete failures: tensor shape mismatch versus out-of-memory during later epochs. They may both occur in object detection training, but they do not look like the same underlying bug and would 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": 23, + "component_count": 26, + "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-46822-2", + "nodes": [ + "issue:46822", + "issue:46823" + ], + "soft_pairs": [ + "issue:46822|issue:46823" + ], + "prepared_review_unit_hash": "2f358da6bf9d1810e867b5dfe70628e52b705d4c844e6a30d95af20ad97e0920", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1672, + "estimated_input_tokens": 418, + "estimated_eval_tokens": 1092 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1672, + "estimated_input_tokens": 418, + "estimated_eval_tokens": 1092 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "The two issues are both model-loading failures, but they describe different code paths and different root causes: one is a ZeRO-3/zero.Init partitioning bug for a composite VL submodule, the other is a 4-bit auto device-mapping bug that wrongly dispatches to CPU despite fitting memory. Not a duplicate pair.", + "confidence": 0.95, + "canonical_issue_reason": "Issue 46822 is the better canonical issue because it has more discussion/activity and is the more established report, while still being a distinct ZeRO-3 loading bug.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 46822 is the strongest representative of its own problem, but it should not be used to cover issue 46823 because the failures involve different mechanisms (ZeRO partitioning vs. 4-bit device_map dispatch).", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:46822", + "right": "issue:46823", + "accept": false, + "reason": "Different underlying bugs: 46822 is about ZeRO-3 zero.Init failing to partition a composite submodule and causing OOM; 46823 is about 4-bit + device_map='auto' incorrectly sending modules to CPU and raising ValueError. Same broad theme, but not the same fixable issue." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 24, + "component_count": 26, + "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-46906-6", + "nodes": [ + "issue:46906", + "issue:46940", + "issue:46948", + "issue:46969", + "issue:46985", + "issue:46998" + ], + "soft_pairs": [ + "issue:46906|issue:46940", + "issue:46906|issue:46948", + "issue:46906|issue:46969", + "issue:46906|issue:46985", + "issue:46906|issue:46998", + "issue:46940|issue:46948", + "issue:46940|issue:46969", + "issue:46940|issue:46985", + "issue:46940|issue:46998", + "issue:46948|issue:46969", + "issue:46948|issue:46985", + "issue:46948|issue:46998", + "issue:46969|issue:46985", + "issue:46969|issue:46998", + "issue:46985|issue:46998" + ], + "prepared_review_unit_hash": "6aa11fa20649608b9d4c07c422eefd75b02b10fa1896b05fbf3fa8ebe51e1d5e", + "component_budget": { + "node_count": 6, + "item_count": 6, + "soft_pair_count": 15, + "serialized_chars": 8797, + "estimated_input_tokens": 2200, + "estimated_eval_tokens": 4656 + }, + "budget": { + "node_count": 6, + "item_count": 6, + "soft_pair_count": 15, + "serialized_chars": 8797, + "estimated_input_tokens": 2200, + "estimated_eval_tokens": 4656 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are a series of similarly titled daily integration-failure triage issues for different dates, not a single duplicated report. They should remain separate unless later evidence links them to the same underlying failure.", + "confidence": 0.95, + "canonical_issue_reason": "issue:46906 is the earliest item and has the strongest thread activity (highest inbound references and discussion), so it is the best anchor for the cluster if one is needed.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:46906 is the most central and most visible of the daily triage issues, with the highest inbound reference count and the earliest timestamp.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:46906", + "right": "issue:46940", + "accept": false, + "reason": "Same triage template, but different date-specific integration reports; no evidence they describe the same failure instance." + }, + { + "left": "issue:46906", + "right": "issue:46948", + "accept": false, + "reason": "These appear to be separate daily failure triage posts rather than one underlying bug report." + }, + { + "left": "issue:46906", + "right": "issue:46969", + "accept": false, + "reason": "Different dates and standalone triage threads; not enough evidence of a shared root cause." + }, + { + "left": "issue:46906", + "right": "issue:46985", + "accept": false, + "reason": "Likely another daily integration-failure triage entry, not a duplicate of the earlier report." + }, + { + "left": "issue:46906", + "right": "issue:46998", + "accept": false, + "reason": "Same naming pattern but distinct date-based triage issue; no concrete overlap showing one bug." + }, + { + "left": "issue:46940", + "right": "issue:46948", + "accept": false, + "reason": "Adjacent daily triage issues, but no direct evidence they track the same incident." + }, + { + "left": "issue:46940", + "right": "issue:46969", + "accept": false, + "reason": "Different day reports with no explicit shared target or linked failure details." + }, + { + "left": "issue:46940", + "right": "issue:46985", + "accept": false, + "reason": "Separate date-stamped integration triage items; not safe to merge as duplicates." + }, + { + "left": "issue:46940", + "right": "issue:46998", + "accept": false, + "reason": "No sign these are the same underlying bug; they look like repeated daily triage posts." + }, + { + "left": "issue:46948", + "right": "issue:46969", + "accept": false, + "reason": "Similar title pattern only; likely different daily integration failures." + }, + { + "left": "issue:46948", + "right": "issue:46985", + "accept": false, + "reason": "Distinct date-based triage issues with no concrete shared root cause evidence." + }, + { + "left": "issue:46948", + "right": "issue:46998", + "accept": false, + "reason": "Not enough overlap beyond the template title to treat as duplicates." + }, + { + "left": "issue:46969", + "right": "issue:46985", + "accept": false, + "reason": "Two separate daily triage entries; no explicit linkage to one bug or change." + }, + { + "left": "issue:46969", + "right": "issue:46998", + "accept": false, + "reason": "Likely independent failure triage reports for different dates." + }, + { + "left": "issue:46985", + "right": "issue:46998", + "accept": false, + "reason": "These are consecutive daily triage issues, but the available evidence does not show the same underlying failure." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 25, + "component_count": 26, + "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-46983-2", + "nodes": [ + "issue:46983", + "issue:46984" + ], + "soft_pairs": [ + "issue:46983|issue:46984" + ], + "prepared_review_unit_hash": "49508a13e178a95af52c8da3150cb18b76164d0fbd682947fe181e079c0c3267", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1490, + "estimated_input_tokens": 373, + "estimated_eval_tokens": 1002 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1490, + "estimated_input_tokens": 373, + "estimated_eval_tokens": 1002 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "The two issues appear to be duplicates of the same feature request: adding RWKV7 model support. The open issue is the better canonical tracker.", + "confidence": 0.98, + "canonical_issue_reason": "Issue 46984 is open and can serve as the active tracker for the same RWKV7 support request; 46983 is closed and appears to cover the same topic.", + "canonical_pr_reason": null, + "best_issue_reason": "46984 is the best issue because it is open, current, and has the exact same title/topic as 46983.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:46983", + "right": "issue:46984", + "accept": true, + "reason": "Same title and same feature request (RWKV7 model support); the only meaningful difference is that one is closed and the other is open, so they look like duplicate issues." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "Grounded and conservative overall: the two issues have identical titles, no conflicting signals, and the open one is a reasonable canonical tracker. The duplicate/soft-match verdict is supported by the packet and the summary appropriately uses \"appear.\"" + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 26, + "component_count": 26, + "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-47007-2", + "nodes": [ + "issue:47007", + "issue:47015" + ], + "soft_pairs": [ + "issue:47007|issue:47015" + ], + "prepared_review_unit_hash": "d2dd4c0bac7afba9ce71e74ec8d070b73ae69dac5ac58ef72233d9ce5fe905a1", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1622, + "estimated_input_tokens": 406, + "estimated_eval_tokens": 1068 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1622, + "estimated_input_tokens": 406, + "estimated_eval_tokens": 1068 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Both items are about AI-driven vulnerability discovery, but they read like separate writeups/ideas rather than the same concrete issue. I would not deduplicate them.", + "confidence": 0.57, + "canonical_issue_reason": "Issue 47015 is the broader, more general framing of the topic and is the better cluster representative by title alone.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 47015 is the most central umbrella for the cluster because it directly names the AI-based vulnerability discovery system without narrowing to a specific lesson/article angle.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:47007", + "right": "issue:47015", + "accept": false, + "reason": "They share a broad cybersecurity/LLM theme, but there is no sign they describe the same underlying bug, request, or change. The titles suggest different articles or proposals, so this is not a safe duplicate." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 1, + "component_count": 35, + "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-40546-41", + "nodes": [ + "pull_request:40546", + "pull_request:43451", + "pull_request:46196", + "pull_request:46266", + "pull_request:46658", + "pull_request:46707", + "pull_request:46836", + "pull_request:46891", + "pull_request:46895", + "pull_request:46902", + "pull_request:46903", + "pull_request:46907", + "pull_request:46908", + "pull_request:46912", + "pull_request:46920", + "pull_request:46926", + "pull_request:46927", + "pull_request:46928", + "pull_request:46933", + "pull_request:46941", + "pull_request:46942", + "pull_request:46949", + "pull_request:46950", + "pull_request:46951", + "pull_request:46954", + "pull_request:46958", + "pull_request:46961", + "pull_request:46965", + "pull_request:46968", + "pull_request:46970", + "pull_request:46971", + "pull_request:46974", + "pull_request:46977", + "pull_request:46986", + "pull_request:46995", + "pull_request:47006", + "pull_request:47014" + ], + "soft_pairs": [ + "pull_request:46895|pull_request:46958", + "pull_request:46903|pull_request:46958", + "pull_request:46950|pull_request:46951", + "pull_request:46927|pull_request:46950", + "pull_request:46950|pull_request:47006", + "pull_request:46927|pull_request:46951", + "pull_request:46951|pull_request:46971", + "pull_request:46949|pull_request:46971", + "pull_request:46927|pull_request:47006", + "pull_request:46951|pull_request:47006", + "pull_request:46971|pull_request:47006", + "pull_request:46941|pull_request:46970", + "pull_request:46908|pull_request:47006", + "pull_request:46908|pull_request:46928", + "pull_request:46902|pull_request:46965", + "pull_request:46907|pull_request:46926", + "pull_request:46970|pull_request:46971", + "pull_request:46926|pull_request:46928", + "pull_request:46927|pull_request:46986", + "pull_request:46951|pull_request:46986", + "pull_request:46933|pull_request:46942", + "pull_request:46926|pull_request:46941", + "pull_request:46928|pull_request:46941", + "pull_request:46907|pull_request:46928", + "pull_request:46928|pull_request:47006", + "pull_request:46941|pull_request:46986", + "pull_request:46941|pull_request:47006", + "pull_request:46907|pull_request:46908", + "pull_request:46908|pull_request:46951", + "pull_request:46954|pull_request:46961", + "pull_request:46907|pull_request:46941", + "pull_request:46954|pull_request:46977", + "pull_request:46926|pull_request:46927", + "pull_request:43451|pull_request:46954", + "pull_request:46954|pull_request:46968", + "pull_request:46954|pull_request:46958", + "pull_request:46958|pull_request:46961", + "pull_request:46958|pull_request:46977", + "pull_request:46961|pull_request:46974", + "pull_request:46958|pull_request:46974", + "pull_request:46974|pull_request:46977", + "pull_request:43451|pull_request:46920", + "pull_request:46920|pull_request:46968", + "pull_request:43451|pull_request:46958", + "pull_request:46971|pull_request:46974", + "pull_request:46968|pull_request:46974", + "pull_request:46920|pull_request:46961", + "pull_request:46961|pull_request:46968", + "pull_request:46920|pull_request:46977", + "pull_request:46968|pull_request:46977", + "pull_request:46958|pull_request:46968", + "pull_request:43451|pull_request:46707", + "pull_request:46266|pull_request:46707", + "pull_request:46196|pull_request:46995", + "pull_request:46707|pull_request:46961", + "pull_request:46891|pull_request:46961", + "pull_request:43451|pull_request:46891", + "pull_request:46707|pull_request:46977", + "pull_request:46891|pull_request:46977", + "pull_request:46707|pull_request:46968", + "pull_request:46891|pull_request:46968", + "pull_request:43451|pull_request:46836", + "pull_request:46912|pull_request:46977", + "pull_request:40546|pull_request:43451", + "pull_request:46196|pull_request:46965", + "pull_request:46836|pull_request:46961", + "pull_request:46836|pull_request:46965", + "pull_request:46836|pull_request:46977", + "pull_request:46196|pull_request:46977", + "pull_request:46266|pull_request:47014", + "pull_request:43451|pull_request:46658", + "pull_request:43451|pull_request:46965" + ], + "prepared_review_unit_hash": "58babfc487e294cd72377c8403cd337594f131ebf84e5da09e9fd540a4dd6d80", + "component_budget": { + "node_count": 41, + "item_count": 41, + "soft_pair_count": 126, + "serialized_chars": 142820, + "estimated_input_tokens": 35705, + "estimated_eval_tokens": 71666 + }, + "budget": { + "node_count": 37, + "item_count": 37, + "soft_pair_count": 72, + "serialized_chars": 105282, + "estimated_input_tokens": 26321, + "estimated_eval_tokens": 52898 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "Mostly unrelated PRs are clustered together by topic and test-failure triage. The CohereASR pair does refer to the same underlying loss bug, but it is not a mergeable duplicate pair because the two PRs are alternative fixes rather than complementary changes.", + "confidence": 0.88, + "canonical_issue_reason": null, + "canonical_pr_reason": "pull_request:46895 is the best canonical CohereASR representative: it directly targets issue 46894, stays scoped to the loss-path fix, and is the cleaner issue-linked patch.", + "best_issue_reason": null, + "best_pr_reason": "pull_request:46895 is the strongest representative PR for the cluster because it is the most direct, narrowly scoped fix for the CohereASR double-shift bug.", + "soft_edge_verdicts": [ + { + "left": "pull_request:46895", + "right": "pull_request:46958", + "accept": false, + "reason": "Same CohereASR bug, but these are alternative loss fixes rather than complementary edits, so they are not mergeable into one PR." + }, + { + "left": "pull_request:46903", + "right": "pull_request:46958", + "accept": false, + "reason": "Same model family, but 46903 also changes PPFormulaNet and is a different fix shape; not the same concrete change." + }, + { + "left": "pull_request:46950", + "right": "pull_request:46951", + "accept": false, + "reason": "Mamba2 OOM behavior vs PEFT loading option; different code paths and fixes." + }, + { + "left": "pull_request:46927", + "right": "pull_request:46950", + "accept": false, + "reason": "GLM46V dtype casting vs Mamba2 OOM; unrelated bugs." + }, + { + "left": "pull_request:46950", + "right": "pull_request:47006", + "accept": false, + "reason": "Mamba2 OOM vs Gemma3 generation/cache behavior; unrelated." + }, + { + "left": "pull_request:46927", + "right": "pull_request:46951", + "accept": false, + "reason": "GLM46V dtype fix vs PEFT integration loading; unrelated." + }, + { + "left": "pull_request:46951", + "right": "pull_request:46971", + "accept": false, + "reason": "PEFT load option vs Mamba2 test dtype argument; different fixes." + }, + { + "left": "pull_request:46949", + "right": "pull_request:46971", + "accept": false, + "reason": "Whisper/audioflamingo cache handling vs Mamba2 OOM test setup; unrelated." + }, + { + "left": "pull_request:46927", + "right": "pull_request:47006", + "accept": false, + "reason": "GLM46V dtype fix vs Gemma3 generation/cache change; unrelated." + }, + { + "left": "pull_request:46951", + "right": "pull_request:47006", + "accept": false, + "reason": "PEFT loading vs Gemma3 generation/cache; unrelated." + }, + { + "left": "pull_request:46971", + "right": "pull_request:47006", + "accept": false, + "reason": "Mamba2 test cleanup vs Gemma3 generation/cache; unrelated." + }, + { + "left": "pull_request:46941", + "right": "pull_request:46970", + "accept": false, + "reason": "DAC output-mismatch fix vs musicgen_melody output mismatch; different models/code paths." + }, + { + "left": "pull_request:46908", + "right": "pull_request:47006", + "accept": false, + "reason": "DeepSeekV3.2 offload test fix vs Gemma3 generation/cache; unrelated." + }, + { + "left": "pull_request:46908", + "right": "pull_request:46928", + "accept": false, + "reason": "DeepSeekV3.2 offload test fix vs BLIP2 tokenizer mapping; unrelated." + }, + { + "left": "pull_request:46902", + "right": "pull_request:46965", + "accept": false, + "reason": "Offloading/save-converted-weights bugfix vs DeepSeekV4 use_cache support; different runtime paths." + }, + { + "left": "pull_request:46907", + "right": "pull_request:46926", + "accept": false, + "reason": "EdgeTam import_or_config fix vs MiniMax-M3-VL load_error test fix; different bugs." + }, + { + "left": "pull_request:46970", + "right": "pull_request:46971", + "accept": false, + "reason": "musicgen_melody model bug vs Mamba2 test-only OOM tweak; different changes." + }, + { + "left": "pull_request:46926", + "right": "pull_request:46928", + "accept": false, + "reason": "MiniMax load error vs BLIP2 tokenizer mapping; unrelated." + }, + { + "left": "pull_request:46927", + "right": "pull_request:46986", + "accept": false, + "reason": "GLM46V input dtype fix vs GLM OCR output mismatch; different models and bugs." + }, + { + "left": "pull_request:46951", + "right": "pull_request:46986", + "accept": false, + "reason": "PEFT loading fix vs GLM OCR model fix; unrelated." + }, + { + "left": "pull_request:46933", + "right": "pull_request:46942", + "accept": false, + "reason": "Both are dtype-related, but Gemma4 quantized embedder casting and TimesFM2.5 block casting are separate concrete fixes." + }, + { + "left": "pull_request:46926", + "right": "pull_request:46941", + "accept": false, + "reason": "MiniMax load error vs DAC output mismatch; unrelated." + }, + { + "left": "pull_request:46928", + "right": "pull_request:46941", + "accept": false, + "reason": "BLIP2 tokenizer handling vs DAC output mismatch; unrelated." + }, + { + "left": "pull_request:46907", + "right": "pull_request:46928", + "accept": false, + "reason": "EdgeTam import_or_config vs BLIP2 tokenizer mapping; unrelated." + }, + { + "left": "pull_request:46928", + "right": "pull_request:47006", + "accept": false, + "reason": "BLIP2 tokenizer mapping vs Gemma3 generation/cache; unrelated." + }, + { + "left": "pull_request:46941", + "right": "pull_request:46986", + "accept": false, + "reason": "DAC output mismatch vs GLM OCR output mismatch; different models and code paths." + }, + { + "left": "pull_request:46941", + "right": "pull_request:47006", + "accept": false, + "reason": "DAC fixes vs Gemma3 generation/cache; unrelated." + }, + { + "left": "pull_request:46907", + "right": "pull_request:46908", + "accept": false, + "reason": "EdgeTam import_or_config vs DeepSeekV3.2 offload test; different failures." + }, + { + "left": "pull_request:46908", + "right": "pull_request:46951", + "accept": false, + "reason": "DeepSeekV3.2 offload test fix vs PEFT integration loading; unrelated." + }, + { + "left": "pull_request:46954", + "right": "pull_request:46961", + "accept": false, + "reason": "GLM-ASR pytest-v9 compatibility fix vs Serge branch-normalization workflow; unrelated." + }, + { + "left": "pull_request:46907", + "right": "pull_request:46941", + "accept": false, + "reason": "EdgeTam import_or_config vs DAC output mismatch; unrelated." + }, + { + "left": "pull_request:46954", + "right": "pull_request:46977", + "accept": false, + "reason": "GLM-ASR pytest compatibility vs Mllama arg-name deprecation; different changes." + }, + { + "left": "pull_request:43451", + "right": "pull_request:46954", + "accept": false, + "reason": "Molmo2 model addition vs GLM-ASR pytest compatibility; unrelated." + }, + { + "left": "pull_request:46954", + "right": "pull_request:46968", + "accept": false, + "reason": "GLM-ASR pytest compatibility vs docs autolink cleanup; unrelated." + }, + { + "left": "pull_request:46954", + "right": "pull_request:46958", + "accept": false, + "reason": "GLM-ASR pytest compatibility vs CohereASR loss fix; unrelated." + }, + { + "left": "pull_request:46958", + "right": "pull_request:46961", + "accept": false, + "reason": "CohereASR loss bugfix vs workflow automation; unrelated." + }, + { + "left": "pull_request:46958", + "right": "pull_request:46977", + "accept": false, + "reason": "CohereASR loss fix vs Mllama arg-name deprecation; different changes." + }, + { + "left": "pull_request:46961", + "right": "pull_request:46974", + "accept": false, + "reason": "Workflow automation vs MusicGen generate typo fix; unrelated." + }, + { + "left": "pull_request:46958", + "right": "pull_request:46974", + "accept": false, + "reason": "CohereASR loss fix vs MusicGen generate typo; unrelated." + }, + { + "left": "pull_request:46974", + "right": "pull_request:46977", + "accept": false, + "reason": "MusicGen generate typo fix vs Mllama arg-name deprecation; unrelated." + }, + { + "left": "pull_request:43451", + "right": "pull_request:46920", + "accept": false, + "reason": "Molmo2 model addition vs Voxtral Realtime API adaptation; unrelated." + }, + { + "left": "pull_request:46920", + "right": "pull_request:46968", + "accept": false, + "reason": "Voxtral Realtime API adaptation vs docs autolink cleanup; unrelated." + }, + { + "left": "pull_request:43451", + "right": "pull_request:46958", + "accept": false, + "reason": "Molmo2 model addition vs CohereASR loss fix; unrelated." + }, + { + "left": "pull_request:46971", + "right": "pull_request:46974", + "accept": false, + "reason": "Mamba2 test OOM fix vs MusicGen generate typo; unrelated." + }, + { + "left": "pull_request:46968", + "right": "pull_request:46974", + "accept": false, + "reason": "Docs autolink cleanup vs MusicGen generate typo fix; unrelated." + }, + { + "left": "pull_request:46920", + "right": "pull_request:46961", + "accept": false, + "reason": "Voxtral Realtime API adaptation vs Serge workflow automation; unrelated." + }, + { + "left": "pull_request:46961", + "right": "pull_request:46968", + "accept": false, + "reason": "Branch-normalization workflow vs docs autolink cleanup; unrelated." + }, + { + "left": "pull_request:46958", + "right": "pull_request:46968", + "accept": false, + "reason": "CohereASR loss fix vs docs autolink cleanup; unrelated." + }, + { + "left": "pull_request:43451", + "right": "pull_request:46707", + "accept": false, + "reason": "Molmo2 model addition vs native FSDP2 support/migration; unrelated." + }, + { + "left": "pull_request:46266", + "right": "pull_request:46707", + "accept": false, + "reason": "DINOv2 modernization/conversion work vs native FSDP2 support; unrelated." + }, + { + "left": "pull_request:46196", + "right": "pull_request:46995", + "accept": false, + "reason": "Wav2Vec2 SDPA dispatch vs flash-linear-attention MLU availability; different acceleration backends." + }, + { + "left": "pull_request:46707", + "right": "pull_request:46961", + "accept": false, + "reason": "FSDP2 support/migration vs Serge workflow normalization; unrelated." + }, + { + "left": "pull_request:46891", + "right": "pull_request:46961", + "accept": false, + "reason": "PEFT integration CI vs Serge workflow normalization; different CI changes." + }, + { + "left": "pull_request:43451", + "right": "pull_request:46891", + "accept": false, + "reason": "Molmo2 model addition vs PEFT integration CI; unrelated." + }, + { + "left": "pull_request:46707", + "right": "pull_request:46977", + "accept": false, + "reason": "FSDP2 support/migration vs Mllama arg-name modernization; unrelated." + }, + { + "left": "pull_request:46891", + "right": "pull_request:46977", + "accept": false, + "reason": "PEFT CI change vs Mllama arg-name cleanup; unrelated." + }, + { + "left": "pull_request:46707", + "right": "pull_request:46968", + "accept": false, + "reason": "FSDP2 support/migration vs docs autolink cleanup; unrelated." + }, + { + "left": "pull_request:46891", + "right": "pull_request:46968", + "accept": false, + "reason": "PEFT CI change vs docs autolink cleanup; unrelated." + }, + { + "left": "pull_request:43451", + "right": "pull_request:46836", + "accept": false, + "reason": "Molmo2 model addition vs Unlimited OCR model addition; separate feature PRs." + }, + { + "left": "pull_request:46912", + "right": "pull_request:46977", + "accept": false, + "reason": "RoPE test/implementation updates vs Mllama arg-name deprecation; unrelated." + }, + { + "left": "pull_request:40546", + "right": "pull_request:43451", + "accept": false, + "reason": "VibeVoice and Molmo2 are separate model additions." + }, + { + "left": "pull_request:46196", + "right": "pull_request:46965", + "accept": false, + "reason": "Wav2Vec2 SDPA dispatch vs DeepSeekV4 use_cache support; unrelated." + }, + { + "left": "pull_request:46836", + "right": "pull_request:46961", + "accept": false, + "reason": "Unlimited OCR model addition vs Serge workflow normalization; unrelated." + }, + { + "left": "pull_request:46836", + "right": "pull_request:46965", + "accept": false, + "reason": "Unlimited OCR model addition vs DeepSeekV4 use_cache support; unrelated." + }, + { + "left": "pull_request:46836", + "right": "pull_request:46977", + "accept": false, + "reason": "Unlimited OCR model addition vs Mllama arg-name cleanup; unrelated." + }, + { + "left": "pull_request:46196", + "right": "pull_request:46977", + "accept": false, + "reason": "Wav2Vec2 SDPA dispatch vs Mllama arg-name cleanup; unrelated." + }, + { + "left": "pull_request:46266", + "right": "pull_request:47014", + "accept": false, + "reason": "DINO modernization/conversion work vs T5 SDPA enablement; unrelated." + }, + { + "left": "pull_request:43451", + "right": "pull_request:46658", + "accept": false, + "reason": "Molmo2 model addition vs Step 3.7 Flash implementation; separate features." + }, + { + "left": "pull_request:43451", + "right": "pull_request:46965", + "accept": false, + "reason": "Molmo2 model addition vs DeepSeekV4 use_cache support; unrelated." + } + ] + }, + "evaluator_result": { + "accept": false, + "feedback": "The cluster summary is mostly grounded, but one soft-edge verdict is too strong: pull_request:46895 and pull_request:46958 are not clearly mergeable into a single PR, since they implement alternative fixes for the same CohereASR loss bug rather than complementary changes. The packet supports that they address the same issue, but not that they should be accepted as a mergeable pair. The confidence/\u2018only clear duplicate pair\u2019 framing is therefore overstated." + }, + "evaluator_used": true, + "retried": true, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 1, + "component_count": 35, + "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-40546-41", + "nodes": [ + "pull_request:43451", + "pull_request:44517", + "pull_request:45694", + "pull_request:46196", + "pull_request:46658", + "pull_request:46707", + "pull_request:46836", + "pull_request:46891", + "pull_request:46907", + "pull_request:46908", + "pull_request:46912", + "pull_request:46926", + "pull_request:46927", + "pull_request:46933", + "pull_request:46949", + "pull_request:46950", + "pull_request:46951", + "pull_request:46957", + "pull_request:46961", + "pull_request:46965", + "pull_request:46968", + "pull_request:46970", + "pull_request:46971", + "pull_request:46977", + "pull_request:46986", + "pull_request:46995", + "pull_request:47006", + "pull_request:47008", + "pull_request:47014" + ], + "soft_pairs": [ + "pull_request:46836|pull_request:46995", + "pull_request:46658|pull_request:46961", + "pull_request:46658|pull_request:46977", + "pull_request:44517|pull_request:46836", + "pull_request:43451|pull_request:46912", + "pull_request:45694|pull_request:46707", + "pull_request:44517|pull_request:47014", + "pull_request:44517|pull_request:46957", + "pull_request:45694|pull_request:46891", + "pull_request:43451|pull_request:46196", + "pull_request:43451|pull_request:47014", + "pull_request:43451|pull_request:46957", + "pull_request:46196|pull_request:46961", + "pull_request:46961|pull_request:47014", + "pull_request:46957|pull_request:46961", + "pull_request:46912|pull_request:46961", + "pull_request:46707|pull_request:47008", + "pull_request:44517|pull_request:46891", + "pull_request:46912|pull_request:47008", + "pull_request:46977|pull_request:47014", + "pull_request:46957|pull_request:46977", + "pull_request:47008|pull_request:47014", + "pull_request:45694|pull_request:47014", + "pull_request:46957|pull_request:47008", + "pull_request:45694|pull_request:46957", + "pull_request:45694|pull_request:46836", + "pull_request:46933|pull_request:47008", + "pull_request:46950|pull_request:46971", + "pull_request:46949|pull_request:46950", + "pull_request:46949|pull_request:46951", + "pull_request:46927|pull_request:46949", + "pull_request:46907|pull_request:46927", + "pull_request:46927|pull_request:46971", + "pull_request:46908|pull_request:46926", + "pull_request:46986|pull_request:47006", + "pull_request:46907|pull_request:46949", + "pull_request:46970|pull_request:47006", + "pull_request:46970|pull_request:46986", + "pull_request:46907|pull_request:47006", + "pull_request:46907|pull_request:46970", + "pull_request:46907|pull_request:46986", + "pull_request:46961|pull_request:46977", + "pull_request:43451|pull_request:46968", + "pull_request:43451|pull_request:46961", + "pull_request:43451|pull_request:46977", + "pull_request:46957|pull_request:46965", + "pull_request:46196|pull_request:47014", + "pull_request:46658|pull_request:46836", + "pull_request:46891|pull_request:47014", + "pull_request:46836|pull_request:47014", + "pull_request:46658|pull_request:47014", + "pull_request:46957|pull_request:47014", + "pull_request:46912|pull_request:47014", + "pull_request:46912|pull_request:46957" + ], + "prepared_review_unit_hash": "835858bae33c6e8900ca86540daafbe022ef924145a0879d499146e70dc05f32", + "component_budget": { + "node_count": 41, + "item_count": 41, + "soft_pair_count": 126, + "serialized_chars": 142820, + "estimated_input_tokens": 35705, + "estimated_eval_tokens": 71666 + }, + "budget": { + "node_count": 29, + "item_count": 29, + "soft_pair_count": 54, + "serialized_chars": 82287, + "estimated_input_tokens": 20572, + "estimated_eval_tokens": 41400 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "Most items are unrelated PRs spanning different models, docs, CI, or infrastructure. The only clear duplicate-like pair is the Gemma4 quantized embedder dtype-casting fix (46933 and 47008).", + "confidence": 0.97, + "canonical_issue_reason": null, + "canonical_pr_reason": "46933 is the better canonical PR: it directly fixes the Gemma4 dtype-casting bug, has the explicit issue target 46932, and is the more established root-fix candidate for the duplicate pair.", + "best_issue_reason": null, + "best_pr_reason": "46933 is the strongest overall PR candidate in this set because it is the only true duplicate anchor and the cleanest concrete code-path fix.", + "soft_edge_verdicts": [ + { + "left": "pull_request:46836", + "right": "pull_request:46995", + "accept": false, + "reason": "Different changes: Unlimited OCR model addition versus MLU support for flash linear attention. Only a broad transformer-model overlap." + }, + { + "left": "pull_request:46658", + "right": "pull_request:46961", + "accept": false, + "reason": "Step-3.7 Flash model build/conversion work is unrelated to a Serge branch-normalization workflow." + }, + { + "left": "pull_request:46658", + "right": "pull_request:46977", + "accept": false, + "reason": "New Step-3.7 Flash model tooling and an Mllama argument cleanup are unrelated." + }, + { + "left": "pull_request:44517", + "right": "pull_request:46836", + "accept": false, + "reason": "Two separate model additions (Qwen3-TTS vs Unlimited OCR); shared auto-registration plumbing is not enough." + }, + { + "left": "pull_request:43451", + "right": "pull_request:46912", + "accept": false, + "reason": "Molmo2 model addition versus RoPE test/implementation updates; not the same bug or change." + }, + { + "left": "pull_request:45694", + "right": "pull_request:46707", + "accept": false, + "reason": "TrainingArguments batch-size fix and native FSDP2 migration address different areas." + }, + { + "left": "pull_request:44517", + "right": "pull_request:47014", + "accept": false, + "reason": "Qwen3-TTS addition and T5 SDPA enablement are different model families and fixes." + }, + { + "left": "pull_request:44517", + "right": "pull_request:46957", + "accept": false, + "reason": "Model addition versus finegrained FP8 decode optimization; no concrete overlap." + }, + { + "left": "pull_request:45694", + "right": "pull_request:46891", + "accept": false, + "reason": "TrainingArguments batch-size behavior and PEFT CI workflow changes are unrelated." + }, + { + "left": "pull_request:43451", + "right": "pull_request:46196", + "accept": false, + "reason": "Molmo2 addition and wav2vec2 SDPA dispatch are unrelated." + }, + { + "left": "pull_request:43451", + "right": "pull_request:47014", + "accept": false, + "reason": "Molmo2 addition versus T5 SDPA enablement; different models and code paths." + }, + { + "left": "pull_request:43451", + "right": "pull_request:46957", + "accept": false, + "reason": "Model addition versus generation/kernel optimization; not the same underlying change." + }, + { + "left": "pull_request:46196", + "right": "pull_request:46961", + "accept": false, + "reason": "wav2vec2 SDPA dispatch and Serge workflow normalization are unrelated." + }, + { + "left": "pull_request:46961", + "right": "pull_request:47014", + "accept": false, + "reason": "Workflow normalization and T5 SDPA enablement have nothing in common beyond repository scope." + }, + { + "left": "pull_request:46957", + "right": "pull_request:46961", + "accept": false, + "reason": "Finegrained FP8 decode optimization versus workflow normalization; different code and intent." + }, + { + "left": "pull_request:46912", + "right": "pull_request:46961", + "accept": false, + "reason": "RoPE testing/behavior changes and workflow normalization are unrelated." + }, + { + "left": "pull_request:46707", + "right": "pull_request:47008", + "accept": false, + "reason": "FSDP2 migration and Gemma4 dtype-casting fix solve different problems." + }, + { + "left": "pull_request:44517", + "right": "pull_request:46891", + "accept": false, + "reason": "Qwen3-TTS model addition and PEFT CI workflow changes are unrelated." + }, + { + "left": "pull_request:46912", + "right": "pull_request:47008", + "accept": false, + "reason": "RoPE test sweep versus Gemma4 dtype-casting fix; no shared concrete bug." + }, + { + "left": "pull_request:46977", + "right": "pull_request:47014", + "accept": false, + "reason": "Mllama arg-name cleanup and T5 SDPA enablement are unrelated." + }, + { + "left": "pull_request:46957", + "right": "pull_request:46977", + "accept": false, + "reason": "Finegrained FP8 decode changes versus Mllama argument cleanup; different bugs." + }, + { + "left": "pull_request:47008", + "right": "pull_request:47014", + "accept": false, + "reason": "Gemma4 dtype-casting fix and T5 SDPA enablement are unrelated." + }, + { + "left": "pull_request:45694", + "right": "pull_request:47014", + "accept": false, + "reason": "TrainingArguments batch-size fix and T5 SDPA enablement are unrelated." + }, + { + "left": "pull_request:46957", + "right": "pull_request:47008", + "accept": false, + "reason": "Finegrained FP8 decode optimization versus Gemma4 dtype-casting fix; different code paths." + }, + { + "left": "pull_request:45694", + "right": "pull_request:46957", + "accept": false, + "reason": "TrainingArguments batch-size fix and finegrained FP8 decode optimization are unrelated." + }, + { + "left": "pull_request:45694", + "right": "pull_request:46836", + "accept": false, + "reason": "TrainingArguments batch-size fix and Unlimited OCR model addition are unrelated." + }, + { + "left": "pull_request:46933", + "right": "pull_request:47008", + "accept": true, + "reason": "Same Gemma4 quantized embedder dtype-casting bug, same files, and shared issue target 46932; these could be merged into one PR." + }, + { + "left": "pull_request:46950", + "right": "pull_request:46971", + "accept": false, + "reason": "Both concern Mamba2 OOM, but one changes model code while the other only adjusts tests; not the same concrete fix." + }, + { + "left": "pull_request:46949", + "right": "pull_request:46950", + "accept": false, + "reason": "Whisper/AudioFlamingo cache behavior and Mamba2 OOM fixes are unrelated model bugs." + }, + { + "left": "pull_request:46949", + "right": "pull_request:46951", + "accept": false, + "reason": "Whisper/AudioFlamingo fix versus PEFT loading tweak; unrelated." + }, + { + "left": "pull_request:46927", + "right": "pull_request:46949", + "accept": false, + "reason": "GLM46V input-id casting fix and Whisper/AudioFlamingo cache fix are different code paths." + }, + { + "left": "pull_request:46907", + "right": "pull_request:46927", + "accept": false, + "reason": "Native FSDP2 migration and GLM46V input-id casting fix are unrelated." + }, + { + "left": "pull_request:46907", + "right": "pull_request:46949", + "accept": false, + "reason": "FSDP2 migration versus Whisper/AudioFlamingo cache fix; different problems." + }, + { + "left": "pull_request:46907", + "right": "pull_request:47006", + "accept": false, + "reason": "FSDP2 migration and Gemma3 generation-config fix do not address the same underlying change." + }, + { + "left": "pull_request:46907", + "right": "pull_request:46970", + "accept": false, + "reason": "FSDP2 migration and MusicGen Melody test fixes are unrelated." + }, + { + "left": "pull_request:46907", + "right": "pull_request:46986", + "accept": false, + "reason": "FSDP2 migration and GLM-OCR output mismatch fix are unrelated." + }, + { + "left": "pull_request:46961", + "right": "pull_request:46977", + "accept": false, + "reason": "Serge workflow normalization and Mllama arg-name cleanup are unrelated." + }, + { + "left": "pull_request:43451", + "right": "pull_request:46968", + "accept": false, + "reason": "Molmo2 model addition and docs autolink cleanup are unrelated." + }, + { + "left": "pull_request:43451", + "right": "pull_request:46961", + "accept": false, + "reason": "Molmo2 addition and workflow normalization are unrelated." + }, + { + "left": "pull_request:43451", + "right": "pull_request:46977", + "accept": false, + "reason": "Molmo2 model addition and Mllama arg-name cleanup are unrelated." + }, + { + "left": "pull_request:46957", + "right": "pull_request:46965", + "accept": false, + "reason": "Finegrained FP8 decode optimization and DeepSeek V4 use_cache=False fix solve different problems." + }, + { + "left": "pull_request:46196", + "right": "pull_request:47014", + "accept": false, + "reason": "wav2vec2 SDPA dispatch and T5 SDPA enablement are different model-family changes." + }, + { + "left": "pull_request:46658", + "right": "pull_request:46836", + "accept": false, + "reason": "Step-3.7 Flash model tooling and Unlimited OCR model addition are unrelated." + }, + { + "left": "pull_request:46891", + "right": "pull_request:47014", + "accept": false, + "reason": "PEFT CI workflow changes and T5 SDPA enablement are unrelated." + }, + { + "left": "pull_request:46836", + "right": "pull_request:47014", + "accept": false, + "reason": "Unlimited OCR model addition and T5 SDPA enablement are unrelated." + }, + { + "left": "pull_request:46658", + "right": "pull_request:47014", + "accept": false, + "reason": "Step-3.7 Flash model tooling and T5 SDPA enablement are unrelated." + }, + { + "left": "pull_request:46957", + "right": "pull_request:47014", + "accept": false, + "reason": "Finegrained FP8 decode optimization and T5 SDPA enablement are unrelated." + }, + { + "left": "pull_request:46912", + "right": "pull_request:47014", + "accept": false, + "reason": "RoPE test sweep and T5 SDPA enablement are unrelated." + }, + { + "left": "pull_request:46912", + "right": "pull_request:46957", + "accept": false, + "reason": "RoPE test sweep and finegrained FP8 decode optimization are unrelated." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 2, + "component_count": 35, + "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": "1f05ae78bccab1e6899c054e9cade9dc22e4581f6982d232ab68f10daf58eecb", + "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": "The two PRs are unrelated feature additions: TimesFM covariate forecasting vs. DINOv3 image classification support. The clustering is a false positive based only on loose similarity.", + "confidence": 0.98, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": "No single PR is a good cluster representative because the PRs target different models, code paths, and user-facing features.", + "soft_edge_verdicts": [ + { + "left": "pull_request:40755", + "right": "pull_request:41224", + "accept": false, + "reason": "Different underlying changes: 40755 adds covariate-based forecasting for TimesFM, while 41224 adds DINOv3ViT image classification support. They do not fix the same bug or implement the same 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": 3, + "component_count": 35, + "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": 4, + "component_count": 35, + "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-41855-13", + "nodes": [ + "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", + "pull_request:46267" + ], + "soft_pairs": [ + "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:46267", + "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": "8ae4c08b65f288f617f07f046eafa22a396d4d4831dc67154054fb3e33eaab9e", + "component_budget": { + "node_count": 13, + "item_count": 13, + "soft_pair_count": 18, + "serialized_chars": 34787, + "estimated_input_tokens": 8697, + "estimated_eval_tokens": 17650 + }, + "budget": { + "node_count": 13, + "item_count": 13, + "soft_pair_count": 18, + "serialized_chars": 34787, + "estimated_input_tokens": 8697, + "estimated_eval_tokens": 17650 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "This set is mostly unrelated open PRs spanning different models, docs, and utilities. The only true duplicate-looking pair is the two `fixed_cross_entropy` changes (43251/43254), which address the same loss-helper bug and could plausibly be merged into one PR.", + "confidence": 0.9, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 43251 is the clearest representative of the duplicate loss-helper fix: it has the explicit issue target and directly forwards the new kwargs into `nn.functional.cross_entropy`. PR 43254 is an alternate, near-identical implementation of the same underlying change.", + "best_issue_reason": null, + "best_pr_reason": "PR 43251 is the best overall representative of the only duplicate pair in this set; it is the most explicit and complete version of the cross-entropy kwargs fix, while the rest of the PRs are unrelated to that bug.", + "soft_edge_verdicts": [ + { + "left": "pull_request:41855", + "right": "pull_request:43823", + "accept": false, + "reason": "Tokenizer API additions for Mistral are unrelated to adding the MobileLLM model; different code paths and different changes." + }, + { + "left": "pull_request:41855", + "right": "pull_request:43424", + "accept": false, + "reason": "Mistral tokenizer methods vs ExecuTorch dynamic-shape export tests; no shared bug or change." + }, + { + "left": "pull_request:42668", + "right": "pull_request:43823", + "accept": false, + "reason": "Processor-auto robustness work is unrelated to introducing MobileLLM; broad infra overlap only." + }, + { + "left": "pull_request:43251", + "right": "pull_request:43823", + "accept": false, + "reason": "Loss-helper kwargs fix and MobileLLM model addition are unrelated." + }, + { + "left": "pull_request:43251", + "right": "pull_request:44827", + "accept": false, + "reason": "Cross-entropy kwargs fix and Mistral4 test/model changes are different issues." + }, + { + "left": "pull_request:43424", + "right": "pull_request:44827", + "accept": false, + "reason": "ExecuTorch dynamic-shape export test is unrelated to Mistral4 test/fix work." + }, + { + "left": "pull_request:43424", + "right": "pull_request:43823", + "accept": false, + "reason": "Exportability test work does not match the MobileLLM model addition." + }, + { + "left": "pull_request:43823", + "right": "pull_request:45655", + "accept": false, + "reason": "MobileLLM addition and config-resolution-order fix are unrelated." + }, + { + "left": "pull_request:43823", + "right": "pull_request:43995", + "accept": false, + "reason": "New model addition vs Falcon output-interface refactor; different model families and code paths." + }, + { + "left": "pull_request:43995", + "right": "pull_request:45213", + "accept": false, + "reason": "Falcon refactor and the model-creation skill are not the same change; one is model code, the other is tooling/docs." + }, + { + "left": "pull_request:43995", + "right": "pull_request:44827", + "accept": false, + "reason": "Falcon output interface refactor is unrelated to Mistral4 test fixes." + }, + { + "left": "pull_request:44827", + "right": "pull_request:45170", + "accept": false, + "reason": "Mistral4 test/model fixes and CLIP-like `layrnorm` renaming are unrelated changes." + }, + { + "left": "pull_request:44827", + "right": "pull_request:46267", + "accept": false, + "reason": "Mistral4 fixes and serve chat-parsing updates address different subsystems and bugs." + }, + { + "left": "pull_request:44827", + "right": "pull_request:46048", + "accept": false, + "reason": "Mistral4 test/model work is unrelated to adding the TDT loss kernel." + }, + { + "left": "pull_request:44827", + "right": "pull_request:45655", + "accept": false, + "reason": "Mistral4 changes and the `cls.config` resolution order fix are unrelated." + }, + { + "left": "pull_request:44827", + "right": "pull_request:45213", + "accept": false, + "reason": "Model/test fixes for Mistral4 are unrelated to the new-model creation skill." + }, + { + "left": "pull_request:43823", + "right": "pull_request:45213", + "accept": false, + "reason": "MobileLLM model addition is not the same as the model-creation skill/documentation PR." + }, + { + "left": "pull_request:43251", + "right": "pull_request:43254", + "accept": true, + "reason": "Both PRs fix the same `fixed_cross_entropy` code path by adding supported kwargs to `nn.functional.cross_entropy`; they are near-identical implementations of the same underlying bug." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet and stays conservative. The only supported duplicate pair is 43251/43254, which shares the same issue target and the same `fixed_cross_entropy` file. The other pair verdicts correctly reject soft similarity alone as insufficient evidence, and the representative PR choice is reasonable." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 5, + "component_count": 35, + "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": "4dad9fc1980a7febe22cd8adea155ec5c51697fcf34f4ec449197a9ba850601b", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 5934, + "estimated_input_tokens": 1484, + "estimated_eval_tokens": 3224 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 5934, + "estimated_input_tokens": 1484, + "estimated_eval_tokens": 3224 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Two open PRs both update GitHub Actions workflow dependencies, but they target different upgrade sets: #42975 is a broad Node 24 compatibility sweep across many workflows, while #42976 bumps Docker-related actions and a few other workflow actions to latest versions. They overlap in workflow files but not in the concrete change being made.", + "confidence": 0.95, + "canonical_issue_reason": null, + "canonical_pr_reason": "No clear duplicate canonical PR: #42975 is a broad Actions/node-compatibility bump, while #42976 is a separate action-version refresh centered on docker/build-push-action and related workflows.", + "best_issue_reason": null, + "best_pr_reason": "If one must be treated as the representative maintenance PR, #42975 is the broader workflow-actions upgrade, but it is not a true duplicate of #42976.", + "soft_edge_verdicts": [ + { + "left": "pull_request:42975", + "right": "pull_request:42976", + "accept": false, + "reason": "Different underlying changes: #42975 upgrades many Actions for Node 24 compatibility; #42976 mainly bumps docker/build-push-action and related workflow action versions. Shared workflow filenames alone are not enough to treat them as the same 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": 6, + "component_count": 35, + "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-42978-4", + "nodes": [ + "pull_request:42978", + "pull_request:43448", + "pull_request:45490", + "pull_request:46562" + ], + "soft_pairs": [ + "pull_request:42978|pull_request:43448", + "pull_request:42978|pull_request:45490", + "pull_request:43448|pull_request:45490", + "pull_request:45490|pull_request:46562" + ], + "prepared_review_unit_hash": "fffca1081be18c7713c1072bc9560bae860d57696874fd494b4b760f4509daee", + "component_budget": { + "node_count": 4, + "item_count": 4, + "soft_pair_count": 4, + "serialized_chars": 13047, + "estimated_input_tokens": 3262, + "estimated_eval_tokens": 6780 + }, + "budget": { + "node_count": 4, + "item_count": 4, + "soft_pair_count": 4, + "serialized_chars": 13047, + "estimated_input_tokens": 3262, + "estimated_eval_tokens": 6780 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "All four PRs are distinct model-addition changes (ViT NEPA, Molmo, CTSM, and Step3p7). They only overlap in shared Transformers registration/docs scaffolding, not in the underlying model or code-path change, so none of the soft edges should be merged as duplicates.", + "confidence": 0.98, + "canonical_issue_reason": null, + "canonical_pr_reason": "No canonical PR: each item introduces a different model and implementation path, so there is no single duplicate representative.", + "best_issue_reason": null, + "best_pr_reason": "No global best PR for duplicate triage; the PRs are unrelated beyond shared boilerplate, so none is a suitable canonical duplicate target.", + "soft_edge_verdicts": [ + { + "left": "pull_request:42978", + "right": "pull_request:43448", + "accept": false, + "reason": "ViT NEPA and Molmo are different model additions with different filenames, configs, and tests; shared auto/docs files are just boilerplate." + }, + { + "left": "pull_request:42978", + "right": "pull_request:45490", + "accept": false, + "reason": "ViT NEPA is a vision model PR, while CTSM is a time-series model PR; they touch different implementation paths and are not the same change." + }, + { + "left": "pull_request:43448", + "right": "pull_request:45490", + "accept": false, + "reason": "Molmo and CTSM add unrelated models with different components (multimodal vs time-series); overlap is only in shared repository wiring." + }, + { + "left": "pull_request:45490", + "right": "pull_request:46562", + "accept": false, + "reason": "CTSM and Step3p7 are different model families and code paths; the shared auto/docs edits do not indicate the same underlying bug 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": 7, + "component_count": 35, + "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-43291-7", + "nodes": [ + "pull_request:43291", + "pull_request:43665", + "pull_request:44815", + "pull_request:46380", + "pull_request:46539", + "pull_request:46886", + "pull_request:46922" + ], + "soft_pairs": [ + "pull_request:43291|pull_request:43665", + "pull_request:43291|pull_request:44815", + "pull_request:43291|pull_request:46380", + "pull_request:43291|pull_request:46922", + "pull_request:43665|pull_request:46539", + "pull_request:43665|pull_request:44815", + "pull_request:43665|pull_request:46380", + "pull_request:43665|pull_request:46922", + "pull_request:44815|pull_request:46539", + "pull_request:44815|pull_request:46380", + "pull_request:44815|pull_request:46922", + "pull_request:46380|pull_request:46539", + "pull_request:46380|pull_request:46922", + "pull_request:46539|pull_request:46922", + "pull_request:46886|pull_request:46922" + ], + "prepared_review_unit_hash": "80596f872106e6d76579de6324d81f6daf1d51d77829f586cd8a5c775a326494", + "component_budget": { + "node_count": 7, + "item_count": 7, + "soft_pair_count": 15, + "serialized_chars": 19594, + "estimated_input_tokens": 4899, + "estimated_eval_tokens": 10054 + }, + "budget": { + "node_count": 7, + "item_count": 7, + "soft_pair_count": 15, + "serialized_chars": 19594, + "estimated_input_tokens": 4899, + "estimated_eval_tokens": 10054 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "This cluster is mostly unrelated PRs across different models and subsystems. The only true duplicate/same-change pair is the YOLOS device-mismatch fix in PRs 46886 and 46922; the rest should stay separate.", + "confidence": 0.97, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 46922 is the best canonical representative for the YOLOS fix because it contains the concrete forward-path device move plus the split-module adjustment, and it matches the explicit issue target shared with 46886.", + "best_issue_reason": null, + "best_pr_reason": "PR 46922 is the strongest standalone representative: it directly fixes the YOLOS device mismatch in the model code and appears to be the more complete version of the same change as 46886.", + "soft_edge_verdicts": [ + { + "left": "pull_request:43291", + "right": "pull_request:43665", + "accept": false, + "reason": "Both are test-only PRs, but they target different models and different failures; no shared concrete bug or change." + }, + { + "left": "pull_request:43291", + "right": "pull_request:44815", + "accept": false, + "reason": "Whisper test cleanup vs FP8/dequant loading logic; unrelated code paths." + }, + { + "left": "pull_request:43291", + "right": "pull_request:46380", + "accept": false, + "reason": "Whisper test fix vs generation/watermarking config fix; different underlying issues." + }, + { + "left": "pull_request:43291", + "right": "pull_request:46922", + "accept": false, + "reason": "Whisper tokenizer/tests and YOLOS device placement are unrelated bugs." + }, + { + "left": "pull_request:43665", + "right": "pull_request:44815", + "accept": false, + "reason": "Test assertions in CLIP/ViT are unrelated to FP8/dequant conversion changes." + }, + { + "left": "pull_request:43665", + "right": "pull_request:46380", + "accept": false, + "reason": "Different model test edits and different product bug; not the same change." + }, + { + "left": "pull_request:43665", + "right": "pull_request:46539", + "accept": false, + "reason": "Failing model tests vs CI memory/device-map infrastructure fix; different underlying problem." + }, + { + "left": "pull_request:43665", + "right": "pull_request:46922", + "accept": false, + "reason": "Unrelated model test modifications vs YOLOS inference/device-mapping fix." + }, + { + "left": "pull_request:44815", + "right": "pull_request:46539", + "accept": false, + "reason": "FP8/dequant loading behavior and CI memory limiting are distinct issues." + }, + { + "left": "pull_request:44815", + "right": "pull_request:46380", + "accept": false, + "reason": "Different subsystems: quantization/dequant vs generation config validation." + }, + { + "left": "pull_request:44815", + "right": "pull_request:46922", + "accept": false, + "reason": "FP8 quantization/dequant changes are not the same as YOLOS device mismatch." + }, + { + "left": "pull_request:46380", + "right": "pull_request:46539", + "accept": false, + "reason": "Generation config watermarking fix and CI Phimoe integration fix are unrelated." + }, + { + "left": "pull_request:46380", + "right": "pull_request:46922", + "accept": false, + "reason": "Generation validation changes do not match the YOLOS device-mismatch code path." + }, + { + "left": "pull_request:46539", + "right": "pull_request:46922", + "accept": false, + "reason": "CI memory/device-map test harness changes are unrelated to YOLOS model placement bugs." + }, + { + "left": "pull_request:46886", + "right": "pull_request:46922", + "accept": true, + "reason": "Same YOLOS device-mismatch issue, same model file, and same explicit issue target; 46922 is essentially the more complete version of the same fix." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The cluster summary is grounded in the packet, and the pairwise rejections are conservative. The YOLOS pair (46886, 46922) is the only clearly mergeable same-bug fix pair, with shared file and issue target; the other PRs address unrelated models or subsystems." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 8, + "component_count": 35, + "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-43488-3", + "nodes": [ + "pull_request:43488", + "pull_request:43532", + "pull_request:46747" + ], + "soft_pairs": [ + "pull_request:43488|pull_request:46747", + "pull_request:43488|pull_request:43532", + "pull_request:43532|pull_request:46747" + ], + "prepared_review_unit_hash": "2015cede188442ba92047922013946fdc638ddefa79014eaa3c975b92333ee76", + "component_budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 3, + "serialized_chars": 6349, + "estimated_input_tokens": 1588, + "estimated_eval_tokens": 3432 + }, + "budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 3, + "serialized_chars": 6349, + "estimated_input_tokens": 1588, + "estimated_eval_tokens": 3432 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Three unrelated '[don't merge]' PRs for repo-bot/workflow testing; they have superficial similarity but change different files and behaviors.", + "confidence": 0.95, + "canonical_issue_reason": null, + "canonical_pr_reason": "No canonical PR: the PRs touch different areas (setup/deps and tests, comment CI workflows, security-gate workflow) and are not the same fix.", + "best_issue_reason": null, + "best_pr_reason": "No best PR for deduping: none of these is a plausible merge target for the others, despite the shared bot/test-oriented framing.", + "soft_edge_verdicts": [ + { + "left": "pull_request:43488", + "right": "pull_request:46747", + "accept": false, + "reason": "Different changes: 43488 updates setup/dependencies and a test, while 46747 only changes the PR security-gate workflow. Not the same underlying bug/change." + }, + { + "left": "pull_request:43488", + "right": "pull_request:43532", + "accept": false, + "reason": "43488 is a setup/dependency/version-bump test PR; 43532 rewires comment/scheduled/slack workflows and test utilities. Shared 'don't merge' framing is too broad." + }, + { + "left": "pull_request:43532", + "right": "pull_request:46747", + "accept": false, + "reason": "43532 alters multiple workflow files plus testing code; 46747 only switches the security-gate trigger from pull_request_target to pull_request. Distinct code paths." + } + ] + }, + "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": 35, + "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": "b00106de8ae0d5471d052ae341518a25ed4869cb78898102ae2d754aee51eb9a", + "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": "These are two separate feature PRs for different segmentation pipelines; they share scaffolding and docs but do not fix the same concrete problem or implement the same change.", + "confidence": 0.97, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "pull_request:43612", + "right": "pull_request:43613", + "accept": false, + "reason": "Different pipeline features: Promptable Concept Segmentation vs Promptable Visual Segmentation. The overlap is mostly template/docs/shared infra, not the same code-path change, so 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": "pull_request", + "component_index": 10, + "component_count": 35, + "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": "ab42c884a4ba5ec065ba641bc05b1b96391fb119417b6b46b650a95ec59a84d2", + "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": "Both items are open PRs in `src/transformers/trainer.py`, but they address different trainer behaviors: PR 43651 changes gradient-accumulation loss scaling, while PR 43663 adds an overridable helper for signature-column filtering. They are related by file location only, not by the same bug or change.", + "confidence": 0.97, + "canonical_issue_reason": null, + "canonical_pr_reason": "No single canonical PR: these are distinct trainer changes with different code paths and goals, so neither can stand in for the other as a duplicate.", + "best_issue_reason": null, + "best_pr_reason": "PR 43651 is the more concrete, narrowly scoped behavior change, but it is still not a duplicate of PR 43663; keeping it separate best reflects the underlying work.", + "soft_edge_verdicts": [ + { + "left": "pull_request:43651", + "right": "pull_request:43663", + "accept": false, + "reason": "Do not merge as duplicates: 43651 changes loss normalization during training_step, while 43663 refactors signature-column detection for column filtering. Same file, different functionality, different code paths." + } + ] + }, + "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": 35, + "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": "a830b2bd6c70b20253b079248963a9567f31ca65aa2d28c4060a6bdb536100a6", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 3663, + "estimated_input_tokens": 916, + "estimated_eval_tokens": 2088 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 3663, + "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 both touch `src/transformers/cli/serve.py` and both mention a `TypeAdapter` NameError, but they target different issues and appear to use different fixes for different trigger conditions. They should not be merged as duplicates.", + "confidence": 0.92, + "canonical_issue_reason": null, + "canonical_pr_reason": "No canonical PR: the PRs address closely related but distinct `TypeAdapter`/typing problems in the CLI, with different issue targets and different corrective approaches.", + "best_issue_reason": null, + "best_pr_reason": "Neither PR is a clear global winner because each is narrowly scoped to a different reported bug; keep both as separate PRs.", + "soft_edge_verdicts": [ + { + "left": "pull_request:43656", + "right": "pull_request:43842", + "accept": false, + "reason": "Reject: despite the shared file and similar wording, these point to different underlying bugs/conditions (one is a general CLI `TypeAdapter` NameError fix, the other is specifically when `pydantic` is not installed) and the diffs show different remedies." + } + ] + }, + "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": 35, + "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": 13, + "component_count": 35, + "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:44722", + "pull_request:44018|pull_request:44116", + "pull_request:44068|pull_request:44116", + "pull_request:44071|pull_request:44722", + "pull_request:44018|pull_request:44085", + "pull_request:44018|pull_request:44086", + "pull_request:44068|pull_request:44086", + "pull_request:44071|pull_request:44085", + "pull_request:44085|pull_request:44086", + "pull_request:44059|pull_request:44086", + "pull_request:44098|pull_request:44101", + "pull_request:44018|pull_request:44071", + "pull_request:44059|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:44071|pull_request:44086", + "pull_request:44017|pull_request:44071", + "pull_request:43997|pull_request:44073", + "pull_request:44073|pull_request:44101", + "pull_request:44073|pull_request:44161", + "pull_request:44059|pull_request:44071", + "pull_request:44018|pull_request:44059", + "pull_request:44056|pull_request:44101", + "pull_request:44059|pull_request:44068", + "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:43997|pull_request:44018", + "pull_request:44056|pull_request:44161", + "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:44018|pull_request:44129", + "pull_request:44010|pull_request:44073", + "pull_request:43997|pull_request:44074", + "pull_request:44010|pull_request:44018", + "pull_request:44071|pull_request:44129", + "pull_request:43997|pull_request:44129", + "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": "bd1bd52f02e9da9a8f849ea57eede3e79eac423dc274c17007d6f18ecdea5661", + "component_budget": { + "node_count": 28, + "item_count": 28, + "soft_pair_count": 78, + "serialized_chars": 84894, + "estimated_input_tokens": 21224, + "estimated_eval_tokens": 42704 + }, + "budget": { + "node_count": 27, + "item_count": 27, + "soft_pair_count": 72, + "serialized_chars": 75088, + "estimated_input_tokens": 18772, + "estimated_eval_tokens": 37800 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "Mostly a heterogeneous set of model-specific output-tracing refactors. The only meaningful duplicate candidates are the GPT-J/CodeGen variants; the rest touch different model code paths and should stay separate.", + "confidence": 0.73, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 44066 is the best canonical PR for the shared GPT-J output-tracing work: it is explicitly issue-linked, covers the core GPT-J change, and is the broadest representative of the overlapping GPT-J/CodeGen refactor family.", + "best_issue_reason": null, + "best_pr_reason": "PR 44066 is the best overall representative in this cluster because it is the clearest, issue-linked example of the shared standardized output-tracing refactor and overlaps with the other GPT-J variants.", + "soft_edge_verdicts": [ + { + "left": "pull_request:43996", + "right": "pull_request:44085", + "accept": false, + "reason": "Different models and code paths (CVT/FNet vs GPT-J). Same broad refactor theme is not enough for duplication." + }, + { + "left": "pull_request:43996", + "right": "pull_request:44044", + "accept": false, + "reason": "CVT/FNet output-tracing cleanup is unrelated to DeBERTa v2's attention/output refactor." + }, + { + "left": "pull_request:44066", + "right": "pull_request:44085", + "accept": true, + "reason": "Both are GPT-J output-tracing refactors in the same file and appear to be the same underlying decorator migration work." + }, + { + "left": "pull_request:44007", + "right": "pull_request:44072", + "accept": false, + "reason": "RegNet and EfficientNet are separate model implementations; these are similar migrations, not the same bug/change." + }, + { + "left": "pull_request:44072", + "right": "pull_request:44722", + "accept": false, + "reason": "EfficientNet vs GPT-J/CodeGen; different model code paths." + }, + { + "left": "pull_request:44066", + "right": "pull_request:44072", + "accept": false, + "reason": "GPT-J/CodeGen output tracing is unrelated to EfficientNet output handling." + }, + { + "left": "pull_request:44013", + "right": "pull_request:44044", + "accept": false, + "reason": "MobileNetV2 and DeBERTa v2 are different implementations with different tracing paths." + }, + { + "left": "pull_request:44066", + "right": "pull_request:44086", + "accept": false, + "reason": "GPT-J/CodeGen versus MGP-STR; same pattern, different code paths." + }, + { + "left": "pull_request:44066", + "right": "pull_request:44071", + "accept": false, + "reason": "GPT-J/CodeGen and MPT are separate refactors, not one concrete change." + }, + { + "left": "pull_request:44018", + "right": "pull_request:44068", + "accept": false, + "reason": "GPT2 and GPT-Neo are distinct models; these are parallel refactors, not duplicates." + }, + { + "left": "pull_request:44066", + "right": "pull_request:44068", + "accept": false, + "reason": "GPT-J/CodeGen output-tracing work is separate from GPT-Neo's refactor." + }, + { + "left": "pull_request:44018", + "right": "pull_request:44066", + "accept": false, + "reason": "GPT2 and GPT-J/CodeGen are different model paths." + }, + { + "left": "pull_request:44019", + "right": "pull_request:44722", + "accept": false, + "reason": "ResNet output tracing is unrelated to GPT-J/CodeGen." + }, + { + "left": "pull_request:44086", + "right": "pull_request:44722", + "accept": false, + "reason": "MGP-STR and GPT-J/CodeGen are different model families." + }, + { + "left": "pull_request:44019", + "right": "pull_request:44085", + "accept": false, + "reason": "ResNet and GPT-J are unrelated code paths." + }, + { + "left": "pull_request:44018", + "right": "pull_request:44019", + "accept": false, + "reason": "GPT2 and ResNet are entirely different implementations." + }, + { + "left": "pull_request:44019", + "right": "pull_request:44071", + "accept": false, + "reason": "ResNet versus MPT; not the same bug/change." + }, + { + "left": "pull_request:44068", + "right": "pull_request:44722", + "accept": false, + "reason": "GPT-Neo and GPT-J/CodeGen are different model code paths." + }, + { + "left": "pull_request:44018", + "right": "pull_request:44116", + "accept": false, + "reason": "GPT2 and Flaubert are different model families." + }, + { + "left": "pull_request:44068", + "right": "pull_request:44116", + "accept": false, + "reason": "GPT-Neo and Flaubert are unrelated model implementations." + }, + { + "left": "pull_request:44071", + "right": "pull_request:44722", + "accept": false, + "reason": "MPT and GPT-J/CodeGen are separate refactors." + }, + { + "left": "pull_request:44018", + "right": "pull_request:44085", + "accept": false, + "reason": "GPT2 versus GPT-J; different concrete code paths." + }, + { + "left": "pull_request:44018", + "right": "pull_request:44086", + "accept": false, + "reason": "GPT2 and MGP-STR are unrelated models." + }, + { + "left": "pull_request:44068", + "right": "pull_request:44086", + "accept": false, + "reason": "GPT-Neo and MGP-STR do not share the same underlying change." + }, + { + "left": "pull_request:44071", + "right": "pull_request:44085", + "accept": false, + "reason": "MPT versus GPT-J; same pattern, different implementation." + }, + { + "left": "pull_request:44085", + "right": "pull_request:44086", + "accept": false, + "reason": "GPT-J and MGP-STR are distinct model code paths." + }, + { + "left": "pull_request:44059", + "right": "pull_request:44086", + "accept": false, + "reason": "GPT2 output-tracing refactor is unrelated to MGP-STR." + }, + { + "left": "pull_request:44098", + "right": "pull_request:44101", + "accept": false, + "reason": "ViLT and Flaubert/XLM are separate model-specific changes." + }, + { + "left": "pull_request:44017", + "right": "pull_request:44161", + "accept": false, + "reason": "SegFormer and LongT5 are unrelated implementations." + }, + { + "left": "pull_request:44017", + "right": "pull_request:44722", + "accept": false, + "reason": "SegFormer output tracing is unrelated to GPT-J/CodeGen." + }, + { + "left": "pull_request:44017", + "right": "pull_request:44073", + "accept": false, + "reason": "SegFormer and VisualBERT are different model paths." + }, + { + "left": "pull_request:44068", + "right": "pull_request:44071", + "accept": false, + "reason": "GPT-Neo and MPT are separate model refactors." + }, + { + "left": "pull_request:44071", + "right": "pull_request:44076", + "accept": false, + "reason": "MPT and ImageGPT are different code paths." + }, + { + "left": "pull_request:44073", + "right": "pull_request:44101", + "accept": false, + "reason": "VisualBERT and Flaubert/XLM are unrelated model families." + }, + { + "left": "pull_request:44074", + "right": "pull_request:44101", + "accept": false, + "reason": "TextNet and Flaubert/XLM are different model implementations." + }, + { + "left": "pull_request:43997", + "right": "pull_request:44017", + "accept": false, + "reason": "RegNet and SegFormer are different model code paths." + }, + { + "left": "pull_request:43997", + "right": "pull_request:44056", + "accept": false, + "reason": "RegNet and MPNet are unrelated refactors." + }, + { + "left": "pull_request:43997", + "right": "pull_request:44018", + "accept": false, + "reason": "RegNet and GPT2 are different models." + }, + { + "left": "pull_request:43996", + "right": "pull_request:44161", + "accept": false, + "reason": "CVT/FNet output tracing is unrelated to LongT5." + }, + { + "left": "pull_request:43996", + "right": "pull_request:44017", + "accept": false, + "reason": "CVT/FNet and SegFormer are different model implementations." + }, + { + "left": "pull_request:44010", + "right": "pull_request:44071", + "accept": false, + "reason": "SqueezeBERT and MPT are unrelated code paths." + }, + { + "left": "pull_request:44010", + "right": "pull_request:44073", + "accept": false, + "reason": "SqueezeBERT and VisualBERT are different model families." + }, + { + "left": "pull_request:43997", + "right": "pull_request:44074", + "accept": false, + "reason": "RegNet and TextNet are separate model-specific changes." + }, + { + "left": "pull_request:43997", + "right": "pull_request:44161", + "accept": false, + "reason": "RegNet and LongT5 are unrelated implementations." + }, + { + "left": "pull_request:44074", + "right": "pull_request:44161", + "accept": false, + "reason": "TextNet and LongT5 are different model code paths." + }, + { + "left": "pull_request:44018", + "right": "pull_request:44129", + "accept": false, + "reason": "GPT2 and SpeechT5 are unrelated changes." + }, + { + "left": "pull_request:44071", + "right": "pull_request:44129", + "accept": false, + "reason": "MPT and SpeechT5 are different model families." + }, + { + "left": "pull_request:44056", + "right": "pull_request:44161", + "accept": false, + "reason": "MPNet and LongT5 are separate refactors." + }, + { + "left": "pull_request:43996", + "right": "pull_request:44007", + "accept": false, + "reason": "CVT/FNet and RegNet are different model code paths." + }, + { + "left": "pull_request:44044", + "right": "pull_request:44129", + "accept": false, + "reason": "DeBERTa v2 and SpeechT5 are unrelated model implementations." + }, + { + "left": "pull_request:44044", + "right": "pull_request:44154", + "accept": false, + "reason": "DeBERTa v2 and VITS are different model families." + }, + { + "left": "pull_request:44066", + "right": "pull_request:44722", + "accept": true, + "reason": "Both touch GPT-J output tracing in the same core file and represent the same standardized decorator migration." + }, + { + "left": "pull_request:44085", + "right": "pull_request:44722", + "accept": true, + "reason": "Both are GPT-J output-tracing refactors on the same code path; they could plausibly be merged into one PR." + }, + { + "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": "CVT/FNet and RegNet are different model implementations." + }, + { + "left": "pull_request:43996", + "right": "pull_request:44129", + "accept": false, + "reason": "CVT/FNet and SpeechT5 are unrelated refactors." + }, + { + "left": "pull_request:44044", + "right": "pull_request:44129", + "accept": false, + "reason": "DeBERTa v2 and SpeechT5 are different models." + }, + { + "left": "pull_request:44044", + "right": "pull_request:44154", + "accept": false, + "reason": "DeBERTa v2 and VITS do not share the same underlying code change." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet and stays conservative about the non-GPT-J pairs. The three accepted GPT-J/CodeGen pairs are supported by shared issue target, shared file coverage, and clearly overlapping output-tracing refactors that look mergeable into one bugfix PR." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 13, + "component_count": 35, + "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": "e7b14536036cb341195189c03dc1e5142c65ebd98975f7c1699ba725f0edaf45", + "component_budget": { + "node_count": 28, + "item_count": 28, + "soft_pair_count": 78, + "serialized_chars": 84894, + "estimated_input_tokens": 21224, + "estimated_eval_tokens": 42704 + }, + "budget": { + "node_count": 8, + "item_count": 8, + "soft_pair_count": 6, + "serialized_chars": 17127, + "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 all model-specific PRs that migrate different architectures to standardized output tracing/capture_outputs patterns. They are thematically related, but they do not fix the same concrete code path and would not plausibly be merged into one PR.", + "confidence": 0.95, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 44073 is the clearest representative of the cluster: it explicitly refactors output tracing with both can_return_tuple and capture_outputs, plus _can_record_outputs wiring, making it the most complete and recognizable example.", + "best_issue_reason": null, + "best_pr_reason": "PR 44073 best represents the shared theme because its title and diff most directly describe the standardized output-tracing migration, while still remaining a single, concrete model implementation change.", + "soft_edge_verdicts": [ + { + "left": "pull_request:44024", + "right": "pull_request:44076", + "accept": false, + "reason": "Both are output-tracing refactors, but for different models (FocalNet vs ImageGPT). Same migration pattern, not the same underlying bug/change, and not mergeable into one PR." + }, + { + "left": "pull_request:44073", + "right": "pull_request:44074", + "accept": false, + "reason": "Both standardize output capture, but one is VisualBert and the other is TextNet. Different model code paths and separate implementation changes." + }, + { + "left": "pull_request:44056", + "right": "pull_request:44073", + "accept": false, + "reason": "Both deal with standardized output tracing, but MPNet and VisualBert are distinct model implementations with different hooks and outputs. Not the same concrete fix." + }, + { + "left": "pull_request:44056", + "right": "pull_request:44074", + "accept": false, + "reason": "Shared refactor theme only; MPNet and TextNet touch different model internals and are independent PRs." + }, + { + "left": "pull_request:43997", + "right": "pull_request:44010", + "accept": false, + "reason": "Both are model-level output tracing migrations, but RegNet and SqueezeBert are unrelated code paths. They do not appear to be duplicate fixes." + }, + { + "left": "pull_request:44010", + "right": "pull_request:44129", + "accept": false, + "reason": "Both relate to output capture/output tracing, but SqueezeBert and SpeechT5 are separate model changes; the SpeechT5 PR also includes its own cache/output signature adjustments." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 14, + "component_count": 35, + "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-43998-13", + "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" + ], + "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" + ], + "prepared_review_unit_hash": "7b141b41d230a881863fc2b9ac27b4ee34f1e247626620a30cd061e1defcbedd", + "component_budget": { + "node_count": 13, + "item_count": 13, + "soft_pair_count": 40, + "serialized_chars": 40183, + "estimated_input_tokens": 10046, + "estimated_eval_tokens": 20348 + }, + "budget": { + "node_count": 13, + "item_count": 13, + "soft_pair_count": 40, + "serialized_chars": 40183, + "estimated_input_tokens": 10046, + "estimated_eval_tokens": 20348 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These PRs are all model-specific refactors around output tracing/can_return_tuple, but they target different files and forward paths, so they are not duplicates of one another.", + "confidence": 0.97, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": "PR 44025 is the most representative anchor for the cluster because it is a fairly complete example of the shared output-tracing refactor pattern and spans two related model files, but it still does not subsume the other PRs as a duplicate.", + "soft_edge_verdicts": [ + { + "left": "pull_request:43998", + "right": "pull_request:44002", + "accept": false, + "reason": "Same refactor theme, but different model implementations and code paths; not the same concrete change." + }, + { + "left": "pull_request:43998", + "right": "pull_request:44000", + "accept": false, + "reason": "Both are output-tracing refactors, but they touch different models and forward paths." + }, + { + "left": "pull_request:43998", + "right": "pull_request:44001", + "accept": false, + "reason": "Same broad pattern, but separate model code paths; not plausibly one PR." + }, + { + "left": "pull_request:43998", + "right": "pull_request:44003", + "accept": false, + "reason": "Different architectures and files; only the refactor style matches." + }, + { + "left": "pull_request:43998", + "right": "pull_request:44029", + "accept": false, + "reason": "Different model family and code path; not the same underlying fix." + }, + { + "left": "pull_request:43999", + "right": "pull_request:44001", + "accept": false, + "reason": "Both are tracing refactors, but for different models and outputs." + }, + { + "left": "pull_request:43999", + "right": "pull_request:44002", + "accept": false, + "reason": "Different model implementations; no shared concrete bug/change." + }, + { + "left": "pull_request:43999", + "right": "pull_request:44025", + "accept": false, + "reason": "Same refactor pattern, but distinct model files and behavior." + }, + { + "left": "pull_request:43999", + "right": "pull_request:44026", + "accept": false, + "reason": "Different encoder-decoder path versus MobileNet; not duplicate changes." + }, + { + "left": "pull_request:43999", + "right": "pull_request:44027", + "accept": false, + "reason": "Different model family and forward logic; only superficially similar." + }, + { + "left": "pull_request:44000", + "right": "pull_request:44001", + "accept": false, + "reason": "Both are output-tracing refactors, but they affect unrelated model code paths." + }, + { + "left": "pull_request:44000", + "right": "pull_request:44002", + "accept": false, + "reason": "Same general theme, different model-specific implementation." + }, + { + "left": "pull_request:44000", + "right": "pull_request:44003", + "accept": false, + "reason": "Distinct model families and source files; not the same change." + }, + { + "left": "pull_request:44000", + "right": "pull_request:44025", + "accept": false, + "reason": "Refactor pattern matches, but the underlying code paths are different." + }, + { + "left": "pull_request:44001", + "right": "pull_request:44002", + "accept": false, + "reason": "Different models and forward paths; not mergeable as one fix." + }, + { + "left": "pull_request:44001", + "right": "pull_request:44003", + "accept": false, + "reason": "Separate model implementations; only the tracing refactor is shared." + }, + { + "left": "pull_request:44001", + "right": "pull_request:44025", + "accept": false, + "reason": "Same broad output-tracing cleanup, but for different models." + }, + { + "left": "pull_request:44002", + "right": "pull_request:44029", + "accept": false, + "reason": "Different architectures and code paths; not the same underlying bug/change." + }, + { + "left": "pull_request:44002", + "right": "pull_request:44003", + "accept": false, + "reason": "Only the refactor motif is shared; the actual model logic differs." + }, + { + "left": "pull_request:44003", + "right": "pull_request:44028", + "accept": false, + "reason": "Different model families and implementations; not a duplicate change." + }, + { + "left": "pull_request:44003", + "right": "pull_request:44004", + "accept": false, + "reason": "Mamba/Falcon-Mamba versus CodeGen are unrelated code paths despite similar refactoring." + }, + { + "left": "pull_request:43998", + "right": "pull_request:44004", + "accept": false, + "reason": "Different model code paths; the similarity is only in the output-tracing refactor pattern." + }, + { + "left": "pull_request:44001", + "right": "pull_request:44004", + "accept": false, + "reason": "Separate model implementations; not the same concrete bug or change." + }, + { + "left": "pull_request:44002", + "right": "pull_request:44004", + "accept": false, + "reason": "Same broad cleanup, but different model-specific forward logic." + }, + { + "left": "pull_request:44025", + "right": "pull_request:44026", + "accept": false, + "reason": "Both are output-tracing refactors, but they apply to different model stacks and code paths." + }, + { + "left": "pull_request:44025", + "right": "pull_request:44027", + "accept": false, + "reason": "Different encoder-decoder models; not the same underlying change." + }, + { + "left": "pull_request:44025", + "right": "pull_request:44028", + "accept": false, + "reason": "Same pattern, but SuperPoint and DepthAnything are unrelated implementations." + }, + { + "left": "pull_request:44025", + "right": "pull_request:44029", + "accept": false, + "reason": "Different model families and forward paths; not duplicate PRs." + }, + { + "left": "pull_request:44025", + "right": "pull_request:44030", + "accept": false, + "reason": "Different model-specific tracing changes; not one mergeable fix." + }, + { + "left": "pull_request:44026", + "right": "pull_request:44029", + "accept": false, + "reason": "Different model families and code paths; only the refactor theme matches." + }, + { + "left": "pull_request:44026", + "right": "pull_request:44027", + "accept": false, + "reason": "Different encoder-decoder implementations; not the same change." + }, + { + "left": "pull_request:44026", + "right": "pull_request:44028", + "accept": false, + "reason": "Distinct model code paths; not a duplicate fix." + }, + { + "left": "pull_request:44026", + "right": "pull_request:44030", + "accept": false, + "reason": "Different architectures and implementation details; same broad pattern only." + }, + { + "left": "pull_request:44027", + "right": "pull_request:44028", + "accept": false, + "reason": "Different model implementations and forward paths; not the same underlying change." + }, + { + "left": "pull_request:44027", + "right": "pull_request:44029", + "accept": false, + "reason": "Separate model families; only the output-tracing refactor is shared." + }, + { + "left": "pull_request:44027", + "right": "pull_request:44030", + "accept": false, + "reason": "Different code paths and model stacks; not duplicate PRs." + }, + { + "left": "pull_request:44028", + "right": "pull_request:44029", + "accept": false, + "reason": "Unrelated models and implementations; not the same fix." + }, + { + "left": "pull_request:44028", + "right": "pull_request:44030", + "accept": false, + "reason": "Different model-specific refactors; no shared concrete bug." + }, + { + "left": "pull_request:44029", + "right": "pull_request:44030", + "accept": false, + "reason": "Different architectures and forward paths; only the refactor pattern is common." + } + ] + }, + "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": 35, + "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-44662-2", + "nodes": [ + "pull_request:44662", + "pull_request:45401" + ], + "soft_pairs": [ + "pull_request:44662|pull_request:45401" + ], + "prepared_review_unit_hash": "3c1c968d7a6051791b1fdb57461c0e4df85bef375e0e79f46e0c11b289ad00ef", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 6552, + "estimated_input_tokens": 1638, + "estimated_eval_tokens": 3532 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 6552, + "estimated_input_tokens": 1638, + "estimated_eval_tokens": 3532 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Both items are model-addition PRs that touch common Transformers registration/docs code, but they implement different models and different code paths: PenguinVL vs Voxtral TTS. They should not be deduplicated.", + "confidence": 0.98, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR #44662 is the better representative: it is a more mature, non-draft model implementation with broader review activity, while #45401 is a draft and explicitly tied to a specific support request.", + "best_issue_reason": null, + "best_pr_reason": "PR #44662 is the strongest standalone PR in the set because it contains a full model implementation with tests and documentation, whereas #45401 is narrower, draft, and targeted at a specific Voxtral TTS issue.", + "soft_edge_verdicts": [ + { + "left": "pull_request:44662", + "right": "pull_request:45401", + "accept": false, + "reason": "Reject: these are different model integrations (PenguinVL vs Voxtral TTS). Overlap is limited to shared Transformers plumbing and docs registration, 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": 16, + "component_count": 35, + "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-44965-2", + "nodes": [ + "pull_request:44965", + "pull_request:46845" + ], + "soft_pairs": [ + "pull_request:44965|pull_request:46845" + ], + "prepared_review_unit_hash": "8aeb0d137a7a1cc533e3f9d0296ec7113e58328c8c9f859252e37885529ee134", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4220, + "estimated_input_tokens": 1055, + "estimated_eval_tokens": 2366 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4220, + "estimated_input_tokens": 1055, + "estimated_eval_tokens": 2366 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These two pull requests are not duplicates: one is about changing a PR documentation workflow to convert PRs to draft, while the other alters CI branch routing and makes unrelated edits in model code/tests. They do not fix the same concrete problem and would not plausibly merge into one PR.", + "confidence": 0.96, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": "PR 46845 is the more coherent and substantive change of the two, with a clear CI-related intent; PR 44965 looks like an unrelated workflow experiment and is not a better canonical representative.", + "soft_edge_verdicts": [ + { + "left": "pull_request:44965", + "right": "pull_request:46845", + "accept": false, + "reason": "Different changes with different goals: draft-conversion workflow vs CI branch/config tweaks plus unrelated code edits. They do not appear to address 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": 17, + "component_count": 35, + "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": "23d13e5ce55be737676cce267bb69eb01557697a8a4b2be58a285754ad2c767f", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4406, + "estimated_input_tokens": 1102, + "estimated_eval_tokens": 2460 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4406, + "estimated_input_tokens": 1102, + "estimated_eval_tokens": 2460 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are related fusion-feature PRs, but they target different layers: one introduces a generic module-fusion API, while the other wires up n-to-1 kernel fusion through `KernelConfig` and model loading. They are not the same concrete change and should not be deduped.", + "confidence": 0.86, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR #45363 is the more concrete, user-facing implementation: it threads kernel fusion through `KernelConfig`, `modeling_utils`, and integration code, making it the better representative of the cluster.", + "best_issue_reason": null, + "best_pr_reason": "PR #45363 is the strongest standalone candidate because it defines the actual end-to-end fusion behavior, whereas PR #44979 is mainly a lower-level API scaffold.", + "soft_edge_verdicts": [ + { + "left": "pull_request:44979", + "right": "pull_request:45363", + "accept": false, + "reason": "Both concern fusion, but #44979 adds a generic module-fusion API and tests, while #45363 implements n-to-1 kernel fusion via `KernelConfig` across loading/integration paths. Different concrete code paths, so not a duplicate." + } + ] + }, + "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": 35, + "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": "e57a0c6f49d394c5745f777e63c3168e8fc6a0a513035b7f892cf30b7c104d8d", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4413, + "estimated_input_tokens": 1104, + "estimated_eval_tokens": 2464 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4413, + "estimated_input_tokens": 1104, + "estimated_eval_tokens": 2464 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Both pull requests address the same concrete bug in `_patch_mistral_regex`: an `AttributeError` caused by using `tokenizer.backend_tokenizer.pre_tokenizer` instead of `tokenizer.pre_tokenizer`. The second PR is a more complete version with tests.", + "confidence": 0.98, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 45317 is the stronger canonical PR because it fixes the same bug as 45086 but also adds test coverage, making it the more complete and reviewable patch.", + "best_issue_reason": null, + "best_pr_reason": "PR 45317 is the best PR choice because it fully addresses the bug and includes tests, while 45086 is just a smaller variant of the same fix.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45086", + "right": "pull_request:45317", + "accept": true, + "reason": "Same underlying AttributeError in the same code path (`_patch_mistral_regex`), same file, same fix direction, and they could plausibly be merged into one PR; 45317 simply adds 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": 35, + "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": "541843e562fa32b789ea91882acd66c9659cc35cc6aaeeac3e1c45c0de9bbdd2", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 6064, + "estimated_input_tokens": 1516, + "estimated_eval_tokens": 3288 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 6064, + "estimated_input_tokens": 1516, + "estimated_eval_tokens": 3288 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Two open NemotronH-related PRs overlap on model registry/config files, but they target different changes: one is a refactor to inherit GraniteMoeHybrid-compatible behavior, the other is a broader model-family split adding NemotronH dense/sparse docs and implementations.", + "confidence": 0.91, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR #45569 is the broader, more representative NemotronH change set: it introduces the dense/sparse model family, updates docs, and touches the central auto/model wiring, making it the better cluster representative.", + "best_issue_reason": null, + "best_pr_reason": "PR #45569 is the best overall candidate because it covers the larger concrete feature/change surface and appears closer to the final intended architecture than the narrower refactor in #45115.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45115", + "right": "pull_request:45569", + "accept": false, + "reason": "They share NemotronH infrastructure files, but the actual changes differ: #45115 is a conversion/refactor toward GraniteMoeHybrid-compatible weight naming, while #45569 splits NemotronH into dense/sparse variants and adds docs/model registry updates. Same subsystem, 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": 20, + "component_count": 35, + "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": "e97b6d069fab3669d692330d4fc4feba0413a0d96ae828f92ee3ddd752af9f6e", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 6090, + "estimated_input_tokens": 1523, + "estimated_eval_tokens": 3302 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 6090, + "estimated_input_tokens": 1523, + "estimated_eval_tokens": 3302 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Two overlapping PRs in the heterogeneous-modeling area, but they are not the same change: one is an end-to-end heterogeneous model support PR, the other is a configuration-only subset.", + "confidence": 0.94, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR #45332 is the broader, more complete change: it adds heterogeneous modeling support plus cache, masking, integration, and test updates, so it best represents the cluster.", + "best_issue_reason": null, + "best_pr_reason": "PR #45332 is the best representative because it covers the full feature scope, while #45333 is limited to per-layer configuration utilities and docs.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45332", + "right": "pull_request:45333", + "accept": false, + "reason": "They share docs and config files, but #45332 implements full heterogeneous model support (modeling, masking, cache, integrations) whereas #45333 only adds configuration support. Different concrete changes, so not a duplicate PR pair." + } + ] + }, + "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": 35, + "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": "2093ad51836a73cd385e362d8b446b7459a937d953c8c29e2f0ddc40902e28f5", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 5514, + "estimated_input_tokens": 1379, + "estimated_eval_tokens": 3014 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 5514, + "estimated_input_tokens": 1379, + "estimated_eval_tokens": 3014 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are two different typing-focused PRs: one is a broad mechanical type-check cleanup across many transformer modules, and the other adds new typing protocols/aliases for modeling_utils, PEFT, and related helpers.", + "confidence": 0.9, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 45425 is the more coherent, self-contained change set with concrete typing API additions; PR 45415 is a broader mechanical enablement pass across many files.", + "best_issue_reason": null, + "best_pr_reason": "PR 45425 is the best representative of the cluster because it makes targeted typing/interface changes rather than a wide lint/config sweep.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45415", + "right": "pull_request:45425", + "accept": false, + "reason": "Both are about typing, but they address different underlying changes: 45415 is broad type-check cleanup and config relaxation, while 45425 introduces specific typing protocols/aliases for model and PEFT integration. They would not plausibly merge into one PR without becoming a much larger unrelated bundle." + } + ] + }, + "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": 35, + "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": "c15cf677d62616d12c453f71657d0160dc1a31012ab1ebbfeadfd4a3bfbd0a9d", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4507, + "estimated_input_tokens": 1127, + "estimated_eval_tokens": 2510 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4507, + "estimated_input_tokens": 1127, + "estimated_eval_tokens": 2510 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Both PRs concern expert parallelism, but they fix different failure modes: one is about EP config/ZeRO-3 loading via accelerate, the other about FSDP2 expert parameters being overwritten during load. They share a subsystem, not the same bug or change.", + "confidence": 0.94, + "canonical_issue_reason": null, + "canonical_pr_reason": "No single canonical PR for duplication: these are separate EP loading bugs with different triggers and fixes (DeepSpeed ZeRO-3/config serialization vs FSDP2 rank-0 broadcast overwrite).", + "best_issue_reason": null, + "best_pr_reason": "45662 is the better representative of the cluster because it fixes a concrete, narrowly defined EP loading corruption path with explicit code-path changes in the loader and MoE handling.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45548", + "right": "pull_request:45662", + "accept": false, + "reason": "Reject: same broad EP area, but the PRs target different concrete bugs and loading paths; they would not plausibly merge into one change without conflating separate fixes." + } + ] + }, + "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": 35, + "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": "76c4d9b1ab6f9f7c1efeb5ba69e045777ca380f7062e20e01c5c9e246460b25a", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4672, + "estimated_input_tokens": 1168, + "estimated_eval_tokens": 2592 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4672, + "estimated_input_tokens": 1168, + "estimated_eval_tokens": 2592 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Both PRs target the same xdist-related captured_info logging collision around the testing debug artifacts, but PR 45645 is the broader, more complete fix because it also updates CI log collection to handle multiple per-worker files.", + "confidence": 0.94, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 45645 is the better canonical PR because it covers the underlying xdist-safe artifact handling end-to-end: the runtime filename fix plus the CI workflow update needed to read multiple captured_info files.", + "best_issue_reason": null, + "best_pr_reason": "PR 45645 is the strongest representative of the cluster since it addresses the same concrete failure mode as 45639 and additionally preserves the debug logs in CI, making it the more complete fix.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45639", + "right": "pull_request:45645", + "accept": true, + "reason": "Both PRs fix the same captured_info xdist collision problem in the testing debug logging path; 45645 appears to subsume 45639 and could plausibly be merged as one combined 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": 24, + "component_count": 35, + "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-45720-2", + "nodes": [ + "pull_request:45720", + "pull_request:46259" + ], + "soft_pairs": [ + "pull_request:45720|pull_request:46259" + ], + "prepared_review_unit_hash": "2a4cb0f8d4a50e298434bd6139d58e5df35d189559395a962a3339827a20d2c8", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 7234, + "estimated_input_tokens": 1809, + "estimated_eval_tokens": 3874 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 7234, + "estimated_input_tokens": 1809, + "estimated_eval_tokens": 3874 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "The two PRs overlap in several transformer model files, but they address different linter rules and different behaviors: 45720 updates forward return handling for TRF003, while 46259 changes weight initialization logic for TRF018. They are not duplicates.", + "confidence": 0.98, + "canonical_issue_reason": null, + "canonical_pr_reason": "No canonical duplicate PR: 45720 targets forward/return_dict behavior, while 46259 targets _init_weights initialization behavior. The overlapping filenames are incidental.", + "best_issue_reason": null, + "best_pr_reason": "46259 is the broader and more substantial PR overall, but it is not a duplicate of 45720.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45720", + "right": "pull_request:46259", + "accept": false, + "reason": "Different underlying changes and code paths: TRF003 modifies forward return handling, while TRF018 modifies weight initialization. Shared model files do not make them the same 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": 25, + "component_count": 35, + "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": 26, + "component_count": 35, + "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-45932-2", + "nodes": [ + "pull_request:45932", + "pull_request:46100" + ], + "soft_pairs": [ + "pull_request:45932|pull_request:46100" + ], + "prepared_review_unit_hash": "c2cee72e260cfd846889ce2d4e492dfed15202861e1173872863eef2542adecc", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 5161, + "estimated_input_tokens": 1291, + "estimated_eval_tokens": 2838 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 5161, + "estimated_input_tokens": 1291, + "estimated_eval_tokens": 2838 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are both open PRs touching testing-related areas, but they address different changes: one is a mixed CI/model patch, the other is a focused test-migration PR. They are not the same underlying change.", + "confidence": 0.94, + "canonical_issue_reason": null, + "canonical_pr_reason": "pull_request:46100 is the more coherent and self-contained change set; pull_request:45932 mixes CI secrets and multiple model/config edits without a single clear theme.", + "best_issue_reason": null, + "best_pr_reason": "pull_request:46100 is the best representative of the cluster because it has a narrowly defined purpose (test migration) and a clearer mergeable scope than pull_request:45932.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45932", + "right": "pull_request:46100", + "accept": false, + "reason": "Both mention testing, but 45932 is a mixed grafana/CI/model patch while 46100 is a dedicated test migration; they do not appear to fix the same concrete problem and would not plausibly be merged as 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": 27, + "component_count": 35, + "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-46126-2", + "nodes": [ + "pull_request:46126", + "pull_request:46269" + ], + "soft_pairs": [ + "pull_request:46126|pull_request:46269" + ], + "prepared_review_unit_hash": "73ea664d7038f01ef1181390ccf390bd83e76aa9ce60c89ba2ad17b02210e6d2", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 5526, + "estimated_input_tokens": 1382, + "estimated_eval_tokens": 3020 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 5526, + "estimated_input_tokens": 1382, + "estimated_eval_tokens": 3020 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are related distributed-training pull requests, but not the same underlying change: one is a focused EP + Trainer integration, while the other is a broad distributed branch base covering many subsystems. The shared files are not enough to treat them as duplicates.", + "confidence": 0.83, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 46269 is the broader umbrella branch base touching many distributed components, so it is the better canonical representative of this cluster than the narrower EP/Trainer integration PR.", + "best_issue_reason": null, + "best_pr_reason": "PR 46269 is the best representative because it is the more comprehensive distributed baseline and subsumes a wider set of related changes.", + "soft_edge_verdicts": [ + { + "left": "pull_request:46126", + "right": "pull_request:46269", + "accept": false, + "reason": "Reject: they overlap in distributed infrastructure files, but 46126 is a specific EP + Trainer integration change while 46269 is a broad distributed branch base with many unrelated edits. They do not look like the same concrete fix and would not plausibly be merged as 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": 28, + "component_count": 35, + "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-46138-10", + "nodes": [ + "pull_request:46138", + "pull_request:46432", + "pull_request:46501", + "pull_request:46502", + "pull_request:46503", + "pull_request:46504", + "pull_request:46554", + "pull_request:46558", + "pull_request:46563", + "pull_request:46716" + ], + "soft_pairs": [ + "pull_request:46138|pull_request:46554", + "pull_request:46138|pull_request:46558", + "pull_request:46138|pull_request:46563", + "pull_request:46138|pull_request:46716", + "pull_request:46138|pull_request:46504", + "pull_request:46432|pull_request:46503", + "pull_request:46432|pull_request:46501", + "pull_request:46432|pull_request:46502", + "pull_request:46432|pull_request:46554", + "pull_request:46432|pull_request:46558", + "pull_request:46501|pull_request:46503", + "pull_request:46501|pull_request:46502", + "pull_request:46501|pull_request:46504", + "pull_request:46501|pull_request:46554", + "pull_request:46502|pull_request:46503", + "pull_request:46502|pull_request:46504", + "pull_request:46502|pull_request:46554", + "pull_request:46503|pull_request:46504", + "pull_request:46503|pull_request:46554", + "pull_request:46504|pull_request:46554", + "pull_request:46504|pull_request:46558", + "pull_request:46554|pull_request:46558", + "pull_request:46554|pull_request:46563", + "pull_request:46554|pull_request:46716", + "pull_request:46558|pull_request:46563", + "pull_request:46558|pull_request:46716", + "pull_request:46563|pull_request:46716", + "pull_request:46504|pull_request:46563", + "pull_request:46432|pull_request:46563", + "pull_request:46504|pull_request:46716", + "pull_request:46432|pull_request:46716" + ], + "prepared_review_unit_hash": "42358afa722ee686b323565b3f3ae86afef7a8cb9c2165b1caf7c1167499d2b4", + "component_budget": { + "node_count": 10, + "item_count": 10, + "soft_pair_count": 31, + "serialized_chars": 26674, + "estimated_input_tokens": 6669, + "estimated_eval_tokens": 13594 + }, + "budget": { + "node_count": 10, + "item_count": 10, + "soft_pair_count": 31, + "serialized_chars": 26674, + "estimated_input_tokens": 6669, + "estimated_eval_tokens": 13594 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "No true duplicates here. The PRs fall into separate workflow-maintenance changes, and the pr-ci-security-gate.yml edits are related in theme but implement different security-hardening fixes, so I would not merge them as duplicates.", + "confidence": 0.74, + "canonical_issue_reason": null, + "canonical_pr_reason": "If one PR must represent the related security-gate subgroup, 46554 is the best candidate because it makes the most concrete self-contained code-path change (checkout plus diff-based checks), while the others are separate workflow adjustments.", + "best_issue_reason": null, + "best_pr_reason": "46554 is the strongest standalone PR in the only partially related subgroup: it addresses a specific security-gate behavior with an explicit logic change and clearer implementation than the follow-up permission/secret/reusable-workflow tweaks.", + "soft_edge_verdicts": [ + { + "left": "pull_request:46138", + "right": "pull_request:46554", + "accept": false, + "reason": "Different workflow areas: self-comment CI command piping vs pr-ci security gate logic." + }, + { + "left": "pull_request:46138", + "right": "pull_request:46558", + "accept": false, + "reason": "Different workflow changes in unrelated files and code paths." + }, + { + "left": "pull_request:46138", + "right": "pull_request:46563", + "accept": false, + "reason": "Unrelated workflow update; no shared concrete bug or fix." + }, + { + "left": "pull_request:46138", + "right": "pull_request:46716", + "accept": false, + "reason": "Self-comment CI update vs pr-ci security-gate secrets handling; not the same change." + }, + { + "left": "pull_request:46138", + "right": "pull_request:46504", + "accept": false, + "reason": "Different workflow files with different behavior; not a duplicate fix." + }, + { + "left": "pull_request:46432", + "right": "pull_request:46503", + "accept": false, + "reason": "SSH runner hardening vs benchmark trigger change; separate workflow concerns." + }, + { + "left": "pull_request:46432", + "right": "pull_request:46501", + "accept": false, + "reason": "SSH runner update and anti-slop permissions change are unrelated." + }, + { + "left": "pull_request:46432", + "right": "pull_request:46502", + "accept": false, + "reason": "Different workflows and purposes; no same underlying bug." + }, + { + "left": "pull_request:46432", + "right": "pull_request:46554", + "accept": false, + "reason": "SSH runner environment cleanup vs security-gate logic; distinct changes." + }, + { + "left": "pull_request:46432", + "right": "pull_request:46558", + "accept": false, + "reason": "SSH runner update vs security-gate permissions/approval tweak; not duplicates." + }, + { + "left": "pull_request:46501", + "right": "pull_request:46503", + "accept": false, + "reason": "Both are workflow edits, but they change different jobs and behavior." + }, + { + "left": "pull_request:46501", + "right": "pull_request:46502", + "accept": false, + "reason": "Anti-slop permissions vs assign-reviewers checkout ref; unrelated fixes." + }, + { + "left": "pull_request:46501", + "right": "pull_request:46504", + "accept": false, + "reason": "Different workflow files and unrelated logic changes." + }, + { + "left": "pull_request:46501", + "right": "pull_request:46554", + "accept": false, + "reason": "Anti-slop permission scoping vs security-gate checkout/diff checks; separate issues." + }, + { + "left": "pull_request:46502", + "right": "pull_request:46503", + "accept": false, + "reason": "Assign-reviewers checkout behavior vs benchmark trigger condition; different changes." + }, + { + "left": "pull_request:46502", + "right": "pull_request:46504", + "accept": false, + "reason": "Unrelated workflow updates in different files." + }, + { + "left": "pull_request:46502", + "right": "pull_request:46554", + "accept": false, + "reason": "Different workflow code paths; not the same bug or fix." + }, + { + "left": "pull_request:46503", + "right": "pull_request:46504", + "accept": false, + "reason": "Benchmark trigger logic vs build-image commit-message parsing; separate changes." + }, + { + "left": "pull_request:46503", + "right": "pull_request:46554", + "accept": false, + "reason": "Benchmark workflow trigger update vs security-gate enforcement; unrelated." + }, + { + "left": "pull_request:46504", + "right": "pull_request:46554", + "accept": false, + "reason": "Build-image tag detection vs pr-ci security-gate logic; not the same change." + }, + { + "left": "pull_request:46504", + "right": "pull_request:46558", + "accept": false, + "reason": "Different workflow logic: commit-message handling vs permissions/approval gating." + }, + { + "left": "pull_request:46554", + "right": "pull_request:46558", + "accept": false, + "reason": "Same file, but different concrete fixes: 46554 adds checkout/diff checks; 46558 tightens permissions and approval flow." + }, + { + "left": "pull_request:46554", + "right": "pull_request:46563", + "accept": false, + "reason": "Different fixes in the same workflow area: one changes the security check implementation, the other swaps to a reusable workflow and adds notes." + }, + { + "left": "pull_request:46554", + "right": "pull_request:46716", + "accept": false, + "reason": "46554 hardens the security gate logic; 46716 changes secrets passing on the reusable workflow. Different code-path problems." + }, + { + "left": "pull_request:46558", + "right": "pull_request:46563", + "accept": false, + "reason": "Permissions/approval gating vs reusable-workflow substitution; distinct changes." + }, + { + "left": "pull_request:46558", + "right": "pull_request:46716", + "accept": false, + "reason": "Permissions and approval guard vs secrets inheritance removal; not the same underlying fix." + }, + { + "left": "pull_request:46563", + "right": "pull_request:46716", + "accept": false, + "reason": "Reusable workflow call/comment change vs secrets handling change; separate PRs." + }, + { + "left": "pull_request:46504", + "right": "pull_request:46563", + "accept": false, + "reason": "Build-image workflow change vs pr-ci-security-gate reusable workflow note; unrelated." + }, + { + "left": "pull_request:46432", + "right": "pull_request:46563", + "accept": false, + "reason": "SSH runner hardening vs security-gate reusable-workflow change; different subsystems and fixes." + }, + { + "left": "pull_request:46504", + "right": "pull_request:46716", + "accept": false, + "reason": "Build-ci docker workflow update vs security-gate secrets change; not duplicates." + }, + { + "left": "pull_request:46432", + "right": "pull_request:46716", + "accept": false, + "reason": "SSH runner update vs security-gate secrets handling; separate changes." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 29, + "component_count": 35, + "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-46174-2", + "nodes": [ + "pull_request:46174", + "pull_request:46780" + ], + "soft_pairs": [ + "pull_request:46174|pull_request:46780" + ], + "prepared_review_unit_hash": "a48ce4ca162f555682f79411d7229ac1ccc08d7822dfab654bb881639fccddbc", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 6030, + "estimated_input_tokens": 1508, + "estimated_eval_tokens": 3272 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 6030, + "estimated_input_tokens": 1508, + "estimated_eval_tokens": 3272 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Two PRs target the same reported MPS crash in the sinusoidal position embedding helper, but they implement different fixes and are not safe to treat as duplicates.", + "confidence": 0.89, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 46174 is the broader, more complete fix for the float64-on-MPS crash: it updates the helper across all affected model files and keeps the internal math on CPU float64 before casting back.", + "best_issue_reason": null, + "best_pr_reason": "PR 46174 is the best representative of the cluster because it is the original, more comprehensive fix and has more discussion/review context than the narrower follow-up.", + "soft_edge_verdicts": [ + { + "left": "pull_request:46174", + "right": "pull_request:46780", + "accept": false, + "reason": "They share the same issue target and helper, but the patches are different solutions: 46174 avoids float64 on MPS by doing the arithmetic on CPU first, while 46780 switches the helper to use dtype for internal arithmetic. Those are alternative implementations, not one mergeable 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": 30, + "component_count": 35, + "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-46281-2", + "nodes": [ + "pull_request:46281", + "pull_request:46890" + ], + "soft_pairs": [ + "pull_request:46281|pull_request:46890" + ], + "prepared_review_unit_hash": "5cd5b95aae12c0bd33443866e33fbf581bbac5dbf0e70a8267a2b7e8056754f8", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4278, + "estimated_input_tokens": 1070, + "estimated_eval_tokens": 2396 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4278, + "estimated_input_tokens": 1070, + "estimated_eval_tokens": 2396 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Two security-related PRs, but they address different path-traversal entry points in different loader code paths. They are thematically similar, not duplicates.", + "confidence": 0.96, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR #46890 is the better cluster representative because it targets a clearly defined, config-driven path-traversal check in the core checkpoint resolution flow and has a self-contained fix in `modeling_utils.py`.", + "best_issue_reason": null, + "best_pr_reason": "PR #46890 is the strongest representative for this cluster: it fixes a concrete unsafe file-path resolution bug, has a clear rejection condition, and is easier to describe as the canonical security hardening change.", + "soft_edge_verdicts": [ + { + "left": "pull_request:46281", + "right": "pull_request:46890", + "accept": false, + "reason": "Different bugs and code paths: #46281 guards sharded-checkpoint filenames from the index `weight_map` in `hub.py`, while #46890 validates `transformers_weights` / explicit checkpoint filenames in `modeling_utils.py`. Same broad path-traversal theme, but not the same underlying fix and not mergeable as 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": 31, + "component_count": 35, + "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-46310-2", + "nodes": [ + "pull_request:46310", + "pull_request:47010" + ], + "soft_pairs": [ + "pull_request:46310|pull_request:47010" + ], + "prepared_review_unit_hash": "3e01bc3b8ecef7c3a9e1f9cd452c02a0146787169c22b819133f5c18bb1a4f92", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4792, + "estimated_input_tokens": 1198, + "estimated_eval_tokens": 2652 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4792, + "estimated_input_tokens": 1198, + "estimated_eval_tokens": 2652 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Both pull requests fix the same off-by-one bug in the sliding window mask\u2019s right boundary (`-dist <= right_window_size` instead of `<`). PR 47010 is the broader, more representative patch because it applies the fix in both Gemma4 and Moonshine Streaming, while 46310 covers only Moonshine Streaming.", + "confidence": 0.98, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 47010 is the best canonical PR because it addresses the same concrete boundary bug and includes the fuller cross-model fix, making it the most representative version of the change.", + "best_issue_reason": null, + "best_pr_reason": "PR 47010 is the strongest standalone candidate: it fixes the exact code-path bug and covers all shown affected locations, whereas PR 46310 is a narrower subset.", + "soft_edge_verdicts": [ + { + "left": "pull_request:46310", + "right": "pull_request:47010", + "accept": true, + "reason": "Accept: both PRs fix the same off-by-one in `sliding_window_mask_function` right-window inclusion on the same code path; 47010 is essentially a superset of 46310, so they plausibly merge into one 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": 32, + "component_count": 35, + "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-46376-2", + "nodes": [ + "pull_request:46376", + "pull_request:46417" + ], + "soft_pairs": [ + "pull_request:46376|pull_request:46417" + ], + "prepared_review_unit_hash": "4f4efbb4b257021dde9b3f12a4185f39f438656fa42d5fe00bf406b02d966dbe", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 5959, + "estimated_input_tokens": 1490, + "estimated_eval_tokens": 3236 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 5959, + "estimated_input_tokens": 1490, + "estimated_eval_tokens": 3236 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "The two pull requests are unrelated: one implements multimodal support for continuous batching, while the other adds the HunYuan VL model stack and docs. They do not target the same bug or code change and should not be deduplicated.", + "confidence": 0.98, + "canonical_issue_reason": null, + "canonical_pr_reason": "No clear canonical PR for a duplicate cluster; if forced, PR #46376 is the more infrastructure-focused change, but it is not a duplicate of PR #46417.", + "best_issue_reason": null, + "best_pr_reason": "PR #46376 is the better representative of a concrete runtime code-path change; PR #46417 is a broader model-addition/documentation PR and is clearly separate.", + "soft_edge_verdicts": [ + { + "left": "pull_request:46376", + "right": "pull_request:46417", + "accept": false, + "reason": "Different underlying work: #46376 adds multimodal support to continuous batching internals, while #46417 adds the HunYuan VL model, processors, docs, and auto-mapping entries. They do not share a concrete code-path fix and would not plausibly 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": 33, + "component_count": 35, + "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-46574-2", + "nodes": [ + "pull_request:46574", + "pull_request:46580" + ], + "soft_pairs": [ + "pull_request:46574|pull_request:46580" + ], + "prepared_review_unit_hash": "7fb932b4185f683ed2dac3b2780642ef377d33283eb4ebff815e56541ff5a2e9", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4410, + "estimated_input_tokens": 1103, + "estimated_eval_tokens": 2462 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4410, + "estimated_input_tokens": 1103, + "estimated_eval_tokens": 2462 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are two unrelated PRs: #46574 fixes CodeLlama decode stripping leading spaces, while #46580 fixes DeepSeek-Coder/DeepSeek-LLM being assigned the wrong tokenizer class in AutoTokenizer. They do not look like the same underlying bug or change.", + "confidence": 0.99, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": "#46574 is the tighter representative of its bugfix class: it has a single concrete tokenizer-decoding issue, an explicit issue link, and a narrowly scoped code/test change.", + "soft_edge_verdicts": [ + { + "left": "pull_request:46574", + "right": "pull_request:46580", + "accept": false, + "reason": "Different failure modes and code paths: one changes CodeLlama decoder setup to preserve leading spaces, the other changes AutoTokenizer checkpoint-prefix routing for DeepSeek models. They are not mergeable as one 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": 34, + "component_count": 35, + "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-46847-2", + "nodes": [ + "pull_request:46847", + "pull_request:47009" + ], + "soft_pairs": [ + "pull_request:46847|pull_request:47009" + ], + "prepared_review_unit_hash": "66783a7c52659440be3dd8a54877ab0ee59de25c65133baeffcce0459576c028", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4639, + "estimated_input_tokens": 1160, + "estimated_eval_tokens": 2576 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4639, + "estimated_input_tokens": 1160, + "estimated_eval_tokens": 2576 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Both PRs make the same Qwen3.5 tensor-parallel plan update for linear_attn, touching the same files with the same added mappings and the same tracked issue. 46847 appears to be the earlier canonical duplicate.", + "confidence": 0.99, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 46847 is the earlier instance of the identical fix, with the same target issue, same files, and the same configuration changes.", + "best_issue_reason": null, + "best_pr_reason": "PR 46847 is the best representative because it is earlier and has the more established review/inbound activity while implementing the same concrete code-path change as 47009.", + "soft_edge_verdicts": [ + { + "left": "pull_request:46847", + "right": "pull_request:47009", + "accept": true, + "reason": "Same underlying change: identical Qwen3.5 linear_attn tp-plan entries in the same files, with the same issue target and matching diff shape." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 35, + "component_count": 35, + "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-46993-2", + "nodes": [ + "pull_request:46993", + "pull_request:46994" + ], + "soft_pairs": [ + "pull_request:46993|pull_request:46994" + ], + "prepared_review_unit_hash": "10cdd3de01bfd45178654e7f3fda490993621f3a5e16ca6120d4a398feb1005e", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4309, + "estimated_input_tokens": 1078, + "estimated_eval_tokens": 2412 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4309, + "estimated_input_tokens": 1078, + "estimated_eval_tokens": 2412 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These two pull requests are not duplicates: one fixes prompt lookup decoding exceeding max_length, while the other tightens validation for empty bad_words_ids / sequence_bias inputs in logits processing.", + "confidence": 0.98, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": "Neither PR is a good canonical representative because they address unrelated bugs in different generation subsystems and would not plausibly be merged into one change.", + "soft_edge_verdicts": [ + { + "left": "pull_request:46993", + "right": "pull_request:46994", + "accept": false, + "reason": "Different concrete problems and code paths: PR 46993 changes candidate generation length capping; PR 46994 adds argument validation for empty token lists in logits processors. They are not the same underlying bug or a single mergeable fix." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + } + ] +} diff --git a/snapshots/20260702T060019Z/analysis-runs/hybrid-model-20260702t060019z/manifest.json b/snapshots/20260702T060019Z/analysis-runs/hybrid-model-20260702t060019z/manifest.json new file mode 100644 index 0000000000000000000000000000000000000000..d2a263093894cdd6ffa7045241e46a5d80933240 --- /dev/null +++ b/snapshots/20260702T060019Z/analysis-runs/hybrid-model-20260702t060019z/manifest.json @@ -0,0 +1,14 @@ +{ + "analysis_id": "hybrid-model-20260702t060019z", + "artifacts": { + "hybrid": "snapshots/20260702T060019Z/analysis-runs/hybrid-model-20260702t060019z/analysis-report-hybrid.json", + "hybrid_reviews": "snapshots/20260702T060019Z/analysis-runs/hybrid-model-20260702t060019z/analysis-report-hybrid.llm-reviews.json" + }, + "channel": "canonical", + "model": null, + "published_at": "2026-07-02T06:11:20Z", + "repo": "huggingface/transformers", + "schema_version": 1, + "snapshot_id": "20260702T060019Z", + "variant": "hybrid" +} diff --git a/snapshots/20260702T060019Z/manifest.json b/snapshots/20260702T060019Z/manifest.json index 1b6e70fef95cf6817e905ab85aa35180d9d0579e..0e6b2de7d0827d89236a44779db9dcc2d40f5423 100644 --- a/snapshots/20260702T060019Z/manifest.json +++ b/snapshots/20260702T060019Z/manifest.json @@ -32,6 +32,24 @@ "timeline_events": 673 }, "extracted_at": "2026-07-02T06:00:19Z", + "published_analysis": { + "canonical_analysis_id": "hybrid-model-20260702t060019z", + "runs": { + "hybrid-model-20260702t060019z": { + "analysis_id": "hybrid-model-20260702t060019z", + "artifacts": { + "hybrid": "snapshots/20260702T060019Z/analysis-runs/hybrid-model-20260702t060019z/analysis-report-hybrid.json", + "hybrid_reviews": "snapshots/20260702T060019Z/analysis-runs/hybrid-model-20260702t060019z/analysis-report-hybrid.llm-reviews.json" + }, + "channel": "canonical", + "manifest_path": "snapshots/20260702T060019Z/analysis-runs/hybrid-model-20260702t060019z/manifest.json", + "model": null, + "published_at": "2026-07-02T06:11:20Z", + "variant": "hybrid" + } + }, + "schema_version": 1 + }, "repo": "huggingface/transformers", "snapshot_id": "20260702T060019Z", "watermark": {